&EPA
EPA/600/R-10/182 | September 2010 | www.epa.gov/ord
United States
Environmental Protection
Agency
San Luis Basin Sustainability
Metrics Project: A Methodology
for Evaluating Regional
Sustainability
Office of Research and Development
National Risk Management Research Laboratory
-------
Front Cover Photo Credits
Top - Medano Creek; Photo M. Hopton; Intermittent flow of Medano Creek in the Great Sand Dimes National Park and Preserve,
San Luis Valley, Colorado.
Bottom Left - solar panel close-up; Photo by A. Kamnanithi; Solar panel at a solar farm under construction in the San Luis Valley.
Bottom Right - SLV Potato Harvest; Photo by Stephen Ausmus; Potatoes being harvested in the San Luis Valley of south-central
Colorado. Rotating potatoes with cover crops provides many benefits, including nitrogen management, improved soil and water
quality, and bigger potatoes and higher yields, http://www.ars.usda.
-------
EPA/600/R-10/182
September 2010
A
for
by
Matthew T. Heberling
Matthew E. Hopton
Authors
Heriberto Cabezas
Daniel Campbell
Tarsha Eason
Ahjond S. Garmestani
Matthew T. Heberling
Matthew E. Hopton
Joshua Templeton
Denis White
Marie Zanowick
National Risk Management Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
and
Richard T. Sparks
U.S. Department of Agriculture
Office of Research and Development
National Risk Management Research Laboratory
-------
The views expressed herein are strictly the opinions of the authors and in no manner represent or
reflect current or planned policy by the federal agencies. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use. The information and data
presented in this product were obtained from sources that are believed to be reliable. However, in
many cases the quality of the information or data was not documented by those sources; therefore.
no claim is made regarding their quality.
-------
of
Notice ii
List of Tables iv
List of Figures x
List of Abbreviations xv
Foreword xvi
Authors, Contributors, and Reviewers xvii
Acknowledgements xx
Abstract xxi
Executive Summary xxii
ES.l Introduction xxii
ES.2 Sustainability Metrics xxii
ES.3 Ecological Footprint xxiii
ES.4 Green Net Regional Product xxiv
ES.5 Emergy xxiv
ES.6 Fisher Information and Order xxvi
ES.7 The San Luis Basin Sustainability Metrics Project: Results, Conclusions, and
Recommendations xxvii
ES.8 References xxix
1.0 Introduction 1
1.1 Concept of Sustainability 1
1.2 The San Luis Basin Sustainability Metrics Project 2
1.2.1 Measuring Sustainability 2
1.2.2 Choice of Region 3
1.2.2.1 Description of the San Luis Basin 4
1.2.3 Community Outreach 4
1.3 Outline of Report 6
1.4 References 6
2.0 Sustainability Metrics.
2.1 Introduction
-------
2.2 Literature Review of Ecological. Economic. Energy, and System Order Metrics 9
2.2.1 Ecological Sustainability 9
2.2.2 Economic Metrics 10
2.2.3 Energy 12
2.2.3.1 Exergy Analysis 12
2.2.3.2 Emergy Analysis 12
2.2.4 System Order 13
2.3 Sustainability Studies That Use Multiple Metrics 14
2.4 Choice of Metrics for San Luis Basin Sustainability Metrics Project 15
2.4.1 Components Basic to Environmental Systems 16
2.5 References 17
3.0 Ecological Footprint 25
3.1 Introduction 25
3.2 Methods 26
3.3 Data, Variables, and Sources 26
3.4 Results 27
3.5 Discussion 28
3.5.1 Biocapacity (supply) 28
3.5.2 Ecological Footprint (demand) 28
3.5.3 Global Hectares 28
3.5.4 Examination of Sustainability 29
3.5.5 Strengths and Weaknesses of Methodology 29
3.6 References 30
Appendix 3-A: Estimation of Energy Use and Determination to Use E1A State Level Data 39
Appendix 3 -B: Ecological Footprint Accounting Ledger for 1980 in the San Luis Basin 42
Appendix 3-C: Twenty-six Years (1980-2005) of Data for the San Luis Basin, Colorado 44
Appendix 3 -D: Ecological Footprint Analysis for the United States, 1980-2005 51
4.0 Green Net Regional Product 53
4.1 Introduction 53
4.2 Model 53
4.3 Data and Sources 53
4.4 Calculation Methodology 55
4.5 Results 56
4.6 Discussion 57
4.7 Conclusion 59
-------
4.8 References 59
Appendix 4-A: Wind Erosion Calculations 67
5.0 Emergy 77
5.1 Introduction 77
5.2. Methods 78
5.2.1 Boundaries 78
5.2.2. Energy Systems Models 78
5.2.3 Emergy Analysis 78
5.2.4. Data and Sources 79
5.2.5 Summary Tables and Indices 80
5.3 Results 80
5.3.1 The Detailed Energy Systems Model of the San Luis Basin 80
5.3.2 Aggregate Diagram 80
5.3.3 Total Emergy and Percent Renewable Use 80
5.3.4 Summary of Renewable and Non-renewable Emergy 81
5.3.5 Summary of Import-Export Exchange 81
5.3.6 Imported Emergy 81
5.3.7 Exported Emergy 81
5.3.8 Population and per Capita Indices 82
5.4 Discussion 82
5.4.1 Spatial Boundaries 82
5.4.2 Data Availability 82
5.4.3 Renewable Inputs to the SLB 83
5.4.4 Temporal Patterns of Development and the Water Cycle 83
5.4.5 Nonrcncwable Resources 83
5.4.6 Patterns of Import and the Emergy Budget of the SLB 83
5.4.7 Patterns of Export 84
5.4.8 Export-Import Exchange 84
5.4.9 Patterns of Population Growth and Consumption 84
5.4.10 Analysis of the Emergy Indices 85
5.4.11 Effects of Missing Data, Unknown Information, and Uncertainty 85
5.4.12 Strengths and Limitations of the Method 86
5.5 Conclusions 86
5.6 References 87
Appendix 5-A: Detailed Steps for Emergy Analysis 113
Appendix 5-B: Additional Indices Calculated for the Emergy Analysis of the
San Luis Basin 114
Appendix 5-C: Calculating Emergy per Unit Factors 118
Appendix 5-D: Further Description of the San Luis Basin Regional Model 139
-------
Appendix 5-E: Emergy Analysis Tables For Renewable Inflows, Renewable Production.
Nonrenewable Productions And Use, Imports And Exports, And Additional Figures Not
Included In Body of Report 141
Appendix 5-F: Data Quality of the Purchased Data 149
Appendix 5-G: Miscellaneous Indices and Results 150
6.0 Fisher Information and Order 163
6.1 Introduction 163
6.1.1 Capturing Dynamic Shifts with Fisher Information 163
6.2 Methods 164
6.2.1 Theory 164
6.2.2 Calculation Methodology 164
6.2.3 Interpreting Fisher Information 166
6.2.4 Computing Fisher Information: A Simple Example 167
6.3 Computing FI for the San Luis Basin: Data and Sources 168
6.4 Calculation Methodology 168
6.5 Results 168
6.6 Discussion 169
6.7 Strengths and Weaknesses 169
6.8 References 170
Appendix 6-A: Derivation of Fisher Information 183
Appendix 6-B: Sensitivity Analysis on Integration Window Parameters 186
Appendix 6-C: San Luis Basin Fisher Code User Manual - Utilizing MATLAB Files
(Nfisher And CalcMeanFisher) to Calculate Fisher Information 192
Appendix 6-D: Using MATLAB to Compute FI for the Simple Example 196
Appendix 6-E: Fisher Information MATLAB Code 198
7.0 The San Luis Basin Sustainability Metrics Project: Results, Conclusions, and
Recommendations 223
7.1 Introduction 223
7.2 Summary of Calculations 223
7.3 Holistic Results for San Luis Basin 223
7.4 Strengths of Methodology 225
7.5 Limitations of Methodology 226
7.6 Recommendations for Future Research on Regional Sustainability Metrics 228
7.7 Conclusions 228
7.8 References 229
-------
Table 2.1- Broad categorical list of data collected on key variables, showing which metric used the
variable, the source of the data, the scale at which the data were collected, and the available years of
data reported 23
Table 3.1- Scale and source of data for the variables used to calculate ecological footprint for the
SLB 33
Table 3.2 - Ecological footprint and biocapacity in the San Luis Basin, Colorado, 1980-2005.
Twelve percent of the biocapacity area was subtracted from the total biocapacity to protect
biodiversity 35
Table 3-B. 1 - Ecological footprint accounting ledger for food consumption for 1980 in the San Luis
Basin. Local Yield and Production are italicized because they were not needed to compute apparent
consumption. Global yields arc from Chambers et al. (2000) 42
Table 3-B.2 - Ecological footprint accounting ledger for energy consumption for 1980 in the San
Luis Basin. Global average energy to land ratio was used to convert joules to hectares (Chambers et
al. 2000) 43
Table 3-B.3 - Ecological footprint accounting ledger for supply and demand of bioproductive land
for 1980 in the San Luis Basin. Equivalence factors are from Mclntyre et al. (2007) and Chambers
et al. (2000) and were treated as constants for the 26 years 43
Table 3-C.l - Twenty-six years (1980-2005) of data for the San Luis Basin, Colorado. Italicized
values were estimated using linear interpolation 44
Table 4.1 - Green Net Regional Product (GNRP): Variables, Purpose, and Data Sources 61
Table 4.2 - Estimates of all components of Green Net Regional Product (GNRP) 64
Table 4-A. 1 - Approximate distribution of crops between the two soil types (Based on Soil
Erodibility Index, I) 69
Table 4-A.2 - Crop rotations included in soil erosion calculation 69
Table 4-A.3 -Total acreage by crop rotation and Soil Erodibility Index 70
Table 4-A.4 - Average annual wind erosion by crop rotation: practicing mulch or noninvcrsion
tillage 72
Table 4-A.5 - Average annual wind erosion by crop rotation: conventional, moldboard plow 72
Table 4-A.6 - Total Erosion (tons/year) for the SLB and Tons per Acre per Year 73
Table 5.1- Tabular format for an emergy analysis, providing a template for the calculation of the
emergy values 89
Table 5.2 - Definition of pathway flows for the systems model of the San Luis Basin (seven county
area), Colorado shown in Fig. 5.1 89
-------
Table 5.3 - Summary of annual emergy flows in the San Luis Basin, 1995-2005 92
Table 5.4 - Emergy indices and indicators for the San Luis Basin, 1995-2005. Our focus for this
study was on Item 106 98
Table 5-B.l - Some emergy indicators and indices used in the analysis of the sustainability of
regional environmental systems. Symbols and expressions refer to Fig. 5-B.l 116
Table 5-C. 1 - Emergy per Unit Factors (solar transforinities and specific emergies) used in the San
Luis Basin study. The units for the emergy7 per unit factors are solar emjoules per joule (sej/j) for
transforinities unless otherwise noted 119
Table 5-C.2 - Specific emergies for imports (sej/g). All commodity specific emergies are without
services and relative to the 9.26E24 sej/y baseline. Values of the emergy per unit factors for the
commodity classes in this table were determined from the data given in Table 5-C. 1 and other data
given in the calculations shown in Table 5-C.3. The note numbers connect to Table 5-C.3.
Wwt = wet weight; otherwise numbers are dwt = dry weight 123
Table 5-C.3 - Determination of the specific emergies for imported commodity classes in
Table 5-C.2. All commodity specific emergies are without sendees and relative to the
9.26E24 sej/y baseline 126
Table 5-C.4 - Specific emergies for exports. All commodity specific emergies are without services
and relative to the 9.26E24 sej/y baseline. Values of the emergy per unit factors for the commodity
classes in this table were determined from the data given in Tables 5-C. 1 and 5-C.3, as well as
data provided in Table 5-C.5. The note numbers connect to Table 5-C.5. Wwt = wet weight,
otherwise numbers are dwt = dry weight. Any commodity class that is the same as that found in
the calculations for imports was determined in the same manner as shown in Table 5-C3 and the
calculation is not repeated here 133
Table 5-C.5 - Specific emergies for imports. All commodity specific emergies are without services
and relative to the 9.26E24 sej/y baseline 135
Table 5-E.l - Renewable emergy inflows to the San Luis Basin 141
Table 5-E.2 - Production in the San Luis Basin that is based primarily on renewable resources.... 142
Table 5-E.3 - Nonrenewable emergy inflows in the San Luis Basin 143
Table 5-E.4 -Imports into the San Luis Basin 144
Table 5-E.5 -Exports from the San Luis Basin 145
Table 5-G.I - Example calculation of emergy and economic value, assumptions, and data sources
used in the emergy evaluation of the San Luis Basin and Upper Rio Grande River Basin for the
year 1995 " 151
Table 6.1- Data were divided into a sequence of overlapping time windows. The eighth window
was removed because there were not enough data in the time series to place in it. Therefore, all
data were accounted for within seven windows, t = time (i.e., 1980-2005). xl = Population and x,
= Personal income (in thousands). Point 1 (ptt) was defined by (x^tj), x,(tt)) and is highlighted as
310352,38469 .". 172
-------
Table 6.2 - Absolute difference of variables Xj (population) and x2 (personal income) in each
window using the first point as the center. For example, the absolute difference of Xj(tj) and Xj(t2)
(i.e., Xj(tj) - Xj(t2)| is 4562 as highlighted here in blue and the absolute difference of point 2 from
point 1 (i.e., 2-1) is 4562 and 511) 173
Table 6.3 - Amplitude (q(s)) for each window. These values were computed by first counting the
number of points in each state within each window, computing a probability for each state in each
window (p(s)) and calculating the amplitude, q(s) = ^p(s) for each state in each window 174
Table 6.4 - Variables used to calculate FI for the San Luis Basin. These variables were selected from
data used to compute EFA, GNRP, and EmA. The data and their sources are discussed in Chapters
3, 4, and 5, respectively. The variables include demographic, energy, land use, food and forest
consumption, environmental and agricultural production categories for the system 175
Table 6.5 - Three-point mean Fisher information () for the overall system and the variables
grouped by the categories defined in Table 6.4 176
Table 6-B. 1 - Design matrix for the sensitivity analysis of the simple example data: Determining
the impact of controllable factors, hwin and winspace on the response variables, average value of FI
(avgFI) and the number of FI results (Npts) for computation 187
Table 6-B.2 -Analysis of Variance (ANOVA) forthe average value of FI (avgFI). The results reveal
that both hwin (F (3,9) = 77.56, P«0.05) and winspace (F (3,9) = 4.64, P< 0.05) were significant
factors affecting avgFI. Moreover, as an indicator of the amount of variability that is accounted for
by factor effects (e.g., hwin or winspace) and random error (i.e., residual), the MS values reflect that
more of the variability was determined by hwin 187
Table 6-B.3 -ANOVA for the resolution of Fisher information results (Npts). Like the ANOVA for
avgFI, both hwin (F (3,9) = 8.33, P O.05) and winspace (F(3,9) = 519, P «0.05) were statistically
significant factors affecting Npts. However, the mean square (MS) values indicate that majority of
the variability in Npts was due to winspace 188
Table 6-B.4 - Criteria for determining the optimal value for hwin and winspace for the sample
exercise. The goal of the numerical optimization was to minimize the deviation in avgFI and
maximize Npts, subject to (a) minimizing the standard deviation of FI, the standard error of the
mean value of FI and standard error of the Npts and (b) maximize Npts. The highest importance was
placed on maximizing Npts and minimizing the standard error of avgFI and Npts. The optimization
was performed in Design Expert 8 (StatEase, Inc.) 188
Table 6-D. 1 - Excel file containing the time series of the variables characterizing the system for
the simple example (SLBExdata.xls). In this exercise, there are two variables (xl and x2) such that
each column contains the time series data for one variable. As described in Appendix 6-C. 1, the
data file is displayed as an m x n matrix with m = the number of time steps and n = the number of
variables 196
Table 6-D.2 - Excel file containing time data for simple example (SLBExtime.xls). In this exercise, the
time period was from 1980 to 2005 and represents 26 time steps. Accordingly, as described in Appendix
6-C. 1, the time file is displayed as anm x l matrix with m = the number of time steps 197
Table 6-D.3 - Excel file containing the size of state (sost) data for simple example (SLBExsost.xls). In
this exercise, there were two variables (xl and x2) and the size of states was computed for each variable
(AXj and Ax2). The values are exhibited as a 1 x n matrix with n = the number of variables 197
Table 7.1 - Summary of four metrics calculated for the San Luis Basin, Colorado 230
-------
List of Figures
Figure 1.1- Seven county region referred to by the investigators as the San Luis Basin (SLB), in
south-central Colorado 8
Figure 3.1- Biocapacity compared to ecological footprint for 26 years in the San Luis Basin.
Twelve percent of the biocapacity area was subtracted from the total BC and be to protect
biodiversity, a - Per capita biocapacity (be) compared to per capita ecological footprint (ef).
b - Biocapacity for the population (BC) compared to total ecological footprint (EF) 36
Figure 3.2 - Available biocapacity (be) for the five footprint land categories over a 26-year period in
the San Luis Basin, Colorado. Twelve percent of the biocapacity area was subtracted from the total
be to protect biodiversity. Note that built-up land has no biocapacity and is not included in
the figure 37
Figure 3.3 - Ecological footprint (ef) and the demand placed on each of the six land categories over
a 26-year period in the San Luis Basin, Colorado 37
Figure 3.4 - Energy consumption (hectares per person) by major energy source over a 26-year
period in the San Luis Basin, Colorado 38
Figure 3.5 - Ecological balance (ha/ca) of the different land categories in the SLB. Remainder
equals biocapacity per capita minus the ecological footprint for the SLB region. Positive values
represent a surplus or remainder of biocapacity and negative values represent a deficit of biocapacity.
Twelve percent of the biocapacity area was subtracted from the total be to protect biodiversity 38
Figure 3-D. 1 - Biocapacity (be) compared to ecological footprint (ef) per capita for 26 years in the US.
Twelve percent of the biocapacity area was subtracted from the total be to protect biodiversity 52
Figure 3-D.2 - Total ecological footprint (EF) and biocapacity (BC) for the US population. Twelve
percent of the biocapacity area was subtracted from the total BCto protect biodiversity 52
Figure 4.1- Net Regional Product, Green Net Regional Product, and the change in Green Net
Regional Product 65
Figure 4.2 - Components of the Depreciation of Natural Capital 65
Figure 4.3 - Sensitivity Analysis Using Upper and Lower Bounds on the Prices for Natural
Capital Depreciation 66
Figure 4-A. 1 - General approach to estimating wind erosion in the San Luis Basin 75
Figure 5.1 - An Energy Systems Diagram of the San Luis Basin showing the major features of this
environmental system 102
Figure 5.2 - An aggregated model used to calculate summary variables and indices for the San Luis
Basin system from 1995 to 2005. Year 2000 is presented here as an example 103
-------
Figure 5.3 - Total emergy used and renewable emergy used in the San Luis Basin from 1995
to 2005 103
Figure 5.4-The total emergy used per person in the San Luis Basin from 1995 to 2005 104
Figure 5.5 - Emergy inflows from renewable sources, imported fuels, electricity, goods and services,
emergy outflows in exported materials, the services in material products, services, and nonrenewable
resources exported directly in the San Luis Basin, 1995-2005 104
Figure 5.6 - A comparison of the total emergy used by the system with the renewable, local
nonrenewable and imported emergy that contribute to it in the San Luis Basin, 1995-2005 105
Figure 5.7 - Fraction of renewable emergy to total emergy use in the San Luis Basin from 1995
to 2005 105
Figure 5.8 - The renewable emergy used per person in the SLB region from 1980 to 2005 106
Figure 5.9-The renewable emergy inputs to the San Luis Basin, 1980-2005 106
Figure 5.10-The renewable emergy base for the San Luis Basin, 1980-2005 107
Figure 5.11- Agricultural production in the San Luis Basin supported primarily by renewable
resources, 1980-2005 107
Figure 5.12- Major categories of nonrenewable resources used and/or produced in the seven
counties of the SanLuis Basin, 1980-2005 108
Figure 5.13- The emergy exported from and imported to the San Luis Basin from 1995
to 2005 108
Figure 5.14-Major categories of emergy imported into the San Luis Basin, 1980-2005 109
Figure 5.15 - The categories of material goods other than fuels imported to the SLB from 1995
to 2005 109
Figure 5.16- Emergy exported from the San Luis Basin, 1980-2005: totals and aggregate categories
Total exports include total materials and total services exported. Total materials exported includes
agricultural and forest products, livestock minerals, and all other materials. Services in materials
exported includes services in services in minerals, all other materials, agricultural and forest
products. Export of pure services was not evaluated and emigration is shown, but was not included
in the total 110
Figure 5.17- Selected categories of agricultural, forest product, and mineral exports from the San
Luis Basin, 1980-2005. The All Other Materials and the services in that category are shown to
facilitate further discussion 110
Figure 5.18- The six largest classes of exports contained in All Other Materials exported from the
San Luis Basin, 1995-2005 Ill
Figure 5.19- The emergy of fuels and electricity used per person in the San Luis Basin from 1980
to 2005 Ill
-------
Figure 5.20 - Renewable emergy sources of the San Luis Basin (1980-2005) that are being used in a
nonrenewable manner 112
Figure 5.21 - The fraction of total emergy use that comes from local renewable emergy and the
fraction of use from home sources in the San Luis Basin, 1995-2005 112
Figure 5-B. 1 - An Energy Systems Diagram used to define the emergy indices (Table 5-B. 1) to
assess regional sustainability 116
Figure 5-B.2 - The Emergy Sustainability Index (ESI = EYR/ELR) and its components the emergy
yield ratio (EYR) and the Environmental Loading Ratio (ELR) in the San Luis Basin from 1995
to 2005 117
Figure 5-E. 1 - Renewable production compared to the emergy of fuels and electricity, and the
renewable emergy base of the SLB 148
Figure 5-E.2 - Minor components of the emergy input to the San Luis Basin 1980-2005 148
Figure 5-G.l - The emergy of processed non-metallic minerals (i.e., perlite and scoria) imported into
the San Luis Basin and the emergy exported in that category 160
Figure 5-G.2 - The change in annual empower from net immigration or emigration from 1985
to 2005 160
Figure 5-G.3 - The emergy of fuels and electricity used per year in the San Luis Basin from 1980
to 2005 161
Figure 5-G.4 - The difference between the total emergy outflows (B+P2E+N2) and total emergy
inflows (R+F+G+P2I) 161
Figure 5-G.5 - Digital elevation model of the Upper Rio Grande River Basin showing the
boundaries of the closed and open basins 162
Figure 6.1- Graphical representation of the sample settings for the Fisher information (FI)
computation indicating the size of the integration window (hwin) and window increment (winspace)
in time steps used to move through the data. In this example, hwin =10 and winspace = 5. The red
star indicates the center point of the window denoting the placement of FI 177
Figure 6.2 - Flow diagram of the binning algorithm in the Fisher information computation.
According to Equation 6.4, if each variable x.(t) - x.(t)\ is less than the size of states (Axi), then the
two points at time /' andy are indistinguishable and are grouped (binned) in the same state 177
Figure 6.3 - Population (xp green triangles) and Personal Income (x2, blue diamonds) values for
the San Luis Basin from 1980 to 2005. These variables were used as data for the example FI
computation exercise provided in Section 6.2.4 178
-------
Figure 6.4 - Bulls-eye plot of binning points in state 1 for the simple example. Point 1 (p^: (x^),
X2(tj)) = (3 10352, 38469)) is the center of this figure and the absolute difference from ptj are plotted
on the corresponding axis, e.g., absolute difference 2-1 = (4562, 511). The values plotted are from
Table 6.2. As described in Section 6.2.2, the level of uncertainty (i.e., size of states (Ax.)) was
computed for each variable such that in this plot AXj and Ax2 correspond to the red and blue line,
respectively. The points are binned in a state when the difference is less than the level of uncertainty
(i.e., size of states) [[[ 178
Figure 6.5 - Probability density (p(s)) for each window in the simple example (Section 6.2.4). These
values were computed by first counting the number of points in each state within each window and
then computing a probability for each state in each window. In window 3, four points were binned
in both state 1 and state 2; therefore, the probability was 50% for each state .................................. 179
Figure 6.6 - Calculating gradients of the amplitude (q(s)) in window 1 of the simple example
(Section 6.2.4). Gradients were used as a convention for mimicking a discrete density function and
were computed as qt-qi+1, where /' is the state. These values used in Equation 6.3 to compute FI... 179
Figure 6.7 - Fisher information for each window from the simple example: (A) Fisher information
(FI) and (B) Smoothed FI using three-point mean (), i.e., 1999 = (FI2002+FI1999+FI1996)/3 and
+FI)/3 [[[ 180
Figure 6.8 - Fisher information calculated for the San Luis Basin over a 26-year period (1980-2006).
Each point represents the FI computed in each window and is reported in the center of that period
(i.e., FI; 1984 = 1980-1987, 1985 = 1981-1988, 1986 = 1982-1989, etc.) ....................................... 180
Figure 6.9 - Mean Fisher information () for the San Luis Basin. was computed as a three-
point average of the FI (Fig. 6.8) in order to smooth out short-term fluctuations and highlight trends,
rather than fluctuations. The is reported in the center of the three points (i.e., ; 2001 =
2002-2000, 1998 = 1999-1997, etc.) [[[ 181
Figure 6.10- Mean Fisher information of the San Luis Basin for the overall system and by category.
Using this plot, we compared changes in system mean FI () to the the variables grouped
in categories. The is reported in the center of the three points (e.g., 2001 = (FI2000 + FI2001 +
FI2002)/3) [[[ 181
Figure 6.11- Percent change in the San Luis Basin demographic variables. The percent change in
population corresponded to the changes in dynamic order of the demographic category. The change
in population decreased initially, peaked in 1995, and decreased thereafter which corresponds to the
changes in dynamic order for the demographic category [[[ 182
Figure 6-A. 1 - Fisher information is proportional to the probability of system states (Pawlowski and
-------
Figure 6-B. 1 - The effect of changing hwin on the FI result. The value of hwin changes from panel
to panel (A-D), while winspace varies within each panel. Therefore, each panel reflects the Fisher
information result over time with hwin constant and a different winspace for each computation: (A)
hwin =8, winspace =1 to 4, (B) hwin =9, winspace =1 to 4, (C) hwin =10, winspace =1 to 4, (D) hwin
= 10, winspace =1 to 4. The average value (avgFI) and standard deviation (SD) reported of FI in
each panel were computed from all the FI results in each panel as a summary statistic. For example,
avgFI and SD in panel A are the mean and standard deviation of the FI values given hwin = 8 and
winspace = 1 to 4. Note that avgFI decreases as hwin increases and small winspace values result in
more fluctuations (e.g., winspace =1 is represented by the blue line) 189
Figure 6-B .2 - Interaction plots for the average value of FI (avgFI). Plotting the value of avgFI
as a function of hwin and winspace affords the ability to evaluate the impact of both parameters
on avgFI. (A) Each hwin value is represented by a different color plot. With winspace varying
along the x-axis, there is little variability in avgFI as winspace increases. Conversely, note that as
hwin increases (e.g., Al: hwin = 8, A4: hwin = 11), the value of avgFI decreases. (B) The strong
relationship between hwin and avgFI is also shown as hwin increases along the x-axis 190
Figure 6-B.3 - Interaction plots for the number of FI results (Npts). Plotting the value of Npts as a
function of hwin and winspace affords the ability to evaluate the impact of both parameters on Npts.
(A) Each computation with a particular hwin value is represented by a different color plot. With
winspace varying along the x-axis, there was a great deal of variability in Npts. (B) However, as
hwin increases along the x-axis, there are minor changes in Npts. Further, there is a greater decrease
in Npts as winspace goes from 1 to 2, than from 2 to 3 or 3 to 4 190
Figure 6-B.4 - Results of the numerical optimization of hwin and winspace using data from the
simple example. Based on the optimization criteria of maximizing Npts and minimizing the
deviation in avgFI, the optimal solution was found (using Design Expert 8) with integration window
parameter settings of hwin = 8 and winspace = 1 191
Figure 6-B.5 - Final Fisher information (FI) result from the simple example. (A) FI result from
the optimal parameter settings (hwin = 8 and winspace =1). (B) Smoothed result: , three-point
mean Fisher information 191
Figure 6-C. 1 - GUI interface developed to compute FI from system data. The GUI is used for (1)
importing pre-formatted data files, (2) choosing parameter settings (3) computing and plotting FI, (4)
saving the FI results and (5) further study of regime shifts. The results plotted in this figure are from
the simple example in Appendix 6-D 194
Figure 6-C.2 - The Advanced GUI was developed to study the pervasiveness and intensity of
ecosystem regime shifts by (1) evaluating and (2) plotting the FI at various tightening levels. (3)
The perfect order tightening level and pervasiveness are computed and shown in the text boxes.
The results plotted are from the simple example as computed in Appendix 6-D. Using this simple
example, note that the pervasiveness is 100% and there is one regime shift between two stable
regimes 195
Figure 7.1 - Graphs of Ecological Balance, Green Net Regional Product (GNRP), Fraction of
Locally Renewable Emergy, and mean Fisher Information for 1980-2005 231
-------
List
aa. Area Appropriated Per capita
ARS Agricultural Research Station
be Bioeapacity Per Capita
BC Bioeapacity for the Population
BEA Bureau of Economic Analysis
BLM Bureau of Land Management
C Countj"
CDSS Colorado's Decision Support System
CDWR Colorado Division of
Water Resource
CFC Consumption of Fixed Capital
C. Consumption Item
CO2 Carbon Dioxide
CON Consumption
CPP Community Partnership Program
CSERGE Center for Social and Economic
Research on Global Environment
CWCB Colorado Water Conservation Board
DEM Demographic
Dwt Dry weight
EBT Estimated by Team
EF Ecological Footprint for population
ef Ecological Footprint per person
EFA Ecological Footprint Analysis
EGY Energy
EIA Energy Information Administration
ELR Environmental Loading Ratio
EmA Emergy Analysis
ENV Environmental
ESI Emergy Sustainability Index
EST Energy Systems Theory
ET Evapotranspiration
ExA Exergy Analysis
EYR Emergy Yield Ratio
FS Forest Service
FI Fisher information
mean Fisher information
GDI Gross Domestic Income
GDP Gross Domestic Product
gha Global Hectares
GI Global Insight
GNDP Green Net Domestic Product
GNRP Green Net Regional Product
GRP Gross Regional Product
GS Genuine Savings
GSA General Sendee Administration
GUI Graphical User Interface
/ Fisher Information (historical usage)
ha Hectares
LEI Local Effect of Investment
MBF Million Board Feet
Misc Miscellaneous
MLRA Major Land Resource Area
MSY Maximum Sustainable Yield
N AIC S North American Industry
Classification System
NASS National Agricultural Statistics
Sendee
NDP Net Domestic Product
NGO Non-governmental organization
NRCS Natural Resource Conservation
Service
NRI National Resource Information
NRP Net Regional Product
ORD Office of Research and Development
PI Personal Income
PCA Principal Components Analysis
PROD Production
R Region
S State
scj Solar Emjoulc
SIC Standard Industrial Classification
SLB San Luis Basin
SLV San Luis Valley
U Emergy Use
USDA United States Department of
Agriculture
USEPA United States Environmental
Protection Agency
UN United Nations
WEQ Wind Erosion Equation
WERU Wind Erosion Research Unit
Wwt Wet weight
-------
The U.S. Environmental Protection Agency (USEPA) is charged by Congress with protecting
the Nation's land, air, and water resources. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between human
activities and the ability of natural systems to support and nurture life. To meet this mandate,
USEPA's research program is providing data and technical support for solving environmental
problems today and building a science knowledge base necessary to manage our ecological resources
wisely, understand how pollutants affect our health, and prevent or reduce environmental risks in the
future.
The National Risk Management Research Laboratory (NRMRL) is the Agency's center for
investigation of technological and management approaches for preventing and reducing risks
from pollution that threaten human health and the environment. The focus of the Laboratory's
research program is on methods and their cost-effectiveness for prevention and control of pollution
to air, land, water, and subsurface resources; protection of water quality in public water systems;
remediation of contaminated sites, sediments and ground water; prevention and control of indoor
air pollution; and restoration of ecosystems. NRMRL collaborates with both public and private
sector partners to foster technologies that reduce the cost of compliance and to anticipate emerging
problems. NRMRL's research provides solutions to environmental problems by: developing
and promoting technologies that protect and improve the environment; advancing scientific and
engineering information to support regulatory and policy decisions; and providing the technical
support and information transfer to ensure implementation of environmental regulations and
strategies at the national, state, and community levels.
This publication has been produced as part of the Laboratory's strategic long-term research plan. It
is published and made available by USEPA's Office of Research and Development to assist the user
community and to link researchers with their clients.
Sally Gutierrez, Director
National Risk Management Research Laboratory
-------
and
The National Risk Management Research Laboratory (NRMRL), within U.S. EPA's Office of
Research and Development, was responsible for preparing this document. Authors are listed
alphabetically and do not necessarily represent level of contribution. Matthew T. Heberling and
Matthew E. Hopton were the document editors.
Chapter Authors
Chapter 1.
Heriberto Cabezas, Tarsha Eason, Ahjond S. Garmestani, Matthew T.
Heberling, Matthew E. Hopton, and Joshua Templeton
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati, OH
Daniel E.Campbell
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Narragansctt, RI
Denis White
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Corvallis, OR
Marie Zanowick
Region 8
U.S. Environmental Protection Agency, Denver, CO
Chapter 2.
Heriberto Cabezas, Tarsha Eason, Ahjond S. Garmestani, Matthew T.
Heberling, Matthew E. Hopton, and Joshua Templeton
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati. OH
Daniel E. Campbell
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Narragansett, RI
Denis White
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Corvallis, OR
Chapter 3.
Matthew E. Hopton
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati, OH
Denis White
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Corvallis, OR
Chapter 4.
Matthew T. Heberling and Joshua Templeton
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati, OH
-------
App. 4-A.
Matthew T. Heberling
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati. OH
Richard T. Sparks
Natural Resources Conservation Service
U.S. Department of Agriculture. Center, CO
Chapter 5.
Daniel E. Campbell
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Narragansett, RI
Ahjond S. Garmestani and Matthew E. Hopton
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati, OH
Chapter 6.
Heriberto Cabezas and Tarsha Eason
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati, OH
Chapter 7.
Heriberto Cabezas, Tarsha Eason, Ahjond S. Garmestani, Matthew T.
Heberling, Matthew E. Hopton, and Joshua Templeton
National Risk Management Research Laboratory
U.S. Environmental Protection Agency, Cincinnati, OH
Daniel E. Campbell
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Narragansett, Rl
Denis White
National Health and Environmental Effects Research Laboratory
U.S. Environmental Protection Agency, Corvallis, OR
Marie Zanowick
Region 8
U.S. Environmental Protection Agency, Denver, CO
-------
This report was externally peer-reviewed by the following professionals with experience in
sustainability research. We acknowledge the effort and insightful comments provided by the
reviewers and accept all remaining errors as our own.
Bhavik R. BaksM
Department of Chemical & Biomolecular Engineeering
The Ohio State University, Columbus, OH
Matt Luck
ISciences, L.L.C,
Burlington, VT
David I. Stern
Arndt-Corden Division of Economics
Australian National University, Canberra ACT
David R. Tilley
Department of Environmental Science & Technology
University of Maryland. College Park, MD
-------
The authors would like to thank Mike Britten, Fred Bunch, Art Hutchinson, Richard Sparks, John
Stump, Andrew Valdez, and Steve Vandiver for directing us to or collecting data for the metric
calculations. We thank several people for providing data, including Kathy Worthington (Xcel
Energy), JoAn Waugh (Rural Electric Co-op), Bruce Short (USDA-FS), and Pete Magee (San Luis
Valley GIS/GPS Authority). The authors wish to acknowledge Fred Bunch for Ms enthusiasm, his
assistance, and the many helpful discussions throughout the three years of this project. The authors
thank Troy Fannon for his support in preparing part of the front matter. We recognize Brent Frakes
for his data management skills.
Matthew T. Heberling acknowledges Ray Bennett and Colorado's Decision Support Systems for
estimating the water balance model for the Rio Grande River Basin and Richard Sparks for his help
estimating wind erosion for the SLB. We thank Rafael Weston for his helpful comments on the
GNRP calculation and information about the region.
Heribcrto Cabczas and Tarsha Eason acknowledge the seminal contributions to Fisher information
theory and computations from Arunprakash Karunanithi, Christopher Pawlowski, Brian Path, andB.
Roy Frieden.
Many individuals contributed to. and assisted with this project; we regret the possibility that we do
not acknowledge even-one.
-------
Sustainability relates to finding and maintaining conditions that can support social and economic
development without adversely altering the environment to too great an extent. Moreover, in order
to satisfy the definition of Sustainability, the environmental, social, and economic characteristics
of the system must effectively meet the needs of current and future generations, indefinitely. Our
definition of Sustainability depends on the extent to which the environment can be altered and
still maintain a high quality of life for people, without jeopardizing the quality of life for future
generations. Consequently, to assess Sustainability, information is needed to understand the
requirements for human well-being and the linkages and demands of human activity (e.g.. society.
economy, government, industry, etc.) on environmental systems. One way to collect needed
information is through metrics that quantify the environmental, social, and economic characteristics
of a system. We used a collaborative, interdisciplinary approach to investigate this complex
problem. Specifically, we set out to: 1) determine the applicability of using existing datasets to
estimate metrics of Sustainability at a regional scale, 2) calculate the metrics through time (1980-
2005), and 3) compare and contrast the results to determine if a regional system is moving toward or
away from Sustainability. This information can help decision makers determine if their region is on a
sustainable path (i.e., moving toward Sustainability).
As a starling point, we identified and tested a set of four metrics that capture some of the most basic
properties of an environmental system. The metrics represent: 1) ecological impacts of human
activity to produce the resources consumed and to assimilate the wastes generated using Ecological
Footprint Analysis; 2) economic well-being or welfare with Green Net Regional Product; 3) flow of
available energy through the system using Emergy Analysis; and 4) overall system order with Fisher
information (i.e.. a well-functioning system is orderly and departure from such a state can lead to
decreased function). Each of these metrics is most sensitive at capturing some particular aspect of
the system, although there is some overlap or redundancy (e.g., consider a physician using multiple
tests to examine a patient's health). We tested the methodology on the San Lids Basin (SLB) in
south-central Colorado. The SLB is a rural, agricultural region with a limited population. This
seven county region contains the Upper Rio Grande River Basin, the San Luis Valley, and the Great
Sand Dunes National Park and Preserve.
Even though data for some of the variables were not available at the county level, we were able
to calculate the metrics through time. From our analyses, Ecological Footprint Analysis indicated
the SLB was moving away from Sustainability. Green Net Regional Product appears to have an
upward trend over the 26-year period providing no indication that the SLB was moving away from
Sustainability (this one-sided test can only tell if a system is moving away from Sustainability). Due
to data unavailability, a complete Emergy Analysis was estimated only for an 11-year period (1995-
2005). One emergy index (fraction of renewable emergy used) that captures Sustainability within
the system suggested a gradual movement away from Sustainability, but a second index (Emergy
Sustainability Index) revealed the region improved its relationship with the larger system, (i.e., other
states and US) from 2003 until the end of the period. Finally, the results of the Fisher information
assessment revealed that, although the system was relatively stable during the 26 years, there was an
indication of slight movement away from Sustainability near the end of the study period.
We consider the entire system moving away from Sustainability if any one metric reveals movement
away from Sustainability. Therefore, the weight of the evidence from our results indicate that the
broad trend of the SLB was moving away from Sustainability over the period examined. Because
this was a pilot study, we offer a number of recommendations for future research based on what we
learned while developing and using this approach.
-------
ES.1 Introduction
There are several established, scientifically supported metrics of sustainability. Many of these
metrics require extensive data collection and detailed computations. Moreover, it is difficult for
one metric to capture all aspects of a system that are relevant to sustainability. A pilot project
was initiated to test an approach to measure, monitor, and maintain prosperity and environmental
quality in the context of sustainability for a regional system. Sustainability relates to finding and
maintaining conditions that can support social and economic development without adversely altering
the environment to too great an extent. The goal of this study was to produce a straightforward,
relatively inexpensive methodology that was simple to use and interpret. This approach required
historical data to be readily accessible, that metrics be applicable to the relevant scale, and that
results meet the needs of decision makers.
Because sustainability is a multidimensional concept, the research group consisted of an
interdisciplinary team in which the members represented expertise associated with several
fundamental components of an environmental system. This project utilized available environmental,
economic, and social data to calculate four sustainability metrics: 1) environmental footprint as
characterized by Ecological Footprint Analysis (EFA), 2) economic well-being as ascertained from
Green Net Regional Product (GNRP), 3) energy and emergy flow through the system as computed
from an Emergy Analysis (EmA), and 4) dynamic order estimated from the computation of Fisher
information (FI).
The San Luis Basin (SLB), a region in south-central Colorado, was selected as the pilot study area
because of its natural hydro logical boundaries, limited population, large amount of publicly owned
land, and interest expressed in the study by USEPA Region 8 and other government officials. We
define the SLB as the following seven counties: Alamosa, Conejos, Costilla, Hinsdale, Mineral, Rio
Grande, and Saguache. This region contains the Upper Rio Grande River Basin, the San Luis Valley,
and the Great Sand Dunes National Park and Preserve.
ES.2
We reviewed the literature to summarize different metrics that quantity' the major components of
an environmental system. This resulted in a review of potential metrics for this pilot study. We
reviewed studies that examined sustainability using multiple metrics and studies that measured
sustainability over time. Both are important attributes of the San Luis Basin Sustainability Metrics
Project
No single metric should be considered a perfect measure of sustainability because of the multiple
definitions and interpretations of sustainability and the difficulties in measuring it. The four
sustainability metrics we selected provide a holistic view of what we considered to be Hie
fundamental aspects of an environmental system.
At a basic level, we recognized that an environmental system has some inherent order, whether the
factors responsible for maintaining that order are clearly identified or understood. Furthermore,
it takes available energy (i.e., energy with the potential to do work) to maintain order in an
environmental system. Finally, sustainability is inherently an anthropocentric issue and how well
off humans are may dictate their influences on the system. For example, an individual or population
that is meeting its needs for existence has the luxury of being concerned about the effect they have
on their environment, whereas a population that is barely surviving may not be intensely concerned
with sustainability.
We selected metrics that capture each of these fundamental aspects; they also represent both strong
and weak measures of sustainability'. Weak sustainability assumes ecological functions and/or
materials can be substituted by technological or man-made surrogates whereas strong sustainability
assumes ecological functions are not replaceable or natural capital cannot be substituted with other
forms of capital (e.g., Dietz and Neumayer 2007, Mayer 2008, Mayer et al. 2004, Neumayer 2010,
Pearce 2002).
-------
EFA was selected to capture the impact a population has on environmental resources resulting from
consumption and waste production. We chose EmA because it measures the flow of available energy
through a system, by converting all types of energy to a common unit (i.e., solar emjoule). Both
of these metrics arc considered measures of strong sustainability. GNRP was chosen to capture
the overall economic well-being of the population and FI was selected to measure dynamic order
of the overall system (both measures of weak sustainability). Out of the four metrics, using FI to
characterize sustainability is a relatively new metric (Mayer 2008). Application of the other metrics
is relatively common and they are established as metrics of sustainability in the scientific literature
(e.g., Mayer 2008). The next four sections summarize each of the metrics.
ES.3 Footprint
This chapter describes the data sources and methodology used to estimate a simplified Ecological
Footprint Analysis (EFA) at a regional scale. EFA captures the human impact on the environmental
system by identifying the amount of biologically productive land necessary to support a person's
level of consumption and waste generation. EFA is a commonly used metric of sustainability
because it is easy to conceptualize and the calculation is relatively straightforward. A typical EFA
requires estimation of the supply of biologically productive land in the geographic region of interest
(called the biocapacity) and the demand placed on the biologically productive land to support the
population (called the ecological footprint). The supply is estimated by identifying the number
of hectares in each of six land categories (arable, pasture, energy, forest, sea, and built-up land).
Consumption levels for food, energy, and other consumables, arc assigned to a land category and
the amount of land necessary to meet the level of consumption is calculated as the demand. An
ecological balance is calculated by subtracting the ecological footprint from the biocapacity. If the
ecological footprint is less than the available biocapacity, the system has an ecological remainder.
If the ecological footprint is greater than the available biocapacity. the system is said to possess an
ecological deficit. If the ecological balance is decreasing, resulting from an increasing ecological
footprint and/or a decreasing biocapacity, the system is said to be moving away from sustainability.
Utilizing free, readily available data, we calculated an EFA for the SLB. Gathering existing data at
a regional scale can be difficult because data are often collected at national or state levels. The lack
of data is further confounded by the fact that data are often collected at intervals greater than one
year. Variables that were missing data for certain years were estimated using linear interpolation.
Other data were from state or national level data and were scaled to the region. For example, energy
consumption data were reported for the State of Colorado. A per capita consumption was calculated
by dividing the annual consumption by the population of the state for the year of interest. This per
capita value was assumed to represent the level of consumption for the region under investigation.
Food consumption was reported as average per capita for the US. This value was assumed to
represent the level of consumption for individuals in the region.
Thirty-five variables from 1980-2005 (26 years) were collected and used to calculate a time-
dependent EFA and the resulting trend was visually examined. The available biocapacity in the
region did not decrease during the period, but per capita biocapacity is decreasing due to population
growth. Per capita biocapacity was at a period high of nearly 36 hectares per person (ha/ca) in
1980 and steadily decreased to a low of just over 27 ha/ca in 2005. Ecological footprint remained
relatively constant over the 26-year period, with a low of 5.10 ha/ca in 1997 to a high of 5.5 ha/ca
in 1985. A steady ecological footprint combined with a decreasing per capita biocapacity, implies
the ecological balance (i.e., remainder) is decreasing and thus the region is moving away from
sustainability. Although per capita consumption did not increase substantially during the 26 years,
more people are drawing on a set quantity of resources. Sustainability, as defined by EFA, requires
that people reduce their footprint to free up resources for a growing population or to increase
biocapacity. Lastly, it is important to remember the region is part of a larger system and provides
much of its natural resources to support the larger system.
Our methodology is a simplified approach to EFA and does not follow standards that are currently
being established. Adhering to the suggested standards would require obtaining data sets that
consist entirely of national data. The national-level data arc replaced with data specific to the
geographic area under examination when they are available. Although national data may represent
-------
the sub-national region under study, the idea requires further investigation, especially in large.
geographically and culturally varied nations such as the US.
ES.4 Green Product
This chapter describes the data sources and methodology used to estimate Green Net Regional
Product (GNRP). GNRP is equal to aggregate consumption minus the depreciation of man-made
and natural capital. It captures the economic well-being, or welfare, measured in US dollars.
Welfare, as defined by economists, incorporates benefits obtained outside of market transactions
such as environmental amenities, but allows for the possibility of trade-offs between market and
non-market goods. GNRP is a measure of weak sustainability that incorporates the assumption that
human and man-made capital can be substituted for natural capital. We measure Hie movement
toward or away from sustainability by examining the change in GNRP over time.
Any attempt at green accounting requires both economic and natural capital data. The economic
data used in our study came from the Bureau of Economic Analysis (BEA). We collected natural
capital data from the Energy Information Administration, Agricultural Statistics for Colorado, The
Colorado Water Conservation Board and the Colorado Division of Water Resources as well as
professional contacts in the SLB.
Our approach to estimating GNRP required transforming BEA economic data at the national,
state, and county levels to the level of the SLB. Nationally, Gross Domestic Product (GDP) and
Consumption of Fixed Capital for the US were required. At the state level, the approach required
Personal Income and GDP for Colorado and New Mexico. For the seven counties in the SLB, we
used Personal Income from the BEA.
Given the contribution of agribusiness to the SLB, we included Hie depletion of both groundwater
and soil as components in the depreciation of natural capital. In addition, we captured the effect
of the consumption of energy on future generations through carbon dioxide (CO2) emissions. We
utilized the Upper Rio Grande database from Colorado's Decision Support System to estimate the
depletion of groundwater in the SLB. We averaged per capita CO, emissions for Colorado and
New Mexico and applied this to the population estimates of the SLB. Wind is Hie main cause of soil
erosion in the SLB. Finding reliable, county-level soil erosion data was challenging and required
us to make many assumptions, consult with experts, and spend many hours doing the calculations.
Future studies are likely to find this to be one of the more difficult aspects of applying our approach.
After collecting data for changes in natural capital, we needed to estimate the value of the change
in the natural capital. The shadow price of groundwater for irrigation, the economic damages from
wind erosion, and the social cost of carbon emissions were obtained from the literature and applied
using benefit transfer.
Aggregating Net Regional Product (NRP; i.e., Gross Regional Product adjusted for Hie depreciation
of man-made capital), Groundwater Storage Depreciation, Soil Erosion Damage, and CO2 Damage
Costs, we estimated GNRP for the SLB from 1980-2005. GNRP had a slight upward trend (based on
a visual examination). Although there are peaks and troughs, the upward trend suggested that there
is no definitive evidence of moving away from sustainability.
We have some concerns about our approach for estimating values for natural capital depreciation.
A sensitivity analysis of the shadow prices and marginal damages indicated trends in GNRP did not
differ for either low or high estimates. We developed and used an approach for estimating GNRP
using publicly available data. Although the natural capital depreciation estimates will differ by
region, the general approach can be applied to other regions.
ES.5
Emergy is an unfamiliar term to many people, yet it is a quantity of great importance for the accurate
evaluation of the condition of systems on all scales from the chemical reactions of molecules to the
birth and death of stars and galaxies. Several aspects of system sustainability can be measured with
emergy indices. The potential for sustaining the condition of the present system is indicated by the
percent of renewable emergy used to support current system structure. The overall sustainability of
the relationship between a system and its next larger system is given by the Emergy Sustainability
-------
Index (ESI; Brown and Ulgiati 1998). This index is the ratio of the emergy yield to the larger
system, the Emergy Yield Ratio (EYR), divided by the potential environmental damage done to the
local system. Environmental Loading Ratio (ELR). Greater emergy yield for less environmental
damage is preferable.
In this summary, our emergy evaluation of the San Luis Basin (SLB) region of Colorado, we focus
on explaining emergy and the results of our analyses. Emergy Analysis (EmA) is another way
to estimate the value that the environment contributes to society. EmA uses nature's accounting
system (that is the flow of available energy2) to determine where a system stands and to judge its
current condition. Nature does not and cannot accept money as payment for the real work that it
does to support society in the SLB and elsewhere, yet emergy analysis shows that almost 50% of
the work that is done in the region is done by the natural systems of the environment. Furthermore,
the greatest amount of nature's work done annually is contributed by the snowpack that stands
on the high mountains surrounding the valley. The existence of stored water at high elevations
allows all of the geopotential energy of this water to be released in a short period of time and in the
process, it recharges the groundwater and maintains unique geological and ecological features of
the valley like the Great Sand Dunes and wetlands. One consequence of this fact is that the natural
and agricultural systems of the region are vulnerable to climate changes that affect the snowpack.
One of the largest natural assets of the region, as shown by EmA, is the stored available energy in
groundwater. Even-one that lives in the region knows this and the history of increasing groundwater
use that was a major factor in the rapid growth of agriculture from the 1880's to 1980 (Emery 1996).
The advantage that EmA gives managers is it allows them to evaluate all aspects of the system in
common biophysical terms (i.e., the amount of solar energy required for environmental, economic.
and social products and services of the system). In this manner, all aspects of the system for which
they are responsible in equal terms, that is disparate quantities (like environmental damage and
economic production) are put on the same biophysical basis so that changes in each are directly
comparable. Economic value serves this purpose in economics, but a different perspective of value
is used and the scaling of relative values is often different in the two approaches. Information on
the emergy value of aspects of a system compared with economic values for the same aspects has
the potential to lead to better and fairer decisions on contentious issues faced by land managers. An
explanation of emergy in more technical language is given in the following paragraph.
EmA assesses the condition of a system based on the flow of available energy required to develop
and operate that system. The quality of energy flows is normalized by expressing all kinds of
available energy in a common unit, the solar equivalent joule (sej), by converting the quantities of
energy in joules or materials in grams used to make a product or sen-ice. The product is converted to
emergy by multiplying the original units (J or g) by an emergy per unit factor, called a transformity
(sej/J) or specific emergy (sej/g), to obtain emergy. For example, a bushel of wheat weighs about 60
Ibs and a gram of wheat contains about 14.200 J of available energy or 3.87 X108 J/bu. If we sum
all the inputs required for the production of the annual wheat crop in Minnesota (Campbell and Ohrt
2009), including the evapotranspiration of water, topsoil erosion, fuel, potash, phosphate, nitrogen,
pesticide, herbicide, electricity, and groundwater we find that 2.89X1013 sej were required to produce
the bushel of wheat. Thus, the transformity of wheat, without human sendees (labor), included was
7.47 10'1 sej/J. This factor can be applied to wheat production in a similar area (i.e., the SLB) to
estimate the emergy of the crop given that the harvest in bushels is known.
The emergy of any environmental, economic, or social product and service can be determined in
this manner as long as the production process for these items is known. Once converted to emergy
the results are directly comparable so the relative advantages and disadvantages of any change
due to alternative policies can be readily determined by direct inspection of the values. Relative
advantage is judged by movement toward greater emergy flow through the system network, that is in
the direction of greater competitiveness, and therefore toward a higher probability that the system's
condition will be sustained given no change in the set of inputs available to the system. According
to Emergy Systems Theory, maximizing empower is hypothesized to be the variable deciding the
outcome in evolutionary competition among alternatives (Odum 1996). If this is true, then system
2 Available energy is energy with the potential to do work. For example water in the mountains can do work, such as drive a
turbine as it flows to a lower elevation.
-------
empower is a master variable that serves as a reference point to characterize system operations
(i.e., increasing system empower flow moves toward greater overall well-being, whereas a decline
indicates a less healthy system that will not compete as successfully with other systems). From this
perspective, every manager should know the projected effect of alternative policies on the empower
(emergy/time) of their systems as well as the effect on system dollar flows, before making a decision.
In this study, information was gathered via government agencies, literature surveys, interviews
with experts, the purchase of a dataset, and meetings with knowledgeable people from the region.
EmA requires data on all aspects of a system; and therefore, data on several important variables
were snared with EFA and GNRP data sets. The data needs to construct an EinA of a region can
be categorized as follows: 1) data on renewable energy inputs to the system and information on
renewable production carried out in the system; 2) data on nonrenewable energy inputs, production.
and consumption, including any renewable energy sources used in a nonrenewable manner; 3) data
on imports to the system including raw and finished materials and sen-ices; and 4) data on exports
from the system including raw and finished materials and sendees. We performed an evaluation
of the annual emergy flows of the SLB region synthesizing data for the seven counties and the
associated Upper Rio Grande River Basin. Out of the numerous indices calculated during a typical
EmA, we selected the fraction of renewable emergy used to total emergy used as the index on which
to focus because it would provide the best indication if the SLB was moving toward or away from
sustainability. The ESI is also used to examine the sustainability of the relationship between the
SLB and its next larger system.
The limiting factor for determining a time series of emergy flows for the SLB was the need for
complete and accurate data on imports and exports. This required the purchase of relevant data from
a private economic data sendee. The firm compiled freight movement by county and these data were
available for the seven-county region for the years 1995 to 2005. Therefore, the complete EmA of
the region was limited to these 11 years.
More often than not, the SLB exported more emergy than it imported. Thus, the SLB screes as
a hinterland supplying raw products and resources to the larger system of Colorado and the US.
Although there was some variability during the 11-year period, the fraction of renewable emergy
to total emergy used had a slight decline over the period. Thus, we can infer that during this time
the SLB was gradually moving away from sustainability. These trends culminated in 2002 when
a large groundwater withdrawal compensating for decreased precipitation in that year brought the
fraction of renewable emergy to a low point. In 2003, renewable emergy inflow recovered, and
exports began to recover, while imports remained about the same level. This led to an increase in
this emergy index of sustainability, which reversed its general downward trend, resulting in upward
movement from 2003 to 2005. However, the total emergy used in the SLB, which is an indicator of
well-being of the local system, recovered much more slowly with a relatively modest increase (6%)
in total emergy flow over this time. The total renewable emergy used per person in the region also
declined for the study period due to population growth. In summary, although both indices show
a decline and resurgence in later years, overall, they indicate an underlying trend for the region of
gradually moving away from sustainability for the period examined.
ES.6 Fisher Information Order
This chapter describes the theory, data sources, and methodology for using Fisher information to
assess the order and stability of the SLB regional system. Fisher information (FI) was developed
as a measure of order in data and is an important method in information theory. The method is
based on the amount of information available in a time series of data and is focused on determining
patterns of behavior. Because the presence of patterns implies a level of organization, FI can be seen
as a measure of dynamic order. In the context of system sustainability, order relates to the ability of
a system to maintain a desirable steady state (i.e. regime) even in the presence of natural fluctuations.
Assessing dynamic order enhances environmental management in its effort to monitor and manage
the impact of human activity on ecosystems. Through this approach, changes in complex system
behavior may be observed over time with the aim of maintaining desired regimes and avoiding
catastrophic shifts.
-------
Our adaptation of FI provides a means of monitoring variables of a system to characterize its
dynamic order and self organization, and. accordingly, its regimes and regime shifts. FI over
time denotes changes in dynamic order such that decreasing FI indicates movement away from
sustainability. no net change in FI denotes the system is maintaining its current state and increasing
FI indicates that system is becoming more orderly and possibly moving toward a new dynamic
regime that is different from the present regime.
In this chapter, we provide the basic theory on the method, describe the computation approach, and
list the variables used in the assessment. To enhance understanding of the method, we provided
a pedagogic exercise to "walk through" the steps for computing Fisher information. Further, we
compute FI for the SLB data and discuss the results, as well as, the strengths and weaknesses of the
approach.
One of the main goals of this project was to use readily available data to assess regional
sustainability. Based on the three pillars of sustainability, it is paramount that data used to compute
FI encompass the environmental, social, and economic components of the system. Because data
compiled to calculate EFA, GNRP. and EmA incorporate the pertinent aspects of activity in the
region, key primar}' variables were selected from these data sets. These variables which capture
the consumption of food and energy, agricultural production, environmental characteristics,
demographic properties, and changes in land use for the SLB system from 1980-2005 are listed in
the chapter. However, the source and details of the data are provided in the corresponding metric
chapters.
Our analysis revealed that there was no indication of a regime shift. The system exhibited small
changes in dynamic order during the study period, with a slight decreasing trend near the end of the
period (i.e., 1998-2001). Accordingly, we deduce that although conditions in the SLB are relatively
steady, there was an indication of slight movement away from sustainability.
A primary benefit of using this information theory approach is the ability to collapse data from
intricate, multivariate systems into a metric that can be computed over time. This is particularly
important considering the complex, multi-disciplinary nature of sustainability. Further, the method
is robust in that it assesses dynamic behavior of both model and real systems. One key output
of this project is the MATLAB code and graphical user interfaces (GUI) developed to simplify
the computation and evaluation procedure. Key considerations surrounding the computation of
FI include establishing the integration window parameters (hwin and winspace), the need for
determining a measure of uncertainty for the state variables (which is used to set the size of states
for each variable), and the availability and quality of data characterizing the state of the system.
Accordingly, we used a simple example to perform sensitivity' analysis on hwin and winspace to
determine the impact of those parameters on the result of FI. In addition, we provided guidance for
determining the size of states parameter.
ES.7 The Project:
Conclusions,
This chapter examines the sustainability metrics holistically, using the results of the previous four
chapters, to decipher whether the SLB is on a sustainable path (i.e., moving toward sustainability).
Although each of the four metrics captures different aspects of the system, they appear to have
similar results based on the concept of weak and strong sustainability. EFA and EmA, measures
of strong sustainability. indicate movement away from sustainability whereas GNRP provided
no definitive evidence of movement away from sustainability, and FI suggests relative stability,
both metrics are measures of weak sustainability. Overall, the weight of evidence led us to the
assessment that the region is moving slowly away from sustainability. The growth of population
from 1980 to 2005 and the increased pressure on environmental resources that is the inevitable
result of human use of the environment also supports our conclusion. However, we recognize the
estimates are not perfect and have identified the limitations of each metric in their respective chapter.
For example, the best we can do for a trend analysis is a visual examination without statistical tests
given the lack of uncertainty measurements for the variables.
-------
The majority of the limitations for the methodology relate to data needs and data manipulation.
Because much of the data needed for these metrics are typically not collected at the county level, we
needed to rescale or convert state or US data to the county or regional level. Other challenges are
related to quantitatively assessing trends for the metrics and using the methodology in other regions
besides the SLB.
We believe that our methodology adequately characterized and evaluated sustainability of the SLB,
in part, because our results are plausible and our observations and conclusions have been validated
to a certain extent in public meetings with stakeholders in the SLB and by anecdotal evidence. The
methodology demonstrated that this framework was not entirely straightforward or simple, nor was it
inexpensive, but it suggests the multidisciplinary approach does a sufficient job of assessing whether
a region is moving toward or away from sustainability. We do not contend that this methodology
is definitive; however, we do argue that each metric identified changes to major components of the
system. We propose that together, these metrics adequately represent the complexity- of a regional
system.
We offer the following recommendations for future research based on the results of our study:
1) Examine previous management decisions in the SLB to see if and how the metrics changed.
For example, the Great Sand Dunes National Park and Preserve was created in 2000
and increased the National Monument land area by almost four times (NFS 2007). This
recommendation would help to determine if any of the changes in the metrics could be linked
back to the management decision to increase the public land area.
2) Continue calculating the metrics in the SLB for subsequent years to examine how sensitive the
metrics are to management decisions.
3) Use stakeholder meetings to determine major sustainability issues in the region and better link
the metrics on those issues and concerns. Determine if there are other established metrics that
may better assess sustainability of a region by addressing known issues. This relationship with
stakeholders will help ensure the metrics are capable of capturing changes pertinent to issues
of public concern. The meetings can be used to gauge whether stakeholders and decision
makers can understand the results of (lie metrics. More research can focus on simplifying
the interpretation of results for various groups of people: managers, regulators, landowners,
educators, policy makers, etc.
4) Develop trend analyses and/or approaches for estimating confidence intervals for individual
metrics. Although a trend may appear obvious, recognizing and identifying the variation in the
data and resulting metric will better identify significant trends. One example might be using
econometric methods (e.g., Greene 1993, Hay et al. 2002) or sensitivity analyses.
5) Develop models of alternative future scenarios and estimate multiple metrics (Baker et al.
2004, Kok and van Delden 2009, Yeh and Chu 2004) that are sensitive to relevant changes in a
system or region. For example, how would the installation of solar farms affect the metrics?
6) Test these multiple metrics in other regions and determine if the metrics are indeed capturing
Hie trend in sustainability of regional systems.
7) Examine the correlation among alternative metrics to determine whether certain metrics could
be dropped from analysis (i.e., one metric moves in a similar or opposite pattern to another
over time) in order to reduce the resources needed (e.g., Bastianoni et al. 2008 used Principal
Components Analysis).
8) Consider the use of other scientific approaches, rather than limiting analysis to visual
examination of graphs, for deciphering the holistic results of multiple metrics, such as
multicriteria methods (e.g., Shmelev and Rodriguez-Labajos 2009); data envelopment analysis
(e.g., Despotis 2005, Kortelainen 2008); mathematical programming (e.g.. Zhou et al. 2007); or
other integrated sustainability metrics (e.g., Mayer 2008).
-------
ES.8
Baker, J.P., Huise, D.W., Gregory, S.V, White, D., Van Sickle, I, Berger, P.A., Dole, D., Schumaker,
N.H. 2004. Alternative Futures for the Willamette River Basin, Oregon. Ecological Applications
14,313-324.
Bastianoni, S., Pulselli, E, Focardi, S., Tiezzi, E., Gramatica, P. 2008. Correlations and complemen-
tarities in data and methods through principal components analysis (PCA) applied to the results of
the SPIn-Eco project. Journal of Environmental Management 86, 419-426.
Brown, M.T. Ulgiati, S. 1997. Emergy-based indicators and ratios to evaluate sustainability: moni-
toring, economies and technology toward environmentally sound innovation. Ecological Engi-
neering 9, 51-69.
Campbell, D.E., Olirt, A. 2009. Environmental Accounting Using Emergy: Evaluation of Minne-
sota. Office of Research and Development, National Health and Environmental Effects Research
Laboratory, Atlantic Ecology Division. EPA/600/R-09/002. 138 p.
Despotis, D.K. 2005. A reassessment of the human development index via data envelopment analy-
sis. Journal of the Operational Research Society- 56. 969-980.
Dietz, S., Neumayer, E. 2007. Weak and strong sustainability in the SEEA: Concepts and measure-
ment. Ecological Economics 61, 617-626.
Emery, P.A. 1996. Hydrology of the San Luis Valley. Colorado, an Overview—and a Look at the Fu-
ture, in: Thompson, R.A., Hudson, M.R., Pillmore, C.L. (Eds.), Geologic Excursions to the Rocky
Mountains and Beyond. Field Trip Guidebook for the 1996 Annual Meeting, Geologic Society of
America; Oct. 28 1996-Oct. 31 1996, Denver, CO. Colorado Geological Survey. Department of
Natural Resources, Denver, CO. Available online at http://www.nps.gov/archive/grsa/resources/
docs/Trip2023.pdf. Last accessed June 23, 2009.
Greene, W. 1993. Econometric Analysis. Prentice Hall, New Jersey.
Kok. K., van Delden, H. 2009. Combining two approaches of integrated scenario development to
combat desertification in the Guadalentin watershed, Spain. Environment and Planning B: Plan-
ning and Design 36, 49-66.
Kortclainen, M. 2008. Dynamic environmental performance analysis: a Malmquist index approach.
Ecological Economics 64, 701-715.
Mayer, A. 2008. Strengths and weaknesses of common sustainability indices for multidimensional
systems. Environment International 34, 277-291.
Mayer, A., Thurston, H., Pawlowski, C. 2004. The multidisciplinary influence of common sustain-
ability indices. Frontiers in Ecology and Environment 2, 419-426.
NPS (National Park Service). 2007. Final Summary of General Management Plan/Wilderness
Study/Environmental Impact Statement. US Department of the Interior. Mosca, CO.
Neumayer, E. 2010. Weak versus Strong Sustainability: Exploring the Limits of Two Opposing
Paradigms. Edward Elgar, Northampton, MA.
Pearce. D. 2002. An intellectual history of environmental economics. Annual Review of Energy
and the Environment 27, 57-81.
Shmelev, S., Rodriguez-Labajos, B. 2009. Dynamic multidimensional assessment of sustainability
at the macro level: the case of Austria. Ecological Economics 68, 2560-2573.
Yeh, S.C., Chu, Y.P. 2004. Economic-environmental scenario simulations to measure sustainable
development in Taiwan. Advances in Architecture Series 18, 483-492.
Zhou, P., Aug. B. W, Poh, K.L. 2007. A mathematical programming approach to constructing com-
posite indicators. Ecological Economics 62, 291-297.
-------
1.0
Introduction
1.1 Concept of Sustainability
The concept of sustainability and the efforts to
implement sustainability are, at: their core, an
undertaking to insure the needs of current and future
generations arc met. This fundamental concept is
embodied in one form or another in a number of public
pronouncements on the subject of sustainability. One
frequently cited quote from the World Commission
on Environment and Development (1987: 43) defines
sustainability as "development that meets the needs of
the present without: compromising the ability of future
generations to meet their own needs." It is important
to note the associated economic, social, and supporting
environmental systems must work in concert to maintain
a desired level of functioning and these are sometimes
referred to as the pillars that support sustainable
development (e.g., UN 2002). Sustainability, however,
is an anthropocentric issue that speaks to the need for
maintaining those interrelated systems over the long term
for the benefit of humanity.
The concept of sustainability is very broad,
complicated, and challenging and it can lead to different
interpretations (e.g., see Prugh et al. 1999). For
example, in the political arena, sustainability is often
associated with ideas from social science including
social justice and equity of wealth distribution (Prugh
et al. 1999). In an economic sense, Pezzey and Toman
(2002: 3-4) state a sustainable economic development
path occurs when "expected well-being per capita
(broadly defined to include more than just material or
market goods) rises over the long term." Stavins et
al. (2003) provide another economic perspective that
requires both dynamic efficiency and intergenerational
equity for their definition of sustainability. In a strict
ecological sense, sustainability is concerned with
the conservation of ecosystem processes, including
aspects such as biological diversity (Callicott and
Mumford 1997, Aarts and Nienhuis 1999). Engineers
focus on the efficiency of energy and material use in
industrial processes, and in reducing waste with some
consideration of social and economic factors (Mihelic et
al. 2003, Sikdaretal. 2004).
Although there is not a consensus on defining
sustainability, there is general recognition the current
level of human activity appears to damage the systems
that support us. Sustainability relates to finding and
maintaining a set of system conditions that can support
the social and economic development of a human
population. To be sustainable, development must not
have major adverse effects on the environmental systems
that support the population. Consequently, metrics are
needed to help us understand the linkages and demands
of human activity (e.g.. society, economy, government,
industry, etc.) on environmental systems (Cabezas et al.
2005a).
When choosing sustainability metrics, researchers should
be aware of relevant issues such as the paradigms of
weak and strong sustainability, which relate to how much
man-made and other capital can substitute for natural
capital (described in Chapter 2 of this report, Neumayer
2010). In addition, there is some amount of uncertainly
and unreliability in measures of sustainability and
most measures have many simplifying assumptions
that complicate this type of research. Given these
complexities, a collaborative, interdisciplinary approach
could alleviate some of the conflict and ambiguities in
these interpretations and approaches.
The Federal government's sustainability research
strategy recognizes that an integration of economic,
social, and environmental policies is required to achieve
sustainability (USEPA 2007). Therefore, the strategy
states (USEPA 2007: 23):
EPA and its partners will develop integrating
decision support tools (models, methodologies,
and technologies) and supporting data and
analysis that will guide decision makers toward
environmental sustainability and sustainable
development.
To translate the aforementioned mandate from the
USEPA strategy into a practical plan of action requires
a working definition of sustainability. Thus, we define
sustainability as identifying and maintaining a set of
conditions that support environmental, social, and
economic systems that meet: the needs of both current
and future generations, thereby addressing the three
pillars of sustainability. Moreover, in order to satisfy
the definition of sustainability, the environmental,
social, and economic systems must: effectively meet the
needs of current and future generations, indefinitely. To
assess sustainability, information is needed to evaluate
human well-being, how the different components of a
system are linked, and the stresses of human activity
(e.g., society, economy, government, industry, etc.) on
-------
environmental systems. One way to approach this and
guide decision makers toward sustainabiiity is through
metrics that quantify changes in environmental, social,
and economic systems.
The research and results presented in this report will
provide decision support through the estimation of
metrics that relate to sustainabiiity. These metrics
will help decision makers understand the movement
of their system toward or away from environmental
sustainabiiity at the regional level. The San Luis
Basin (SLB), in south-central Colorado, was used as a
pilot study to develop and test the applicability of the
methodology.
1.2 The Luis
Project
Over the long-term, the maintenance of intact and
sustainable ecosystems depends on the ability of all
landowners to manage and utilize their lands in a
sustainable manner. Because the Federal Government
owns more than 265.9 million hectares (GSA 2004),
the US has a vested interest in managing these lands
in a sustainable manner. For example, the US Forest
Service Mission is to "Sustain the health, diversity,
and productivity of the Nation's forests and grasslands
to meet the needs of present and future generations"
(USDA 2007: 2). Understanding the effects of
facility operations and management decisions on the
environment at the regional scale is essential to the
work of several programs within the USEPA. This is
especially relevant in the west, where the majority of
the US public lands are located. In fact, over thirty-
three percent of the land within USEPA Region 8, where
the SLB is located, is publicly owned (USEPA 2009a).
Thus, choices made daily by land managers are an
important component of how well the environment in the
Region's six states (Colorado, Montana, North Dakota,
South Dakota, Utah, and Wyoming) is protected.
Public land managers make decisions daily that
can affect the quality of the land for which they are
responsible. These decisions are based on several
factors, including the mission of the agency they serve;
the directives of the current political administration; and
the wishes of the surrounding population. To assist land
managers and the surrounding communities in making
sustainable choices, USEPA needs to develop integrated,
system-based tools that provide feedback on how these
collective decisions are affecting ecosystems and their
sendees. To fulfill this need, the USEPA Office of
Research and Development initiated the San Luis Basin
Sustainabiiity Metrics Project in 2006. The results of
this project will provide a set of sustainabiiity metrics for
use on the regional scale that are scientifically credible
and assist land managers. Land managers can use the
methodology to identify trends as the system progresses
through time, and provide early warning of abnormal
conditions. This will help them better protect and
manage healthy ecosystems.
This report presents an overview of the San Luis Basin
Sustainabiiity Metrics Project, a five-year research
program. The purpose of the report is to present the
methodology, discuss approaches used to estimate the
individual metrics, and examine results that could inform
decision makers in the region. The goal of this study
was to produce a straightforward, relatively inexpensive
methodology that was simple to use and interpret.3
This required historical data to be readily accessible,
that metrics be applicable to the relevant scale, and
that results meet the needs of decision makers. The
objectives were: 1) determine the applicability of using
existing datasets to estimate metrics of sustainabiiity at
a regional scale, 2) calculate the metrics through lime
(initial proposal was 1980-2005), and 3) compare and
contrast the results to determine if the system is moving
toward or away from sustainabiiity.
1.2.1 Sustainabiiity
In general, identifying major properties (e.g., climatic,
soil, biological diversity, etc.) and cycles (e.g., water,
carbon, phosphorus, oxygen, nitrogen, etc.) of a system
may seem relatively simple and straightforward, but
it is difficult to list and quantify many of the specific
details of the properties and cycles critical to the needs
of humans and, therefore, to sustainabiiity. As a starting
point, we identified and tested a set of four metrics
that capture some of the most basic properties of a
system. The four metrics included Ecological Footprint:
Analysis (EFA), Green Net Regional Product (GNRP),
Emergy Analysis (EmA), and Fisher information (FI:
more discussion on the choice of metrics can be found
in Chapter 2). These metrics, respectively, represent:
human burden on the environment, economic well-
being, flow of available energy through the system, and
overall system order. Each metric captures particular
aspects of Hie system, but may not account for every
aspect pertinent to sustainabiiity. There is some overlap
or redundancy (e.g., consider a physician using multiple
tests to examine a patient's health) in the variables
used and, therefore, the pillars they quantify and how-
sensitive each metric is to the pillars of sustainabiiity. In
3 We believe the method is relatively inexpensive because we have
identified the majority of data sources for these types of calculations
leading to reduced time and resources necessary for others to compute
the metrics. However, whether we met the goal and objectives would
be based on anecdotal evidence and discussions with stakeholders in
flic region.
-------
fact, this issue of sensitivity is one of the main reasons
we initially used multiple metrics.
Note that for this report, we use metric and index
interchangeably; however, we acknowledge a difference
between indicator and either of the terms we use.
Mayer (2008) defines an indicator as measuring one
characteristic (e.g., CO2 emissions or biodiversity),
but an index (or metric) combines many indicators (or
variables) through aggregation (e.g.. GNRP). Therefore.
indices or metrics can provide a multidimensional view
of sustainability.
The four metrics we utilized are capable of measuring
sustainability in an absolute sense. For example, if
the ecological footprint (demand) is larger than the
available biological capacity (supply), then the system is
considered unsustainable following the assumptions and
methods of that approach. It is, however, conceptually
and practically difficult to compute values for these
metrics with absolute accuracy. The conceptual
difficult}' arises because each of these metrics attempts
to convert many disparate variables or indicators into
one common measure. For example, consider that EFA
converts everything into land area, EmA converts all
quantities into solar emjoules, GNRP requires that all
quantities be expressed monetarily, and FI converts
system variables into a measure of dynamic order. This
conversion exercise is relatively simple and accurate as
long as the quantity is closely related to the common
measure of the metric (e.g., converting agricultural
production into land area). This is more difficult and less
accurate when converting a quantity7 that is not closely
related to the common measure (e.g.. converting human
transportation needs into information). The practical
difficult}7 comes from the impossibility of measuring
or estimating values for the innumerable variables or
indicators needed to characterize a system well enough
to compute accurate values for the metrics. Hence, it
is likely that even under the most extraordinary efforts
inaccurate values for the metrics would result because of
measurement errors, errors in the conversion of unrelated
variables into a common measure, and/or missing
variables.
The aforementioned reasons have led us to adopt the
strategy of focusing not on computing absolute values
for the metrics but on identifying changes in the value
of the metrics through time;1 We believe the benefits
of this strategy7 are: 1) errors incurred in complicated
calculations due to either conceptual or practical issues
often tend to cancel out when differences are taken, 2)
4 Without a measure of uncertainty for any of the metrics, it is not pos-
sible to determine if any change over time is statistically significant.
We address this in detail in Chapter 7.
it is easier to formulate scientifically defensible criteria
for changes than for absolute values (but see Alonso
1968). 3) components that do not change over time can
be dropped or simply approximated, and 4) trends in the
metrics rather than absolute values are most likely to
indicate movement toward or away from sustainability.
Knowing how the system is evolving over time rather
than its state at a point in time is, perhaps, the most
critical issue. There is a body of scientific theory that
tells us what changes in these metrics mean in terms of
sustainability (e.g., Campbell 2001, Heal and Kristom
2005, Mayer 2008). We established criteria for each
metric to determine if a system is moving toward
sustainability'. For EFA, ecological balance (biocapacity
- ecological footprint) is stable or increasing with time.
For EmA, the fraction of renewable emergy used to
total emergy used is trending to one. FI and GNRP
are one-sided tests of weak sustainability and they
cannot technically show if the system is moving toward
sustainability. Criteria for these one-sided tests indicate
if the region is moving away from sustainability when
FI is steadily decreasing with time (Karunanithi et al.
2008) and GNRP is decreasing with time (i.e., Pezzey
et al. 2006). Note these criteria do not presuppose any
particular version of the system; rather they address
the maintenance and preservation of basic properties
and processes that are necessary for continued human
existence. An unsustainable path is defined as one where
any of the following are observed - human welfare (as
defined by GNRP), ecological balance, emergy flows
into the system, or Fisher information - are in violation
of the aforementioned criteria. We considered a
system that is moving away from sustainability, for that
particular aspect of the system, to be on an unsustainable
path, when one or more of the four criteria are not
satisfied and/or a one-sided test is satisfied.
1.2.2 Choice of Region
The SLB in Colorado (Fig. 1.1) was selected as the
study area because of its distinct natural hydrological
boundaries, limited population, large amount of publicly
owned land, and interest expressed in the study by
USEPA Region 8 and other government officials.
Surrounded by the Sangre de Cristo Mountains to
the east and the San Juan Mountains to the west,
approximately seventy'-eight percent of this region
is publicly owned, making it an optimal system for
our study. This study has fostered a partnership for
sustainability research between USEPA Region 8 and
the USEPA's Office of Research and Development.
Moreover, it will expand existing partnerships between
land management agencies within USEPA Region 8 such
as the National Park Service, USD A Forest Sen-ice,
-------
US Fish and Wildlife Sendee, and Bureau of Land
Management (BLM). Some 202342 hectares (ha) on
the borders of the valley (adjacent to National Forest
lands) are managed by BLM and much of this land is
leased to neighboring ranches for grazing. The Great
Sand Dunes National Park and Preserve is the most well
known feature of the valley. Directly west of the Sangre
de Cristo Mountains, the dunes (the largest in North
America) rise to a height of over 213 m, and are under
the protection of the Park and Preserve.
1.2.2.1 Description of the San Luis Basin
We define the SLB as the following seven counties:
Alamosa, Conejos, Costilla, Hinsdale, Mineral, Rio
Grande, and Saguache. This region contains the Upper
Rio Grande River Basin, the San Luis Valley, and the
Great Sand Dunes national Park and Preserve. The
Upper Rio Grande River Basin is nearly 21000 km2 in
size (USDA 1978) and the San Luis Valley, a primary
feature of the basin, is a high desert, sitting at an average
altitude of approximately 2300 m (CWCB 2000).
The SLB is home to approximately 50,000 people (US
Census Bureau 2002). Although the study area is in
the State of Colorado, the region borders New Mexico
to the south and, based on conversations with local
stakeholders, is closely related to that state culturally
and economically. This relationship is not surprising
considering the geographic features that define the
region to the north, east, and west. The economy is
predominantly based on agriculture (San Luis Valley
Development Resources Group 2007). The agriculture
is dominated by potatoes, small grains, and alfalfa
and is dependent on irrigation (approximately 85% of
the water consumption in the valley is for agriculture;
CWCB 2000, Wurster et al. 2003, NASS 2009) and
future demand for irrigation water could increase one-
hundred percent (HRS Water Consultants 1987). Most
of the irrigation water comes from surface water that
is collected and channeled through a series of historic
canals (San Luis Valley Development Resources Group
2007).
The SLB sits atop the Rio Grande Rift and is drained to
the south by the Rio Grande River (Mayo et al. 2007).
The northern section does not drain into the Rio Grande
River and is referred to as the Closed Basin (CWCB
2000). The Rio Grande River begins to flow just south
of the Closed Basin (Mayo et al. 2007). The San Luis
Valley has an annual average precipitation in the basin
ranging from 17 cm at Alamosa to 114 cm at Wolf Creek
Pass (CWCB 2000), but much of the surface water
begins as snowmelt coming out of the mountains (Emery
1979).
Given the structure of the region, the groundwater
system is complex. According to Mayo et al. (2007).
there are three groundwater sources within 1200 m of
the surface: 1) an unconfined aquifer, 2) an upper active
confined aquifer, and 3) a lower active confined aquifer.
The unconfined aquifer is the closest to the surface
compared to the confined aquifer (Mayo et al. 2007)
and it provides water to valley residents and farmers
and maintains the levels of the San Luis Lake and other
wetlands just to the west of the dunes (Wurster et al.
2003). The confined aquifer occurs between about 305
m to 1829 m below the surface (HRS Water Consultants
1987). Although layers of clay separate the different
aquifers, there is some leakage between the unconfined
and confined aquifers (Mayo et al. 2007). In addition
to aquifer leakage, recharge occurs through stream
infiltration, mountain underflow, overland infiltration,
and canal and irrigation infiltration (Mayo et al. 2007).
The combined use of ground and surface water has been.
and will continue to be, of major importance to proper
management of the valley's water resources (Emery
1979). The Conejos River has lost some of its flow from
withdrawal from the confined aquifer and increased
consumptive use near the river (Emery 1979). This
loss of water has caused problems with meeting die Rio
Grande Compact which requires Colorado to deliver
a certain amount of water to New Mexico and Texas
(Emery 1.979, CWCB 2000). In order to meet the water
requirements of the Compact, the Closed Basin Project
pumps water from die Closed Basin using wells and
canals and discharges the water into the river (CWCB
2000).
1.2.3 Community Outreach
In order to produce a useful product, the USEPA needed
to bring the local residents into the project to let diem
know what we were doing, why we were doing it and
what the potential outcomes might be. The purpose
of the outreach is that once the project is complete,
the metric tools would be accepted and used by the
community. In addition, USEPA needed to advertise the
project well enough so as we collected information in the
SLB, local residents could contact us with questions.
The outreach approach developed and implemented by
USEPA Region 8 focused on involving people who were
identified as community representatives and leaders:
* Federal Land Managers - Because 69% of the land
is federally owned and managed (San Luis Valley
Development Resources Group 2007), federal
land managers from the National Park Service,
USD A Forest Sendee, BLM, and US Fish and
Wildlife Sendee played a significant role in the
-------
outreach efforts.5 These managers were engaged
in developing the metric tool for their own use
and are seen as leaders within the SLB community
and its members. Our connection with the federal
land managers in the valley served to identify key
community leaders. Because these individuals were
well known and respected within the community,
their support and introduction of ORD scientists
to the community was vital to being accepted by
community leaders.
« Community Organizers and Non-profit groups - We
had a list of community organizers and interested
parties from previous work the USEPA performed
in the San Luis Valley on drinking water (USEPA
2009b).
* Adams State College (ASC) - ASC, a prominent
institution in the community, has a stated goal of
reaching out as a partner in (lie community and has a
well-established Community Partnerships Program
(CPP) whose mission is to connect ASC resources
with the community in order to increase the quality
of life for all residents of the SLB. ASC CPP was
instrumental in helping USEPA scientists connect to
the community.
* National and State representatives - The Secretary
of the Interior (Kenneth Salazar) and Colorado
Congressman (John Salazar) arc residents.
landowners, and fanners in the community. These
local representatives, as well as US Senators
Michael Bennel and Mark Udall, were informed of
the project and engaged in the community outreach
efforts.
Once we established contact with local leaders, a USEPA
Region 8 representative spent a week in the community
meeting with these leaders. Desired outcomes included
informing the individuals about the project; receiving
feedback on how to inform the community about the
project; integrating their concerns into the project
design; receiving feedback on the use of the tool within
the community: and developing an expanded list of
community leaders that we should contact. More than 30
such meetings were held and the results were shared with
the project team. Community leaders collectively asked
for a formal project overview in order to understand
better the usefulness and application of the metrics to
the local community. Working with ASC and CPP, the
USEPA representative scheduled a project overview
meeting and held a question and answer event. Seventy-
5 The San Luis Valley Development Resources Group defines the Val-
ley as the following six counties: Alamosa, Conejos, Costilla, Mineral,
Rio Grande, and Saguache. This differs from our definition of the SLB
by not including Hinsdale County and accounts for the difference in the
proportion of land publicly owned (-78%) stated previously.
five letters of invitation were sent to community leaders
who were identified during the first phase of the outreach
effort. Over 40 people attended the project overview
meeting and USEPA received positive feedback and an
indication the community welcomed the project focus.
CPP notified the USEPA of an opportunity to develop
and present a challenge to an upper level advertising-
class (BUS 345) in the School of Business as their
class project. The USEPA representative presented an
overview of the project and challenged the students
to develop an outreach campaign directed to the SLB
community. The campaign would inform the community
about the project and provide information on who to
contact within the project to get questions answered or
provide feedback from the general community about the
project. The students selected the challenge as the class
project and developed a work plan that included a public
meeting, press releases, and identification of potential
media connections for additional outreach efforts.
The class research identified five areas of outreach
opportunity and/or concern: 1) the SLB residents were
unaware of the project, 2) the most effective form of
advertising in the SLB is through the local newspaper,
3) the primary concern the residents had with the project
was USEPAs purpose for conducting the research, 4)
public meetings are an effective form of informing the
SLB residents, and 5) SLB residents initially viewed
the project as having a negative impact on the region.
Based on these findings, the class requested a conference
meeting with a USEPA representative to establish
firm goals and objectives for the outreach activities
and brainstorm ideas for an advertising project. The
approach developed by the class included an organized
town meeting to inform the community about the
project. The class advertised the town meeting in the
newspaper and placed notification of the meeting on the
Chamber of Commerce web page. This effort revealed
the community was concerned the project could result in
new regulations, more government control of land use
decisions in the region, and concern that USEPA would
acquire land from farmers and residents. Based on these
findings, we sought opportunities to meet with the public
and hand out information describing the overall project.
The information on these postcards included a contact
name and phone number for the Region 8 representative
and had the following text (see footnote 3 above for the
six counties of the Valley):
This is to let you know that the Environmental
Protection Agency is working in your community
to gather information for a research project. The
project will help build a scientific foundation
for sound environmental decision-making. This
-------
is a research project only. Our goal is to provide
the San Luis Valley with useful tools for future
planning. This research project will not result in
any action by the USEPA in the San Luis Valley.
These postcards were distributed at public events
identified by the community leaders as places where the
community gathered and would be receptive to hearing
about the project. Additional outreach was conducted
by CPP and included seminars open to the public. For
example. USEPA received invitations to speak at public
events. One such event was Focus the Nation, a national
effort to have colleges and universities focus on one
significant issue for one day. Another public event was
the annual meeting of the San Luis Valley Resource
Conservation and Development Board, a USD A
program designed to assist communities to accelerate
the conservation, development, and utilization of natural
resources in order to improve their economic activity.
USEPA held its annual team meetings in the region
and in 2007 and 2008 invited community leaders to
participate and provide valuable community perspectives
to the project.
This community outreach approach was effective in
meeting our goal of informing the community in order to
increase the probability the community would accept and
use the tools. We presented preliminary results from the
project to the community and they were well received
by the community and the attendees provided valuable
feedback and comments to the research group.
1.3 Outline of Report
Having introduced some of the background for this
pilot project, the report proceeds as follows: first, we
present a literature review of sustainability metrics in
Chapter 2. Because of the abundance of literature on
sustainability metrics and classification systems, we only
reference studies that report lists of metrics. The chapter
focuses on metrics that fit within categories of ecology,
economics, energy, and system order. In addition, this
chapter summarizes studies that have looked at multiple
metrics over time. Chapters 3 through 6 present the
four metrics used in the SLB pilot project: EFA, GNRP,
EmA, and FI, respectively. Each of these chapters
presents an explanation of the methodology, the data and
their sources, results, and discussion. The remaining
chapter summarizes the results of all the metrics and
discusses the strengths and limitations of estimating four
sustainability metrics over time using publicly available
data. Seventeen appendices cover additional results or
methodologies (e.g., estimates for wind erosion) used in
estimating the metrics.
1.4
Aarts, B.G.W., Nienhuis, P.M. 1999. Ecological
Sustainability and Biodiversity. International Journal
of Sustainable Development and World Ecology 6,
89-102.
Alonso, W. 1968. Predicting best with imperfect data.
Journal of the American Institute of Planners 34, 248-
255.
Callicott, J.B., Mumford, K. 1997. Ecological
Sustainability as a Conservation Concept.
Conservation Biology 11, 32-40.
Campbell, D. 2001. Proposal for Including What
Is Valuable to Ecosystems in Environmental
Assessments. Environmental Science and Technology
35, 2867-2873.
CWCB (Colorado Water Conservation Board). 2000.
Rio Grande Decision Support System: An Overview.
Section 2.0 Rio Grande Basin Conditions. CWCB,
State of Colorado. Available online at http://cdss.sMe.
co.us/DNN/RioGrande/tabid/57/Default.aspx. Last
Accessed July 21, 2009.
Emery, P. 1979. Geohydrology of the San Luis Valley,
Colorado, USA. Proceedings of the Canberra
Symposium 1979. IAHS-AISH Publication no. 128,
297-305.
GSA (General Sendees Administration). 2004. Federal
Real Property' Profile for FY2004. US General
Sendees Administration, Office of Governmenrwide
Policy. Available online at http://www.gsa.gov/
realpropertyprofile. Last accessed July 21, 2009.
Heal, G.. Kristom, B. 2005. National income
and the environment, in: Maler, K., Vincent, J.
(Eds.). Handbook of Environmental Economics,
Economy wide and International Environmental Issues,
Vol. 3. Elscvicr, North-Holland, Amsterdam.
HRS Water Consultants. 1987. San Luis Valley
Confined Aquifer Study Final Report (Phase I).
Consulting report, 87001-17, Lakewood, CO.
Mayer, A. 2008. Strengths and weaknesses of common
sustainability indices for multidimensional systems.
Environment International 34, 277-291.
Mayo, A.L.. Davey, A., Christiansen D. 2007.
Groundwater flow patterns in the San Luis Valley,
Colorado, USA revisited: an evaluation of solute and
isotopic data. Hydrogeology Journal 15, 383-408.
-------
Mihelic, J.R., Crittenden, J.C., Small, M.J., Shonnard,
D.R., Hokanson, D.R., Zhang, Q., Chen, H., Sorby,
S.A.. James, V.U., Sutherland, J.W., Schnoor, J.L.
2003. Sustainability Science and Engineering: The
Emergence of a New Metadiscipline. Environmental
Science and Technology 37. 5314-5324.
NASS (National Agricultural Statistics Sen-ice, United
States Department of Commerce, United States
Department of Agriculture). 2009. United States
Census of Agriculture: 2002. Available online
at http://www nass.usda.gov/Statisticsj3y __State/
Colorado/index.asp. Last accessed June 26, 2009.
Pezzey, J., Hanley, N., Turner, K., Tinch, D. 2006.
Comparing augmented Sustainability measures for
Scotland: Is there a mismatch? Ecological Economics
57, 60-74.
Pczzcy, J., Toman, M. 2002. Making Sense of
"Sustainability." Issue Brief 02-25. Resources for the
Future. Washington, DC.
Pragk T, Costanza, R., Daly, H. 1999. The local
politics of global Sustainability. Island Press,
Washington, DC.
San Luis Valley Development Resources Group. 2007.
Comprehensive economic development strategy.
Available online at http://www.slvdrg.org/ceds.php.
Last accessed June 17, 2009.
Sikdar, S.K., Glavic, P., Jain, R. (Eds.). 2004.
Technological Choices for Sustainability. Springer-
Verlag, Berlin, Germany.
Stavins, R., Wagner, A., Wagner, G. 2003. Interpreting
Sustainability in economic terms: dynamic efficiency
plus intergenerational equity. Economic Letters 79,
339-343.
UN (United Nations). 2002. Report of the World
Summit on Sustainable Development. Johannesburg,
South Africa, 26 August-4 September 2002. A/
CONE 199/20. ISBN 92-1-104521-5. United Nations,
New York, NY.
US Census Bureau. 2002. 2000 Census of Population
and Housing, Summary Population and Housing
Characteristics, PHC-1-7, Colorado. Washington, DC.
USDA (United States Department of Agriculture). 1978.
Water and Related Land Resources, Rio Grande Basin
Colorado. A Cooperative Study with the Colorado
Water Conservation Board. U. S. Government
Printing Office. Volumes 1 and 2.
USDA (United States Department of Agriculture). 2007.
USDAForest Sendee Strategic PlanFY2007-2012.
FS-880. Forest Sendee. Washington, DC.
USEPA (US Environmental Protection Agency). 2007.
Sustainability Research Strategy. EPA600/S-07/001.
Office of Research and Development. Washington,
DC.
USEPA (US Environmental Protection Agency). 2009a.
About EPA Region 8. Region 8. Denver, CO.
Available online at http://www.epa.gov/region8/about/
index html. Last accessed July 21, 2009.
USEPA (US Environmental Protection Agency). 2009b.
San Luis Valley Drinking Water Well Project. Region
8. Denver, CO. Available online at http://www.epa.
gov/region08/ej/ejinitiatives.html#slv. Last accessed
August 19, 2009.
World Commission on Environment and Development.
1987. Our Common Future. Oxford University Press,
Oxford, UK
Wurslcr, F.C., Cooper, D.J., Sanford, W.E. 2003.
Stream/aquifer interactions at Great Sand dunes
National Monument. Colorado: influences on
interdunal wetland disappearance. Journal of
Hydrology 271, 77-100.
-------
[ Upper Rio Grande River Basin
SLB counties
I I
0 12.5 25
I
50 Kilometers
Figure 1.1 - Seven county region referred to by the investigators as the San Luis Basin (SLB), in south-central
Colorado.
-------
2.0
Sustainability Metrics
2.1 Introduction
In this chapter, we review the literature of different
measures of sustainability for the four categories
described in Chapter 1: ecology, economics, energy,
and system order. Once we provide background on our
selected metrics, we review the literature of studies that
have examined sustainability using multiple metrics,
and those studies that examined suslainabilily over
time. Both arc important attributes of the San Luis Basin
Sustainability Metrics Project. The last section justifies
our choice of the four metrics from within the four broad
categories.
2.2 of Ecological,
Economic, Energy, Order
Metrics
Articles such as Bohringer and Jochem (2007),
Gasparatos et al. (2007), Hak et al. (2007), Mayer et al.
(2004), Mayer (2008), Ness et al. (2007), Singh et al.
(2009), UN (2007), and Wilson et al. (2007) summarize
different sustainability metrics and indicators. Many of
these papers group the indices and indicators into broad
categories and then summarize the individual measures
in those groups. Neumayer (2010), for example,
presents the two opposing paradigms of weak and strong
sustainability along with metrics that fall under these
paradigms. Weak sustainability assumes ecological
functions and/or resources (i.e., natural capital) can
be replaced by technological or man-made surrogates,
whereas strong sustainability assumes ecological
functions are not replaceable and natural capital cannot
be substituted for with other forms of capital (e.g., Dietz
and Neumayer 2007, Mayer et al. 2004, Mayer 2008,
Neumayer 2010, Pearce 2002). For additional detail on
how the specific indices mentioned in this report relate
to sustainability, we suggest examining some of the
papers listed above. Rather than repeating the extensive
efforts above, we present a subset within the group of
ecological, economic, energy, and system order metrics
of sustainability.
2.2.1 Ecological Sustainability
A number of indices have been proposed that measure
environmental or ecological sustainability and are based
on population biology. Specifically, the indices are based
on the concept of carrying capacity and they attempt
to assess the number of individuals an ecosystem can
maintain by quantifying the load a system can support
and the impact organisms have on the system (Caughley
1976). In other words, carrying capacity is the number
of individuals of a given species that can be supported
by the available resources in an ecosystem (Sharkey
1970). Three such metrics based on carrying capacity
that have been suggested as measures of sustainability
are maximum sustainable yield (MSY), IPAT, and
Ecological Footprint Analysis (EFA).
MSY attempts to identify the resilience of a population
to harvest or extraction (e.g.. Fry 1947, Ricker 1958,
Schaeffer 1954, but sec Pitcher and Hart 1982). To be
sustainable, resources must be harvested at a rate that
allows a population to replace or renew the extracted
biomass to avoid depleting llie supply. Implementing
this management strategy has been difficult (Larkin
1977) due to extreme variability in stock recruitment
for some species, which was noted as a possible flaw by
Russell (1931).
Another such metric adapted from carrying capacity
by Ehrlich and Holdren (1971) is TPAT. TPAT proposes
the Impact a population lias on its environment is
directly related to the product of Population size,
level of Affluence, and the Technology employed by
the population (e.g., York et al. 2003). Ehrlich and
Holdren (1971) assumed an increase in any of the
parameters would result in higher overall impact on
the environment. A number of modifications have
been made to Hie metric to examine if each parameter
should have equal contribution (e.g., Sutton 2003,
Waggoner and Ausubcl 2002, York et al. 2002 and
2003) or if additional factors should be included (e.g.,
Fan et al. 2006, Schulze 2002). For example, Dietz and
Rosa (1994) converted IPAT to STIRPAT, a non-linear
regression equation that allows random errors (i.e.,
stochasticity) in the estimation of the parameters and
permits hypothesis testing (e.g., Dietz et al. 2007, Mayer
2008). Interestingly, Dietz et al. (2007) use ecological
footprint (see below) as their measure of anthropogenic
stressors in their STRIPAT approach.
Rees (1992) and Wackernagel and Rees (1996)
applied the concept of carrying capacity to humans
by inverting the number of individuals a system can
support and measuring the amount of biologically
productive (bioproductive) land that is required to
support a population of a given size. This area of land
necessary for the production and maintenance of goods
and services consumed by the population was termed
-------
the ecological footprint (the metric is named after this
component of the overall analysis). Ecological footprint
is based on the concept of strong sustainability and
it attempts to capture ecological impacts of human
activity from resources consumed and wastes generated
(Neumayer 2004). Specifically, ecological footprint
includes consumption of crops, meat, seafood, wood
and fiber, energy, and living space, which are converted
to one of six biologically productive land types (i.e.,
arable, pasture, sea, forest, energy, and built lands,
respectively). Comparing the "supply" of bioproductive
land to the "demand" of resource consumption and waste
generation can reveal if the system is sustainable at the
current levels of human impact. Although there has been
a number of modifications and adaptations to improve
the ecological footprint methodology since its inception
(e.g., Lenzen and Murray 2001, Luck et al. 2001,
Monfreda et al. 2004, Medved 2006, Venetoulis and
Talberth 2005, 2008, Wackernagel 2009), there has been
a push for standardizing the methods (e.g., Wackernagel
2009). Nonetheless, new metrics have been proposed
that combine existing indexes with ecological footprint
such as emergy (e.g., Hu et al. 2009, Siche et al. 2010a,
2010b, Zhao et al. 2005), exergy (e.g., Hau and Bakshi
2004), and net primary production (e.g., Siche et al.
2010b, Venetoulis and Talberth 2005 and 2008), for
example.
Each of these metrics relates, perhaps indirectly, to
a fourth sustainability metric, resilience. Resilience
refers to the amount of perturbation a system can absorb
before shifting to an alternate dynamic regime (Grimm
and Wissel 1997, Holling 1973). Human activities can
increase or decrease the resilience of a regime, or set of
system conditions, that provides desirable ecosystem
services (Mayer et al. 2004) and is beneficial to humans
(Carpenter et al. 2001, Jackson et al. 2001, Orr 2002),
thereby increasing or decreasing the carrying capacity.
2.2.2 Economic Metrics
Economists define sustainability as non-declining
welfare or well-being over generations (Neumayer 2010,
Pearce and Atkinson 1995, Tietenberg 2000). Welfare,
as defined by economists, incorporates benefits obtained
outside of market transactions such as environmental
amenities, but allows for the possibility of trade-offs
between market and non-market goods. In this sense, the
field of economics can offer useful metrics that are based
on the concept of weak sustainability.
Augmented economic accounts, often referred to as
green accounting, is a common approach to estimating
welfare, with the trends in augmented accounts
measuring sustainability (Adamowicz 2003). Green Net
Domestic Product (GNDP) and Genuine Savings are two
of the most important economic metrics that account for
depreciation of natural resources (e.g., Dasgupta 2001,
Pearce and Atkinson 1993, 1995, Pezzey et al. 2006).
We define both GNDP and Genuine Savings in this
Section.
Gross Domestic Product (GDP) is one measure that
captures the market value of a country's output (and
therefore, its aggregate consumption). However, it
does not measure welfare because it excludes the value
of leisure time and services produced by nature (van
den Bergh 2009). Changes in GDP over time are often
assumed in casual discourse to approximate changes
in welfare over time. However, such a measure is
unsatisfactory for purposes of measuring sustainability.
An economy that eats the proverbial seed corn or allows
natural capital to depreciate faster than it is replaced,
destroys the region's capacity to maintain the current
standard of living into the future and is, by definition,
unsustainable. The productive capacity of the economy
will be reduced in future years, eventually resulting in a
decline in human welfare. However, such an economy
is likely to show positive GDP growth for some period
of time. GDP measures the consumption and production
of market goods, while ignoring depreciation, pollution,
and the services produced by nature. Consumption of
seed corn and unsustainable natural resource extraction
contribute to GDP by increasing current consumption,
while the costs of reduced productive capacity in the
future are not reflected in current GDP. Therefore, trends
in GDP do not indicate the sustainability of an economy.
For the economic accounts to measure sustainability, the
depreciation of all capital stocks should be included in
the calculation. Net Domestic Product (NDP) is a step
in the right direction. It modifies GDP by subtracting
the depreciation of man-made capital. However, it does
not incorporate all types of capital such as human, social,
and natural capital; therefore, it is not an ideal measure
of sustainability (Azqueta and Sotelsek 2007, Hartwick
1990, Weitzman 1976). Natural capital stock can include
both non-renewable and renewable resources, which use
different approaches for their estimation (see Hartwick
1990, Vincent 2000). Hartwick (1990) shows how
pollution control costs could be incorporated into GNDP
and how the stock of pollution, when it enters the utility
function (i.e., pollution is a capital bad rather than a
capital good), changes the calculation of GNDP.
GNDP explicitly incorporates the depreciation of man-
made, human, and natural capital. Following Weitzman
(1976, 2000) and Vincent (2000) who provide a good
summary, we define GNDP as
GNDP(t) =
C(t) + \
h(t)Kh(t)
n(t)K"(t)
(2.1)
-------
For tliis equation, C(t) is aggregate consumption through
time, t; X(t) are the shadow prices (i.e., the value of
having an additional unit) of the different capital stocks;
total capital stock includes man-made capital (Km).
human capital (Kh), and natural capital (Kn); and the
stock of man-made, human, and natural capital changes
over time is represented with the dot notation. The
first two terms on the right side of Equation 2.1 are
the standard definition of net domestic product (NDP)
which includes the depreciation of man-made capital.
The last two terms transform NDP into GNDP (Cairns
2002, Pezzey et al. 2006). Although it is not included in
Equation 2.1, Dasgupta (2001) includes the changes in
the stock of knowledge in GNDP.
A single year measurement of GNDP does not
demonstrate sustainabilily (Heal and Krislom 2005).
Only trends in GNDP can give an indication of
whether an economy is moving toward or away from
sustainabilily. Dasgupta and Maler (2000). and later
corrected by Aslieim and Weilzman (2001), test to see if
AGA'I)P> 0
(2.2)
GNDP measured in real prices (adjusted for inflation)
must be increasing over time for a region to be moving
toward sustainability. Using this type of intertemporal
welfare analysis will allow stocks of capital to be
dropped from the analysis if they are not changing over
time (Usher 1981 in Ahlroth 2003 and Hartwick 1990).
Some arguments against using the change in GNDP
as a measure of sustainability and issues related to
estimating the depreciation of natural capital can be
found in Azqueta and Sotelsek (2007). Dasgupta (2001).
Hanley et al. (1999), Mayer (2008), and Nordhaus and
Kokkelenbcrg (1999). For example, Dasgupta (2001)
concludes that GNDP could increase over time while
the country becomes poorer and well-being decreases.
Rather than Equation 2.2 representing sustainability,
Dasgupta (2001: 151) states
social well-being increases during a brief interval
of time, if, and only if. the value of changes in
consumption services plus the change in the value
of net changes in capital assets is positive.
Genuine Savings (GS), which is based on the Hartwick
Rule, was introduced by Pearce and Atkinson (1993).
As stated above, a constant capital stock is necessary in
order to achieve a sustainable consumption flow, which
is the basis for GS. It is defined as
S VK
Y"J
(2.3)
stock that is defined as above in Equation 2.1 but with
different depreciation rates for each stock (Pearce and
Atkinson 1993, 1995). To move toward sustainability,
a region's GS would have to be 0 or positive. This
condition for sustainability does not have any restrictions
for substitution between man-made and natural capital.
As long as GS remains constant or increasing, natural
capital could be degraded (Pearce and Atkinson 1995).
Hamilton and Clemens (1999) have published GS results
for developing countries through the World Bank. An
argument against GS can be found in Pillarisetti (2005)
and criticisms related to GS are described in Dietz and
Neumayer (2004).
Before we discuss whether GNDP or GS are appropriate
for measuring sustainability. we recognize there is a
relationship between GNDP and GS such that GNDP
is equal to consumption plus GS (Dasgupta 2001).
Put another way, GS is measured as GNDP minus
consumption (Hamilton et al. 1997, Pearce and Atkinson
1995). Pezzey et al. (2006) define augmented Green Net
National Product6 equal to consumption expenditures
plus augmented GS. where the augmentation term relates
to exogenous technical progress.
Recent papers have stated that examining the trend of
GNDP or GS actually is not an appropriate measure for
sustainability. Some describe it as a one-sided measure
and it only reveals when the region is moving away from
sustainability (e.g., Pezzey 2004, Pezzey et al. 2006).
Hanley et al. (2002: 6) summarizes this discussion:
The academic consensus at the moment is that
both green NNP [what we call GNDP] and GS
are one-sided indicators of weak sustainability
in that negative GS or falling green NNP signal
non-sustainability. However, the measurement of
a positive genuine savings rate/rising green NNP
at a given point in time is not sufficient to lead to
the conclusion that the economy is on a sustainable
path and further evidence must be sought before
any firm judgement can be made.
Apart from the one-sidedness identified by Hanley
above, several factors may influence the relationship
between GNDP (and GS) and sustainability. Exogenous
technological progress changing world interest rates
or terms of trade may compensate for changes in net
investment. Some of these factors are reflected in trends
in GNDP, but Pezzey (2004) addresses exactly how this
For this equation, S is savings. Y is income, yr is
depreciation on all capital stock, and K is total capital
6 Gross Domestic Product and Gross National Product both measure
the market value of goods and services but they differ in terms of who
produces the goods and services (US residents regardless of location
counts towards Gross National Product) and where the goods and
services are produced (within the US regardless of nationality counts
towards Gross Domestic Product).
-------
should be accounted for on a theoretical level. Like
most practical estimates of GNDP, we do not address
these issues in our estimates. For instance, the business
cycle may lead to deceptive trends in GNDP. or a
recession may lead to declining GNDP due to a drop in
consumption that reflects the business cycle rather than
moving away from sustainability.
Whereas some studies have attempted to estimate GS or
GNDP for countries (e.g., Hamilton and Clemens 1999,
Hanley et al. 1999), other studies have focused on a
particular sector of an economy. For example, Matero
and Saastamoinen (2007) examined forest ecosystem
sendees in Finland's forestry accounts and Scroa da
Motta and Ferraz do Amaral (2000) estimated the
depreciation from timber harvest in Brazil. Hrubovcak
et al. (2000) developed a green accounting framework
for US agriculture.
Other indices, such as the Index of Sustainable
Economic Welfare (Daly and Cobb 1989) and the
Genuine Progress Indicator (Cobb et al. 1995), could
be described in more detail in this section, although
Hanley et al. (1999) describe these more as socio-
political measures. The Index of Sustainable Economic
Welfare and the Genuine Progress Indicator focus on
expenditures (e.g., defensive expenditures that mitigate
pollution damage) that reduce sustainability because
they do not increase welfare (Cobb et al. 1995, Daly and
Cobb 1989, Mayer 2008).
2.2.3 Energy
The transformation of available energy potentials
underlies all action in the observable universe and, as a
result, all known processes and systems can be analyzed
using energy-based methods. Although all processes
have an energetic aspect, most energy-based methods
limit their scope of application. There are several
energetic approaches that might be used to determine if
a given system regime is sustainable, but two energy-
based methods are examined in this section; Exergy
Analysis (ExA) and Emergy Analysis (EmA). ExA
was first developed to enhance scientific understanding
of thermal and chemical processes and to identify the
tliermodynamically most-efficient manner in which a
process can be carried out within its environment. EmA
recognizes that Energy Systems Theory (Odum 1994),
conceptually, can be applied to analyze all known
systems, but that boundaries must be arbitrarily set,
based on the particular problem or research question
under consideration. In general, the boundaries set for
an EmA tend to be more broad and more comprehensive
than those that would be set for an ExA of an equivalent
problem (Sciubba and Ulgiati 2005).
2.2.3,1 Exergy Analysis
Preliminary work that eventually gave rise to exergy
analysis was carried out in Europe during the latter
half of the 19th Century. The work of Joule on energy
conservation established the first law of thermodynamics
and that of Clausius, Thomson, Tail, Maxwell, and
primarily Gibbs gave rise to the idea of the maximum
work that can be obtained from an energy potential or
Hie second law of thermodynamics (Szargut. et al. 1988).
Gouy (1889) and Stodola (1898) independently showed
that for work performed under real conditions, the
amount of work actually obtained from a given thermal
energy potential is always smaller than the maximum
possible work, because of the irreversibility of thermal
processes. Because of these different interpretations of
the same word "energy," Rant (1956) proposed the term
"exergy" to describe the ability of energy to perform
work in a transformation (energy quality), especially
under the conditions occurring in technical processes
(Szargut et al. 1988). In an exergy analysis, the quality
of energy is accounted for, not just the quantity as in
an energy analysis (Dewulf et al. 2005). The primary
purpose for exergy analysis has been to give information
about the possibilities for improving thermal processes,
but it cannot say whether such improvements are
practicable. However, exergy and economic analyses
have been combined in an attempt to answer such
questions (El-Sayed and Evans 1970, Evans 1961, Evans
and Tribus 1965).
Exergy can be used as an indicator of the renewability of
industrial processes (Berthiaume et al. 2001), as a metric
of sustainability for technology (Dewulf et al. 2005) and
bioenergy systems (Dewulf et al. 2005). Further, ExA
allows for the assessment of the efficiency of production
chain, which has obvious potential benefits for industry
(Dewulf et al. 2005), and has been made possible via
Cumulative Exergy Consumption, which sums the
exergy values of all resource inflows (Zhang et al. 2010).
For assessing the sustainability of systems, an extended
form of exergy analysis, (e.g., cumulative exergy
analysis; Sciubba and Ulgiati 2005) should be employed.
If ExA is extended to the boundaries of the biosphere,
it effectively approaches an EmA as evidenced by the
analysis of Ukidwe and Bakshi (2007), who defined a
quantity7 "ecologically cumulative exergy," which was
effectively emergy. Exergy does not take into account
ecosystem sendees, except some provisioning sen-ices,
which is an obvious limitation when exergy is advocated
as a "holistic" sustainability metric (Zhang et al. 2010).
2.2.3.2 Emergy Analysis
EmA developed from Energy Systems Theory (EST), a
discipline derived from the union of the fields of ecology,
-------
irreversible thermodynamics, and general systems theory
(Odum 1994). EST identifies and develops general laws
and principles that help understand and interpret natural
phenomena. The transformation of available energy
is the fundamental, underlying causal factor behind all
action in systems of many kinds, including ecological.
social, and economic systems, all of which can be
characterized and analyzed in terms of a network of
energy flows.
EmA builds on the tools of the energy systems approach
(Odum 1971, 1994, 1996) to evaluate the sustainability
of present and possible future human activities. EmA
focuses on the system's use and exchange of emcrgy,
which is the available energy of one kind previously
used up, directly and indirectly, to make a service
or product. The balance of emergy for a system is
expressed in terms of the emergy inputs from renewable
and nonrenewable resources of local or imported origin
observed relative to the economic and social activities
supported by the system. Emergy like exergy is a
concept grounded in the second law of thermodynamics
(i.e., based on the quantity of available energy [exergy]
used up in an irreversible transformation). Both exergy
and emergy must initially satisfy first law constraints
(i.e., the conservation of energy).
Sustainability is not a unitary idea because it can have
different meaning for different individuals or groups.
In addition, sustainability itself is a relative term. In
this context, emergy analyses have been conducted
at various spatial scales; historical studies of nations
have been performed (e.g.. Rydberg and Jansen 2002.
Tilley 2006) and nations have been evaluated over
decades (e.g., Huang et al. 2006). Once the various
spatial and temporal parameters of sustainability have
been decided, different aspects of sustainability can be
examined using the appropriate emergy indices (Brown
and Ulgiati 1997, Lu et al. 2003, Lu and Campbell 2009,
Ulgiati and Brown 1998). Finally, emergy and exergy
methods have been used together to examine energy
conversion processes (Tonon et al. 2006) and emergy has
been combined with, and compared to, other metrics of
sustainability such as ecological footprint (e.g., Giannetti
et al. 2010, Hu et al. 2009, Siche et al. 2010a, 2010b,
Zhao et al. 2005).
2.2.4 System Order
Researchers have studied the impact of global trends on
sustainability recognizing both positive and negative
movement of classes of indicators affecting the transition
to a sustainable future (Kates and Parris 2003). Such
work underscores the importance of evaluating and
managing patterns of change related to growth and
development. Because human and natural systems are
complex and integrated, no action exists in isolation. As
a result, any action starts a chain of events propagating
through multi-dimensional spatial and temporal scales.
Although some systems are brittle and unstable, others
(e.g., ecosystems) are evolutionary and adaptable. If a
system is resilient, it typically will be able to withstand
periodic fluctuations and still maintain function.
However, it is possible for a system to reach a dynamic
threshold and abruptly shift from one set of system
conditions (i.e.. regimes) to another.
Regime shifts have been demonstrated for a multitude
of ecological and social systems, and often have
significant ecological and economic consequences. A
classic example of a regime shift is a shallow lake
where a system may shift from an oligotrophic to an
eutrophic regime due to inflow of phosphorus resulting
in algae overgrowth, depletion of oxygen needed for
survival offish, a loss of biodiversity, and reduced water
quality. Subsequently, the human-desired function
and utility of the system is lost. In the context of
dynamic change and regimes, sustainability is related
to finding and maintaining a set of system conditions
(i.e., a dynamic regime) that can support the social and
economic development of human and environmental
systems without major, irrecoverable environmental
consequences (Karunanithi et al. 2008). Furthermore,
sustainability is predicated on the human preference for
the persistence of a particular regime over another in
order to maintain desired system function and support
current and future generations. (Mayer et al. 2006).
Hence, quantifying and identifying regime shifts is
critically important to sustainability.
There has been a great deal of research on developing
regime shifts indicators (Biggs et al. 2009, Carpenter
2003, Carpenter and Brock 2006. Chisholm and Filotas
2009, Dakos et al. 2008, Scheffer et al. 2009, van
Nes and Scheffer 2007). In real, complex systems,
this task involves tracking multiple system variables
simultaneously overtime. Although variance, skewness,
kurtosis, and critical slowing down are common
indicators of regime shift, research is still needed to
determine whether these indicators will signal shifts
in real, complex systems (Scheffer et al. 2009). One
method of assessing the dynamic changes in complex
systems is information theory.
Information theory generally relates to the quantification
of information in data and has enhanced the ability to
assess organizational complexity even in the presence
of imperfect observations (e.g., noisy data; Mayer
et al. 2006). It has been used to understand various
types of systems (Path et al. 2003) including food
web models (Path and Cabezas 2004) and ecosystem
-------
functioning (Patricio et al. 2004), and is commonly
used for monitoring ecosystem complexity (Anand
and (Mod 2000, Svirezhev 2000) and stability (Path
and Cabezas 2004, Ulanowicz 1997). Key metliods
within information theory include Shannon information
(Shannon and Weaver 1949), Gird-Simpson information
(Colubi 1996), and Fisher information (Cabezas et al.
2005a). Both Shannon and Gini-Simpson information
have been used, for example, as estimates of biological
diversity as they use number of species and species
abundance to calculate system information. One feature
of these measures is that given a different ordering of
the same species probabilities, both indices result in the
same level of information. However, changes in order
are crucial for evaluating the dynamic behavior and
regimes in complex systems (Cabezas et al. 2005a. Path
et al. 2003).
Unlike other measures of system information, Fisher
information (FI) provides a means of monitoring
system variables to characterize a system's dynamic
order including its regimes and regime shifts (Cabezas
et al. 2003). The ability to detect regimes and regime
shifts permits the identification of fundamental changes
occurring in the system and provides insight into what
can be done to abate negative consequences. In practice.
FI has been applied to deriving fundamental equations
of physics, thermodynamics, and population genetics
(Frieden 1998, Frieden 2001). It has been used as a
measure of dynamic order in complex systems (Path ct
al. 2003, Mayer et al. 2006). It has been proposed as a
quantitative index for the detection and assessment of
ecosystem regime shifts (Karunanilhi et al. 2008, Mayer
ct al. 2006) and as a sustainability metric (Cabezas
and Path 2002). Moreover, FI has been used to study
model systems including the stability of a multiple
compartment food web (Cabezas et al. 2005a, Cabezas
ct al. 2005b, Path and Cabezas 2004) and to optimize
control of dynamic model systems for sustainable
environmental management (Shastri etal. 2008a, Shastri
et al. 2008b).
2.3 Sustainability That
Multiple
No single metric should be considered a perfect measure
of sustainability because there are multiple definitions of
sustainability and few, if any, metrics encapsulate all the
definitions; most metrics also are difficult to calculate
(Hanley et al. 1999). Neumayer (2010) finds that no
conclusion can be made about which sustainability
paradigm (i.e., weak vs. strong sustainability) is most
appropriate (but see Dietz and Neumayer 2007).
This would suggest using various ways to measure
suslainabilily in order to cover both paradigms.
Several papers suggest there are three dimensions, or
pillars, of sustainability: economic, environmental.
and social (Adamowicz 2003, Bohringer and Jochem
2007, Hanley et al. 1999, Tanzil and Beloff 2006, UN
2002). It seems clear that different types of indices will
be required to assess the three pillars of sustainability.
To help select metrics Bohringer and Jochem (2007)
summarize five criteria for choosing the right
sustainability metrics: 1) a connection to the definition
of sustainability. 2) indicators from holistic fields. 3)
good data that are available for many years, 4) process-
oriented selection, and 5) the ability to derive political
objectives.
Whereas some studies have focused on particular
metrics, a number have applied multiple metrics in an
attempt to address sustainability. Although some of
these studies discussed major differences in comparing
measures of sustainability, others focused on capturing
multiple aspects of sustainability. For example, Hanley
et al. (1999) used a time series analysis to compare
seven sustainability measures for Scotland. The seven
measures included: (green) Net National Product, GS,
EFA, Environmental Space. Net Primary Productivity,
Index of Sustainable Economic Welfare, and Genuine
Progress Indicator. These measures were used to
examine Scotland from 1980-1993. Overall, they found
Scotland to be unsustainable; however, the individual
measures provided results for different aspects of the
system under study.
Pezzey et al. (2006) compared Green Net National
Product and the interest on augmented GS for Scotland
during 1992-1999. The theory suggests that both the
augmented Green Net National Product and the interest
on augmented GS should be equal. The authors found
that both measures were positive, suggesting that
Scotland was sustainable over the time period, but the
two measures differed in magnitude. They believe
the mismatch was related to assumptions used in their
approach.
In a special issue of the Journal of Environmental
Management, Pulselli et al. (2008) and Tiezzi
and Bastianoni (2008) presented a summary of a
large sustainability project for the Siena Province
using ecological metrics. The authors proposed a
"'sustainability diagnosis" where the diagnosis is
accomplished using a systemic approach, in order to
develop sustainable policies. The methods included
Emergy Evaluation (i.e., EmA), EFA, Greenhouse Gas
Inventory, Life Cycle Assessment, and Extended Exergy
Analysis. The results from these measures were then
analyzed using Principal Component Analysis (PCA) to
examine similarities and differences among the methods.
-------
Bastianoni et al. (2008) discussed the comparisons
of the different metrics using PCA for the territory.
PCA examines the correlation among many metrics
and creates a smaller set of uncorrelated variables, or
principal components (Jolliffe 2002). The authors found
some interesting results when examining correlations
related to particular territories. For example, the largest
urban areas have similarities with population, population
density, total income, CO2, and purchased non-local
resources. In addition, Bastianoni et al. (2008) were
surprised to see a low correlation between Emergy and
EFA. They suggested the low correlation could be
caused by EFA not calculating the extraction of non-
renewable materials explicitly and only focusing on
household consumption. Because Emergy did include
extraction and incorporated industrial consumption, the
result was a low correlation between the two metrics.
Wilson et al. (2007) compared six sustainable
development indices for 132 nations. The metrics
included EFA, Surplus Biocapacity Measure, llie
Environmental Sustainability Index, the Well-being
Index, Human Development Index, and GDP. Using
the Pearson product-moment correlation, they found the
environmental Sustainability index and well-being index
were positively correlated, as were the well-being index,
the human development index, and GDP. Negative
correlations were identified among EFA. the Well-being-
Index, the Human Development Index and GDP.
Eight metrics were estimated for 1990-2000 to examine
the Sustainability pattern of France (Nourry 2008).
The eight metrics were (green) Net National Product,
GS, EFA, Indicator of Sustainable Economic Welfare.
Genuine Progress Indicator. Pollution-sensitive
Human Development Indicator, Sustainable Human
Development Indicator, and French Dashboard on
Sustainable Development. Nourry (2008: 10) defines
this last indicator as a "non-monetary measure composed
of non-aggregated indicators."7 Some of the 15
indicators used were life expectancy without disability,
overfishing, GS, waste production and population,
and public debt (Nourry 2008). The results suggest
that France satisfies weak Sustainability but not strong
Sustainability.
Lee and Huang (2007) examined 51 indicators of
Sustainability' for Taipei from 1994-2004. Those
indicators were categorized into four dimensions and
reduced to estimate a composite Sustainability index.
For example, the environmental dimension included
indicators such as per capita CO, emissions, reservoir
'There is not a consensus on the use of the terms indicator and metric.
Nourry's (2008) use of indicator is equivalent to our use of metric in
this report.
water quality-, and tap water quality. Social indicators
included urban population density, the wealth gap, and
crime rate among others. The economic dimension
included average personal income, female/male
unemployment rate, and percentage of households
with internet connection. The institutional dimension
incorporated enforcement of local environmental plans,
joint international cooperation related to sustainable
development, and the ratio of the environmental and
ecological budget to total budget in its calculation.
The four different dimensions were aggregated into
one measure (i.e., metric) using equal weighting and
suggested that Taipei was moving towards Sustainability.
However, only the environmental and social dimensions
indicated this movement. The economic dimension
only started increasing after 2002 and the institutional
dimension was inconsistent.
Gray more et al. (2008) focused on South East
Queensland as a regional case study. The metrics
examined included EFA, Well-being Assessment,
Ecosystem Health Assessment, Quality of Life, and
Natural Resource Availability. The authors determined
that calculating these metrics at the regional scale was
not informative for a few reasons. For example, data
limitations at the regional scale caused problems with
calculating the metrics, especially when attempting
to examine trends. In addition, the metrics examined
were not always easy to understand or did not reveal
information about Sustainability. Graymorc ct al. (2008)
conclude that a new approach is necessary given the
limitations of their approach for calculating regional
Sustainability metrics.
Similar to Hanley et al. (1999), the special issue of the
Journal of Environmental Management (2008), and
Nourry7 (2008), we propose an approach that includes
multiple metrics in order to measure the Sustainability
of a region. Rather than comparing the metrics for
different areas (e.g., Wilson et al. 2007), we think the
Sustainability' measures we chose will capture various
attributes of the SLB region, providing a holistic view
similar to that of the Pulselli et al. (2008) "Sustainability
diagnosis."
2.4 Choice of for Luis
Project
This project utilized available environmental, economic.
and social data to calculate four indices of Sustainability.
The metrics or indices included: 1) environmental
footprint as characterized by EFA, 2) economic well-
being as ascertained from Green Net Regional Product
(GNRP; similar to GNDP but defined on a smaller spatial
scale). 3) energy flow through the system as computed
-------
from ail EmA, and 4) dynamic order estimated from
FI. We chose several metrics for a few reasons. First,
we suspect that no single metric will provide sufficient
information for planners to assess adequately different
aspects of the system. We also think that multiple
metrics are necessary because no metric is perfect in
capturing a specific aspect of the system. In addition,
these four metrics cover both paradigms of weak and
strong sustainability (as described above). GNRP is a
measure of weak sustainability (e.g.. Neumayer 2010.
Pezzey et al. 2006) and we believe FI and the associated
Sustainable Regimes Hypothesis are measures of weak
sustainability. Although the approach is not directly
indicative of resource substitutability and focuses on
the functionality of a system, it is possible for resources
(e.g., species) to be replaced as long as the system does
not become disorderly (i.e., it continues to function).
EFA is based on the concept of strong sustainability
(e.g., Ferguson 2002, Mayer 2008, Neumayer 2010, Rees
2002, but see Dietz and Neumayer 2007, McManus and
Haughton 2006) and so is EmA (e.g., Giannetti et al.
2010).
We note these multiple metrics do use similar
variables, and therefore, (here is some redundancy in
the metrics. This redundancy in variables could result
in similar behavior in the metrics over time, under
some circumstances, depending on Hie influence of
the particular variables. Table 2.1 summarizes the
different metrics and the data needed for estimation. It
presents the spatial scale of the data and the years the
data were available. EmA is Hie most data intensive
with FI following. However, EmA has specific data
requirements (e.g., energy inflows and outflows),
whereas FI has no requirements other than data that
adequately represent the system behavior. Of the broad
categories of variables presented, EFA and GNRP are Hie
least data intensive. The only variable that all metrics
include is population, but each uses it in different ways.
For example. GNRP uses it to estimate CO, emissions
for the SLB based on per capita estimates for Colorado
and New Mexico. EFA calculates the land required for
each land category to support both an individual and
entire population of Hie region. It may be one of the
easiest of the suslainabilily metrics for decision makers
to understand and conceptualize and may be one of
the most commonly computed. Although most of the
variables are required by at least two of the metrics, each
metric plays a unique and important role in measuring
sustainability for the SLB as we describe below.
This suite of metrics will give public land managers,
the local community, and the U SEPA a methodology
for monitoring the effect of land management and
decision-making on environmental quality and economic
sustainability. We believe the sustainability metrics
selected will assist with the review of public land
development and management decisions and promoting
environmental quality.
2.4.1 Components to Environmental
Systems
The four metrics we selected capture what we
determined to be fundamental aspects of an
environmental system and its sustainability. We built
our methodology based on these primary aspects and
selected metrics in an attempt to capture or quantify
those aspects. At a basic level, we recognized that an
environmental system has some inherent order, whether
the factors responsible for maintaining that order are
clearly identified or understood. Furthermore, it takes
energy to maintain order in an environmental system.
Finally, sustainability is inherently an anthropocentric
issue and how well off humans are may dictate their
influences on the system. An individual or population
that is meeting its needs for existence has the luxury
of being concerned about the effect it has on its
environment whereas a population that is barely
surviving may not be intensely concerned with long-term
sustainability.
We selected metrics that capture each of these
fundamental aspects. FI was selected to measure
dynamic order of the overall system. We chose EmA
because it measures the quality-normalized flow of
energy through the system. GNRP was chosen to capture
the overall economic well-being of the population and
EFA was selected to capture the impact a population has
on environmental resources resulting from consumption
and waste production. Out of the four metrics, using
FI to characterize sustainability is a relatively new
metric (Mayer 2008). Application of the other metrics
is relatively common and they are established as metrics
of sustainability in the scientific literature (e.g., Mayer
2008).
Although the metrics were selected because they
measure what we identified as primary or fundamental
aspects of an environmental system, they assess
some component of the economic, social, and
environmental pillars of sustainability and these
three pillars, traditionally, are considered to support
sustainable development (Bobylev 2009, UN 2002).
Measurements relevant to each pillar of sustainability
will enable decision makers to assess where the regional
environmental system is in need of management actions
and help them make choices that can directly guide the
path of the regional system toward sustainability.
-------
Finally, although a certain suite of metrics was chosen
for this study, it does not mean that we view these as
providing the final word on system sustainability. The
issue of sustainability is complex and we expect that
other metrics exist, or may be developed, that will
provide improvements in the methods of calculating
indices in the future. In the next four chapters, we
present the methodology and results of EFA, GNRP,
EmA, and FI, respectively.
2.5
Adamowicz, W. 2003. Economic indicators of
sustainable forest management: theory versus practice.
Journal of Forest Economics 9, 27-40.
Ahlroth, S. 2003. Correcting NDP for SO2 and NOx
emissions: implementation of a theoretical model in
practice. Review of Income and Wealth 49. 425-440.
Anand, M., (Mod, L. 2000. On partitioning of an
ecological complexity function. Ecological Modelling
132,51-61.
Asheim, G., Weitzman, M. 2001. Does NNP growth
indicate welfare improvement? Economics Letters 73.
233-239.
Azqueta, D., Sotelsek, D. 2007. Valuing nature: from
environmental impacts to natural capital. Ecological
Economics 63, 22-30.
Bastianoni, S., Pulselli, P., Focardi, S., Tiezzi,
E., Gramalica, P. 2008. Correlations and
complementarities in data and methods through
principal components analysis (PCA) applied to
the results of the SPIn-Eco project. Journal of
Environmental Management 86, 419-426.
Berthiaume, R., Bouchard. C., Rosea M.A. 2001.
Exergetic evaluation of the renewabiliry of a biofuel.
Exergy 1, 256-268.
Biggs, R., Carpenter, S.R., Brock, W.A. 2009. Turning
back from the brink: detecting an impending regime
shift in time to avert it, PNAS 106, 826-831.
Bobylev, N. 2009. Mainstreaming sustainable
development into a city's master plan: a case of urban
underground space use. Land Use Policy 26, 1128-
1137.
Bohringcr, C., Jochem, P. 2007. Measuring the
immeasurable—A survey of sustainability indices.
Ecological Economics 63, 1-8.
Brown, M.T., Ulgiati, S. 1997. Emergy based Indices
and ratios to evaluate sustainability: monitoring
technology and economies toward environmentally
sound nnovation. Ecological Engineering 9, 51-69.
Cabezas, H., Path, B.D. 2002. Towards a theory of
sustainable systems. Fluid Phase Equilibria 3-14,
194-197.
Cabezas. H., Pawlowski, C.W., Mayer, A.L., Hoagland,
N.T. 2003. Suslainability: ecological, economic,
technological and systems perspectives. Clean
Technologies and Environmental Policy 5, 176-180.
Cabezas, H., Pawlowski, C.W., Mayer, A.L., Hoagland,
N.T. 2005a. Sustainable systems theory: ecological
and other aspects. Journal of Cleaner Production 13.
455-467.
Cabe/as, H., Pawlowski, C.W., Mayer, A.L., Hoagland,
N.T. 2005b. Simulated experiments with complex
sustainable systems: ecology and technology.
Resources, Conservation and Recycling 44, 279-291.
Cairns, R. 2002. Green accounting using imperfect
current prices. Environment and Development
Economics 7, 207-214.
Carpenter, S.R., Walker, B.H., Anderies J.M., Abel, N.
2001. From metaphor to measurement: resilience of
what to what? Ecosystems 4, 765-781.
Carpenter, S. R. 2003. Regime shifts in lake
ecosystems: pattern and variation, (Excellence in
Ecology Series. Ecology Institute, Oldendorf/Luhe.
Germany), vol. 15.
Carpenter, S.R., Brock, W. A. 2006. Rising variance:
a leading indicator of ecological transition. Ecology
Letters 9, 311-318.
Caughley, G. 1976. Wildlife management and the
dynamics of ungulate populations. Applied Biology 1,
183-246.
Chisholm, R.A., Filotas, E. 2009. Critical slowing
down as an indicator of transitions in two-species
models. Journal of Theoretical Biology 257, 142-149.
Cobb, C., Halstead, T, Rowe, I 1995. The
Genuine Progress Indicator: Summary of Data and
Methodology. Redefining Progress, San Francisco,
CA.
Colubi A. 1996. Comparative studies of two diversity
indices based on entropy measures: Gini-Simpson's
vs Shannon's. In: dc Andalicia J (Ed), Information
processing and management of uncertainty in
knowledge-based systems. International Conference
on Information Processing and Management of
Uncertainty in Knowledge-based Systems 1996.
IPMU'96. 2. 675-80.
-------
Dakos, V., Scheffer, M., vaiiNes, E. H., Brovkin, V.,
Petoukhov , V., Held, H. 2008. Slowing down as
an early warning signal for abrupt climate change.
Proceedings of National Academy of Sciences 105,
14308-14312.
Daly, H., Cobb, C. (Eds.). 1989. For die common good.
Beacon Press, Boston, MA.
Dasgupta, P., Maler, K. 2000. Net national product,
wealth, and social well-being. Environment and
Development Economics 5, 69-93.
Dasgupta. P. 2001. Human Well-Being and die Natural
Environment. Oxford University Press, New York.
Dewulf, I, Van Langenhove, H. Van de Velde. B. 2005.
Exergy-based efficiency and renewability assessment
of biofuel production. Environmental Science and
Technology 39, 3878-3882.
Dietz, S., Neumayer, E. 2004. Genuine savings:
a critical analysis of its policy-guiding value.
International Journal of Environment and Sustainable
Development 3, 276-292.
Dietz, S., Neumayer, E. 2007. Weak and strong
sustainability in the SEEA: Concepts and
measurement. Ecological Economics 61, 617-626.
Dietz, T., Rosa, E.A. 1994. Rethinking the
environmental impacts of population, affluence and
technology. Human Ecology Review 1,277-300.
Dietz, T.. Rosa. E.A., York, R. 2007. Driving the human
ecological footprint. Frontiers in Ecology and the
Environment 5, 13-18.
Ehrlich, PR., Holdrcn, J.P. 1971. Impact of Population
Growth. Science 171, 1212-1217.
El-Sayed, Y.M., Evans, R.B. 1970. Thermoeconomics
and die design of heat systems. ASME Journal of
Engineering Power 102, 29-36.
Evans. R.B. 1961. A contribution to die theory of
diermoeconomics. M.S. Thesis, University of
California, Los Angeles, CA.
Evans, R.B. and Tribus, M. 1965. Thermoeconomic
design under conditions of variable price structure, in:
Proceedings of 1st International Symposium on Water
Desalination, Washington, October. SWD-89.
Fan, Y, Liu, L.-C., Wu, G., Wie, Y.-M. 2006. Analyzing-
impact factors of CO2 emissions using the STIRPAT
model. Environmental Impact Assessment 26, 377-
395.
Path, B., Cabezas, H., Pawlowski, C.W. 2003. Regime
changes in ecological systems: an information theory
approach. Journal of Theoretical Biology 222-4, 517-
530.
Path, B.D., Cabezas, H. 2004. Exergy and Fisher
information as ecological indices. Ecological
Modeling 174, 25-35.
Frieden, B.R. 1998. Physics from Fisher Information: A
Unification, Cambridge University Press, Cambridge,
UK.
Frieden, B.R. 2001. Probability. Statistical Optics and
Data Testing, third ed. Springer-Verlag, New York.
Fry, F.E. J. 1947. Statistics of a lake trout fishery.
Biometrics 5, 27-67.
Gasparalos, A., El-Haram, M., Homer, M. 2007. A
critical review of reductionist approaches for assessing
the progress towards sustainability. Environmental
Impact Assessment Review doi: 10.1016/j.
eiar.2007.09.002.
Giannetti, B., Almeida. C., Bonilla, S. 2010. Comparing
emergy accounting with well-known sustainability
metrics: The case of Southern Cone Common Market,
Mercosur. Energy Policy 38, 3518-3526.
Gouy, L.G. 1889. On utilizable energy. Journal de
Physiques, 501.
Graymore, M., Sipe, N., Rickson, R. 2008. Regional
sustainability: how useful are current tools of
sustainability assessment at the regional scale?
Ecological Economics 67, 362-372.
Grimm, V., Wisscl, C. 1997. Babel, or the ecological
stability discussions: an inventory and analysis of
terminology and a guide for avoiding confusion.
Oecologia 109, 323-334.
Hak, T, Moldaa B.. Dahl, A. (Eds.). 2007.
Sustainability7 Indicators: A Scientific Assessment.
SCOPE Vol. 67. Island Press, Washington, DC.
Hamilton, K., Atkinson, G., Pcarce, D. 1997. Genuine
savings as an indicator of sustainability. Centre
for Social and Economic Research on the Global
Environment (CSERGE) Working Paper GEC 97-03.
ISSN 0967-8875.
Hamilton, K., Clemens, M. 1999. Genuine savings
rates in developing countries. World Bank Economic
Review 13, 333-356.
-------
Hanley, N., Moffatt, I., Faichney, R., Wilson, M. 1999.
Measuring sustainability: a time series of alternative
indicators for Scotland. Ecological Economics 28,
55-73.
Hanley, N., McGregor, P., Turner, K., Wilson, M.
2002. Greening Hie National Accounts for Scotland.
Scottish Economic Policy Network (scotecon). ISBN
1-904365-03-5. University of Stirling, Stirling.
Hartwick, J. 1990. Natural resources, national
accounting and economic depreciation. Journal of
Public Economics 43, 291-304.
Hau J.L., BakshiB.R. 2004. Expanding exergy analysis
to account for ecosystem products and services.
Environmental Science & Technology 38, 3768-77.
Heal, G., Kristom, B. 2005. National income
and the environment, in: Maler, K., Vincent, J.
(Eds.), Handbook of Environmental Economics,
Economywide and International Environmental Issues,
Vol. 3. Elsevier, North-Holland, Amsterdam.
Holling, C.S. 1973. Resilience and stability of
ecological systems. Annual Review of Ecology and
Systematics 4, 1-23.
Hrubovcak, J., LeBlanc, M., Eakin, B. 2000.
Agriculture, natural resources, and environmental
accounting. Environmental and Resource Economics
17, 145-162.
Hu, D., Wang, R.-S., Lei, K.-R, Li, E, Wang, Z., Wang,
B.-N. 2009. Expanding ecological appropriation
approach: solar space method and a case study in
Yang/lion city, East China. Ecological Complexity 6,
473-483.
Huang, S.L., Lee, C.L., Chen, C.W. 2006.
Socioeconomic metabolism in Taiwan: emergy
synthesis versus material flow analysis. Resources,
Conservation and Recycling 48, 166-196.
Jackson, J.B.C., Kirby, M.X., Berger, W.H., Bjorndal,
K.A., Botsford, L.W., Bourque, B.J., Bradbury, R.H.,
Cooke, R., Erlandson, J., Estcs, J.A., Hughes, T.P.,
Kidwell, S., Lange, C.B., Lenihan, H.S., Pandolfi,
J.M., Peterson, C.H., Steneck, R.S., Tegner, M.J.,
Warner, R.R. 2001. Historical overfishing and the
recent collapse of coastal ecosystems. Science 293,
629.
Jolliffe, I. 2002. Principal component analysis.
Springer-Verlag, New York.
Karunanithi, A.T., Cabezas, H., Frieden, B.R.,
Pawlowski, C.W. 2008. Detection and Assessment
of ecosystem regime shifts from Fisher Information.
Ecology & Society 13, 22.
Kates, R.W., Parris, T. M. 2003. Long term trends and
a sustainability transition. Proceedings of National
Academy of Sciences 100, 8062-8067.
Larkin, P.A. 1977. An epitaph for the concept of
maximum sustainable yield. Transactions of the
American Fisheries Society 106, 1-11.
Lee, Y, Huang, C. 2007. Sustainability index for
Taipei. Environmental Impact Assessment Review 27,
505-521.
Lenzen, M., Murray, S.A. 2001. A modified ecological
footprint method and its application to Australia.
Ecological Economics 37, 229-255.
Lu, H., Campbell, D.E. 2009. Ecological and economic
dynamics of the Shunde agricultural system under
China's small city development strategy. Journal of
Environmental Management 90, 2589-2600.
Lu, H.F., Lan, S.F., Li, L., Peng, S.L. 2003. New
emergy indices for sustainable development. Journal
of Environmental Sciences, 15, 562-569.
Luck, M.A., Jenerette, G.D., Wu, J., Grimm, N.B.
2001. The urban funnel model and the spatially
heterogeneous ecological footprint. Ecosystems 4,
782-796.
Matero, J., Saastamoinen, O. 2007. In search of
marginal environmental valuations—ecosystem
services in Finnish forest accounting. Ecological
Economics 61, 101-114.
Mayer, A. 2008. Strengths and weaknesses of common
sustainability indices for multidimensional systems.
Environment International 34, 277-291
Mayer, A.L., Pawlowski, C.W., Cabezas, H. 2006.
Fisher Information and dynamic regime changes in
ecological systems. Ecological Modeling 195, 72-82.
Mayer, A., Thurston, H., Pawlowski, C. 2004. The
multidisciplinary influence of common sustainability
indices. Frontiers in Ecology and Environment 2, 419-
426.
Medved, S. 2006. Present and future ecological
footprint of Slovenia—the influence of energy demand
scenarios. Ecological Modelling 19, 25-36.
-------
Moirfreda, C., Wackernagel, M., Deuniling, D. 2004.
Establishing national natural capital accounts based on
detailed Ecological Footprint and biological capacity
assessments. Land Use Policy 21, 231-246.
Ness, B., Urbel-Piirsalu, E., Anderberg, S., Olsson, L.
2007. Categorising tools for sustainability assessment.
Ecological Economics 60. 498-508.
Neumayer, E. 2004. Indicators of sustainability.
In: Tietenberg. T., Folmer, H. (Eds.), International
Yearbook of Environmental and Resource Economics
2004/05. Edward Elgar, Cheltenham, UK.
Neumayer, E. 2010. Weak versus Strong Sustainability:
Exploring the Limits of Two Opposing Paradigms.
Edward Elgar, Northampton, MA.
Nordhaus, W., Kokkelenberg, E. 1999. Nature's
numbers: expanding the national economic accounts
to include the environment. Report of the Panel on
Integrated Environmental and Economic Accounting.
National Academy Press: Washington, DC.
Nourry, M. 2008. Measuring sustainable development:
some empirical evidence for France from eight
alternative indicators. Ecological Economics 67, 441-
456. doi: 10.1016/j.ecolecon.2007.12.019.
Odum, H.T. 1971. Environment. Power, and Society.
Wiley, New York.
Odum, H.T. 1994. Ecological and General Systems:
An Introduction to Systems Ecology, Second Edition.
University Press of Colorado, Niwot, Colorado.
Odum, H.T. 1996. Environmental Accounting: Emergy
and Environmental Decision Making. John Wiley &
Sons, New York.
Orr, D.W. 2002. Four challenges of sustainability.
Conservation Biology 16, 1457-60.
Patrick) I, Ulanowicz, R.. Pardal, M. A.. Marques, J.C.
2004. Ascendency as an ecological indicator: a case
study of estuarine pulse eutrophication. Estuarine,
Coastal and Shelf Science 60. 23-35.
Pearce, D. 2002. An intellectual history of
environmental economics. Annual Review of Energy
and the Environment 27, 57-81.
Pearce. D., Atkinson. G. 1995. Measuring sustainable
development, in: Bromley, D. (Ed.), The Handbook
of Environmental Economics. Blackwell Publishers,
Cambridge, MA.
Pearce, D., G. Atkinson. 1993. Capital theory and the
measurement of sustainable development: an indicator
of "weak" sustainability. Ecological Economics 8,
103-108.
Pezzey, J. 2004. One-sided sustainability tests with
amenities, and changes in technology, trade, and
population. Journal of Environmental Economics and
Management 48, 613-631.
Pezzey, J., Hanley, N., Turner, K., Tinch, D. 2006.
Comparing augmented sustainability measures for
Scotland: is there a mismatch? Ecological Economics
57, 60-74.
Pillarisetti, J. 2005. The World Bank's'genuine
savings' measure and sustainability. Ecological
Economics 55, 599-609.
Pitcher, T.I, Hart, P.IB. 1982. Fisheries Ecology.
Chapman & Hall, London.
Pulselli, E, Ciampalini. E, Leipert, C., Tiezzi, E.
2008. Integrating methods for the environmental
sustainability: the SPIn-Eco project in the province of
Siena (Italy). Journal of Environmental Management
86, 332-341.
Rant, Z. 1956. "Exergie, ein neues Wort fur
"Technische Arbcitsfahigkeit" (Excrgy, a new word
for "technical available work")". Forschung auf dem
Gebiete des fngenieurwesens 22, 36-37.
Rees, W.E. 1992. Ecological footprints and
appropriated carrying capacity: what urban economics
leaves out. Environment and Urbanization 4, 120-130.
Rico-Ramirez, VR., Qunitana-Hernandez, P.A., Ortiz-
Cruz, J.A., Hernandez-Castro, S. 2008. Fisher
information: a generalized sustainability index.
Computer-Aided Chemical Engineering 25, 1155-
1160.
Ricker, W.E. 1958. Handbook of computations for
biological statistics of fish populations. Bulletin 119,
Fisheries Research Board, Canada, Ottawa. 300 pp.
Russell, F.S. 1931. Some theoretical considerations
on the "overfishing" problem. Journal de Conseil
International pour 1'Explaration de la mer 6, 3-20.
Rydberg, T., Jansen, J. 2002. Comparison of horse and
tractor traction using emergy analysis. Ecological
Engineering 19, 13-28.
Schaefer, MB. 1954. Some aspects of the dynamics
of populations important to the management of the
commercial marine fisheries. Bulletin Inter-American
Tropicl Tuna Commission 1(2), 27-56.
Scheffer, M., Bascompte, J., Brock, W. A., Brovkin,
V.. Carpenter. S. R., Dakos, V., Held, H., van Nes, E.
H., Rietkerk, M., Sugihara, G. 2009. Early-warning
signals for critical transitions. Nature 461, 53-59.
-------
Schulze, P.C., 2002. I=PAT. Ecological economics 40,
149-150.
Sciubba, E., Ulgiati. S. 2005. Emergy and exergy
analyses: complementary methods of irreducible
ideological options? Energy 30, 1953-1988.
Seroa da Motta, R., Ferraz do Aniaral, C. 2000.
Estimating timber depreciation in the Brazilian
Amazon. Environment and Development Economics
5, 129-142.
Shannon. C., Weaver W. 1949. The mathematical
theory of communication. University of Illinois Press.
Champaign, IL.
Sharkey, M.J. 1970. The carrying capacity of natural
and improved land in different climatic zones.
Mammalia 34, 564-572.
Shastri, Y, Diwckar, U., Cabezas, H. 2008a. Optimal
Control Theory for sustainable environment
management. Environmental Science and Technology
42, 5322-5328.
Shastri. Y, Diwekar, U., Cabezas, H., Williamson.
I 2008b. Is sustainability achievable? Exploring
the limits of sustainability with model systems.
Environmental Science and Technology 42, 6710-
6716.
Siche, R., Agostinho, K, Ortega, E. 2010a. Emergy Net
Primary Production (ENPP) as a basis for calculation
of ecological footprint. Ecological Indicators 10,475-
483.
Siche, R., Pereira, L., Agostinlio, E, Ortega, E. 2010b.
Convergence of ecological footprint and emergy
analysis as a sustainability indicator of contries: Peru
as case study. Communications in Nonlinear Science
and Numerical Simulation 15, 3182-3192.
Singh, R., Murty, H., Gupta, S., Dikshit, A. 2009. An
overview of sustainability assessment methodologies.
Ecological Indicators 9, 189-212.
Stodola, A. 1898. The cyclic processes of the gas
engine [in German]: Zeitschrift des Vereines Deutscher
Ingenieure 32, 1086-1091.
Sutton, P. 2003. An empirical environmental
sustainability index derived solely from nighttime
satellite imagery and ecosystem service valuation.
Population and Environment 24, 293-311.
Svirezhev. YM. 2000. Thermodynamics and ecology.
Ecological Modelling 132, 11-22.
Szargut, I, Morris, D.R., Steward, F.R. 1988. Exergy
Analysis of Thermal, Chemical, and Metallurgical
Processes. Hemisphere Publishing Corporation. New
York.
Tanzil, D., Beloff, B. 2006. Assessing Impacts:
Overview on sustainability indicators and metrics.
Environmental Quality Management 15. 41-56.
Tietenberg, T. 2000. Environmental and Natural
Resource Economics, Fifth Edition. Addison-Wesley
Longman. Inc., New York.
Tiezzi, E., Bastianoni, S. 2008. Sustainability of the
Siena Province through ecodynamic indicators.
Journal of Environmental Management 86, 329-331.
Tilley, D.R. 2006. National metabolism and
communication technology development in the United
States: 1790 to 2000. Environment and History 12,
165-190.
Tonon, S., Brown. M.T., Luchi, F. Mirandola, A.,
Stoppato, A., Ulgiati, S. 2006. An integrated
assessment of energy conversion processes by means
of thermodynamic, economic, and environmental
parameters. Energy 31. 149-163.
Ukidwe, N.U., Bakshi,B.R. 2007. Industrial and
ecological cumulative exergy consumption of the
United States via the 1997 input-output benchmark
model. Energy 32, 1560-1592.
Ulanowicz, R.E. 1997. Ecology, the Ascendent
Perspective. Columbia University Press, New York.
Ulgiati, S., Brown, M.T. 1998. Monitoring patterns of
sustainability in natural and man-made ecosystems.
Ecological Modelling 108. 23-36.
UN (United Nations). 2002. Report of the World
Summit on Sustainable Development. Johannesburg,
South Africa, 26 August-4 September 2002. A/
CONF. 199/20. ISBN 92-1-104521-5. United Nations,
New York.
UN (United Nations). 2007. Indicators of Sustainable
Development: Guidelines and Methodologies. Third
Edition. Division for Sustainable Development,
Department of Economic and Social Affairs. United
Nations, New York.
Usher, D. 1981. The Measurement of Economic
Growth. Blackwell, Oxford.
vandenBerghJ.C.J.M. 2009. The GDP paradox.
Journal of Economic Psychology 30, 117-135.
-------
van Nes, E.H., Scheffer, M. 2007. Slow recovery
from perturbations as a generic indicator of a nearby
catastrophic shift. American Naturalist 169, 738-747.
Venetoulis, J, Talberth, J. 2005. Ecological footprint of
nations: 2005 update. Redefining Progress, Oakland,
California.
Venetoulis, J, Talberth, J. 2008. Refining the
ecological footprint. Environment, Development and
Sustainability 10, 441-469.
Vincent, J. 2000. Green accounting: from theory to
practice. Environment and Development Economics
5, 13-24.
Wackernagel, M., Rees, W. 1996. Our ecological
footprint: reducing human impact on the earth. New
Society Publishers, Gabriola Island, BC.
Wackernagel, M. 2009. Introduction: methodological
advancement in footprint analysis. Ecological
Economics 68, 1925-1927.
Waggoner, P.E., Ausubel. J.H. 2002. A framework
for sustainability science: a renovated 1PAT identity.
Proceedings of the National Academy of Science, 99,
7860-7865.
Wcitzman. M. 1976. On the welfare significance of
national product in a dynamic economy. Quarterly
Journal of Economics 90, 156-162.
Weitzman. M. 2000. The linearised Hamiltonian as
comprehensive NDP. Environment and Development
Economics 5, 55-68.
Wilson, J., Tyediners, P., Pelot, R. 2007. Contrasting
and comparing sustainable development indicator
metrics. Ecological Indicators 7, 299-314.
York R., Rosa, E.A., Dietz, T. 2002. Bridging
environmental scidne with environmental policy:
plasticity of population, affluence, and technology.
Social Science Quarterly 83, 18-34.
York R., Rosa, E.A., Diet/, T. 2003. STIRPAT, IPAT
and ImPACT: analytic tools for unpacking the
driving forces of environmental impacts. Ecological
Economics 46, 351-365.
Zhang, Y., Singh, S., Bakshi, B.R. 2010. Accounting for
ecosystem services in life cycle assessment, Part I: a
critical review. Environmental Science & Technology
44, 2232-2242.
Zhao, S., Li, Z., Li, W. 2005. A modified method of
ecological footprint calculation and its application.
Ecological Modelling 185, 65-75.
-------
Table 2.1- Broad categorical list of data collected on key variables, showing which metric used the variable, the
source of the data, the scale at which the data were collected, and the available years of data reported.
Variable
Population
Personal income
Land area
Precipitation
Solar and Wind
Food consumption
Food production
Imports
Exports
Forest harvest
Energy consumption
CO, emissions
Water balance
Wind erosion
Metric'
Em A
EmA
EmA
EmA
EmA
EmA
EmA
EmA
EmA
EmA
EmA
EmA
EFA
EFA
EFA
EFA
EFA
GNRP
GNRP
GNRP
GNRP
GNRP
F!
FI
FI
FI
FI
FI
FI
FI
FI
FI
FI
FI
Source2
BEA
BEA
NASS
PRISM Group
NASS
USDA-ARS
NASS
GI
GI
USDA-FS
EIA
EIA
CDSS
Multiple sources
Scale'
C, S
c, s
C
c
c
N
C
c
c
c
s
s
R
C
Years
1980-2005
1980-2005
1980-2005
1980-2005
1980-2005
1980-2005
1980-2005
1995-2005
1995-2005
1980-2005
1980-2005
1980-2005
1980-2005
1980-2005
'EmA — Emergy Analysis; EFA — Ecological Footprint Analysis; GNRP — Green Net Regional Product; FI — Fisher information. 2 BEA — Bureau of
Economic Analysis; NASS — National Agricultural Statistics Service; PRISM Group — Parameter-elevation Regressions on Independent Slopes
Model Group; USDA-ARS — United States Department of Agriculture, Agricultural Research Station; USDA-FS — United States Department of
Agriculture, Forest Service; CDSS — Colorado's Decision Support Systems; GI — Global Insight, Inc.; EIA — Energy Information Administration.
3C — county; R — region; S — state; N — national.
-------
-------
3.0
Ecological Footprint
3.1 Introduction
In this chapter, we introduce data sources and the
methodology employed to perform an Ecological
Footprint Analysis (EFA) for the San Luis Basin (SLB).
We present and discuss the results, interpret what they
mean, and discuss methodological issues in an attempt to
make the approach more useful to decision makers.
EFA. more appropriately called environmental footprint
(Eaton et al. 2007) analysis, is a commonly used
metric of sustainability because it is straightforward
in theory and easy to conceptualize. EFA attempts to
capture human impacts on the regenerative capacity
of an environmental system (Chambers et al. 2000) by
identifying the amount of bioproductive land required
to support a person's average annual consumption
and waste production (Wackernagel and Rees 1996).
Since Wackernagel and Rees (1996) introduced it,
EFA methodology has been continually evolving as
researchers have attempted to improve the methodology
(e.g., Lenzen and Murray 2001, Luck et al. 2001,
Monfreda et al. 2004, Medved 2006, Wackernagel 2009).
In general, EFA categorizes bioproductive land into one
of six types: energy, arable, forest, pasture, built, and
sea. Land that does not meet one of these six categories
is deemed non-productive and is not included in EFA
accounting (Chambers et al. 2000). Furthermore, in
the accounting twelve percent of the land is set aside
for biodiversity (Chambers et al. 2000, but see Noss
and Cooperrider 1994 for discussion of the appropriate
percentage). EFA calculates the supply (biocapacity) of
land available to support a population and the demand
(ecological footprint) a population places on these six
land categories by estimating how much land is being
used to support the population. The biocapacity in a
given region is compared to the ecological footprint
of the population by subtracting demand from supply.
Examining this balance sheet over time will reveal
if there is an ecological reserve or deficit and if a
population is moving toward or away from sustainability
as defined by the EFA method.
This metric has been applied to numerous systems. For
instance, EFA accounting has been conducted at global
(e.g., WWF 2008), national (e.g., Erb 2004, Lammers
et al. 2008, Lenzen and Murray 2001, Medved 2006,
Wackernagel and Yount 1998, Wackernagel et al. 1999,
Wackernagel ct al. 2004a, WWF 2008), regional (e.g.,
Gray more et al. 2008, Huang et al. 2007, Pulselli et al.
2008, Zhao et al. 2008). and municipal (e.g., Collins
et al. 2006, Scotti et al. 2009, Wood and Garnett 2009)
spatial scales. EFA has been conducted for industry
or product level (e.g., wine - Niccolucci et al. 2008,
university campus - Li et al. 2008. tourism - Patterson
et al. 2007), and individual or household levels (e.g.,
Chambers et al. 2000) as well. Some researchers have
conducted EFA analysis by comparing and contrasting
multiple spatial scales (e.g.. Lenzen and Murray 2003,
Luck etal. 2001).
We conducted conventional EFA accounting using the
compound approach as introduced by Wackernagel and
Rees (1996) and expanded by Chambers et al. (2000)
because it is more inclusive and robust compared to the
component-based approach (Chambers et al. 2000). In
general, an EFA calculation requires four components
or sets of variables: 1) the size of the population in the
region of interest, 2) the amount of consumables in
various categories used per individual, 3) the amount
of energy of various types consumed per individual,
and 4) the amount of biologically productive land, as
defined by EFA, available in the region of interest. The
consumption component includes the biotic resources
such as meat, dairy, fruits, and vegetables. Consumption
is usually calculated by adding the quantity of imported
consumables to the amount produced in the system
under study and subtracting the amount exported from
the system. The remaining quantity is assumed to be
consumed by the population (Chambers et al. 2000,
Wackernagel and Rees 1996). The energy balance
component considers both locally generated energy and,
if known, energy embodied in traded goods (Chambers
et al. 2000, Wackernagel and Rees 1996). Primary fuel
consumption is adjusted for carbon content and used to
derive the amount of forested land necessary to sequester
the resulting CO2 emissions (Chambers et al. 2000).
Lastly, the area (in hectares) of each of the six land
categories of biologically productive land is measured to
calculate the biocapacity. Each of these components is
used in one of three accounting type ledgers to determine
the supply (biocapacity) and demand (ecological
footprint) of the environmental system (Chambers et al.
2000, Wackernagel 1996).
Note that we deviate slightly from the naming
convention commonly employed in an EFA (e.g.,
Chambers ct al. 2000, Wackernagel 1996) in an attempt
to make the text easier to follow. It is typical for the
-------
population's ecological footprint to be denoted with EF
and the per capita ecological footprint identified with
ef. To be consistent with this format, we have deviated
from the convention by representing the population's
biocapacity (typically denoted as BC) as BC and the
per capita biocapacity as be (compared to the frequently
usedSQ.
The amount of land area appropriated per capita (cm) for
each major consumption item (c.) is estimated as:
(3.1)
where A' is the population of the region. The average
demand or ecological footprint per capita (ef) is
computed by summing all the ecosystem areas
appropriated (aa) by all purchased items in the annual
consumption of goods and services.
ef=Eaal (3.2)
Each item is converted into a footprint value, represented
as hectares per capita (ha/ca). Often these areas are
adjusted to express a world average and are converted
to global hectares. An ecological footprint (EF) for the
entire system or region is calculated by multiplying efby
the population size (A7).
EF = N(ef)
(3.3)
The biocapacity for the population (BC) is the number of
hectares of each of the six bioproductive land categories
in the area under study.
BC=LL, (3.4)
where L. indexes each of the six categories of land. A
per capita biocapacity' (be) is calculated by dividing the
region's 5Cby the population of the region.
• L.
(3.5)
An ecological balance is calculated by subtracting ef
from be to determine if there is an ecological deficit
or reserve. If e/exceeds be, the system is considered
unsustainable at the individual level, and if EF exceeds
BC, an ecological deficit exists and the system is
considered unsustainable regionally. Conversely,
the system is considered sustainable if there is an
ecological reserve (ef< be or EF < BC). However,
because we question the accuracy of identifying a
system as sustainable (see Chapter 1), we identify a
system's movement toward or away from sustainability
by examining the ecological balance. A system is
considered moving away from sustainability if the
ecological reserve is decreasing or, if there is an
ecological deficit the deficit is increasing through time.
3.2
The data needs of traditional EFA accounting are
relatively straightforward, although they can be intensive.
EFA requires food consumables such as meat, dairy, fish,
fruits and vegetables, animal feed, roots and tubers, and
pulses (i.e., legumes), and per capita energy consumption.
This level of detail creates difficulty when analyzing sub-
national regions because it is often difficult to obtain data
for local areas, especially parts of states. To overcome
this quandary, researchers often approximate values for
many of the variables or use values from disparate years
(e.g., Bagliani et al. 2008). This occurs because data,
on consumption patterns arc more likely to be collected
for national and even state level rather than for smaller
political units. Thus, depending on the data required, we
were limited by the availability of data sources due to
the regional nature of our study and the requirement for
26 consecutive yearly (1980-2005) values. Rather than
calculating a national EFA and scaling it to the region by
multiplying by the population of the region, we decided
to reduce the number of variables to those that were
necessary to calculate the EFA for the region. Because
our goal was to calculate a simplified ecological footprint
for the region and not necessarily a more detailed
ecological footprint as calculated in Wackcrnagel et al.
(2005), we attempted to restrict our variables to data
specific to Hie region (i.e., county-level data). However,
because data for some essential components of a footprint
analysis were not available at the individual or county
level, we had to utilize data that were not specific to the
region. When necessary, data that were available only
at the state or national level were scaled to Hie region
of study based on per capita rates in die larger systems
(Table 3.1). Any data that were reported less frequently
than annually were linearly interpolated in order to
calculate EFA for all 26 years.
3.3
We selected variables easily obtained and freely
available, thereby enabling a calculation that could
be undertaken by virtually any entity interested in
conducting its own EFA. Most variables were recorded
as reported from the original data source (Table 3.1).
However, biocapacity was calculated for some land
categories. Specifically, arable land was reported for
each county in the National Agricultural Statistics Sendee
(NASS) report at 5-year intervals (NASS 1984, 1989,
1994, 2009a, 2009b). Pasture was estimated using the
NASS reports by subtracting die hectares of arable land
from total farmland. Forest land was estimated from
United States Department of Agriculture (USD A) Forest
Service data and subtracting the amount of harvested
-------
forest for each count}'. Built land was obtained from the
US Census Bureau from the land area reported for each
Census-designated place in each county. Because the
region is land-locked, there was no bioproductive area
for the sea category. Lastly, energy land is the amount
of forested land required to sequester CO, produced
from energy consumption. This value was zero hectares
(ha) on the supply side of the equation because it is not
a supply, per se. Energy land is obtained by subtracting
from forest land the area necessary to sequester CO., and,
therefore, does not exist before CO, is produced.
Energy consumption daia for the State of Colorado
were obtained from the Department of Energy's
Energy Information Administration (EIA; EIA 2006).
We assumed that EIA data calculated on a per capita
usage and multiplied by the population of the SLB
was sufficient to estimate the energy consumption.
We justified this assumption by comparing coal,
natural gas. and propane consumption in the SLB to
their consumptive use for the State of Colorado. We
decided the more readily available EIA data adequately
represented consumption for these sources (see Appendix
3-A). Using EIA data for Colorado, consumption was
calculated as per capita, for the state and included coal,
natural gas, petroleum, nuclear, hydroelectric, and wood
and waste categories. An added benefit of using EIA
data, the petroleum category includes motor fuel, thereby
capturing energy consumption for transportation in the
SLB. Energy consumption was converted to area of
forest and the corresponding energy land required for
CO2 absorption, following Chambers et al. (2000).
Per capita food consumption for the US, from USDA
Economic Research Sendee (USDA-ERS 2009), was
used to estimate food consumption for the region under
study. Because data were already in quantity consumed,
it was not necessary to subtract exports from imports
and production to estimate consumption. Each food
item was assigned to an appropriate land category and
a per capita footprint was calculated for each food item
by dividing the kg consumed by the global yield for that
kind of land and dividing the result by the population
of the region. The footprint for each land category
was summed to produce the amount of land required to
support levels of consumption of the population of the
region.
Global yields were used because it is unlikely that
most items consumed in the region were produced in
the region. Global yields were taken from Chambers
et al. (2000: 70-73) and were constant for all years.
Ideally, global yields and conversion factors (see below)
would be calculated for each year and would result in
a better estimate of the EFA for a region. However.
more accurate annual values are unlikely to have a large
influence on the results. Another compromise in our
analysis is that many items are not included in this EFA
because every consumable is not reported or tracked at
the regional level and, thus, it is not currently possible to
include every item consumed.
Many studies suggest converting footprints into global
hectares (gha) in order to allow comparison of EFAs
between systems (e.g., Wackcrnagcl et al. 2004b,
Wiedmann and Lenzen 2007). A global hectare is
"normalized to the area-weighted average productivity of
biologically productive land and water in a given year"
(Global Footprint Network 2008). In other words, a
global hectare is the global average productivity (i.e., kg
produced) per one hectare of land or water and facilitates
comparison between EFAs for different nations.
Equivalence factors and yield factors are necessary to
convert actual ha into gha and they can be obtained
from the literature for certain years (e.g., Chambers et
al. 2000, Wackernagel and Yount 1998), but a complete
set is difficult to obtain without additional expense. To
overcome this lack of factors, we recorded values from
Chambers et al. (2000) and Mclntyre et al. (2007) for
alternative calculations using gha. However, our goal
was not to produce a mechanism by which regions can
be compared but rather to enable a region to analyze
their movement toward or away from sustainabiliry and
manage accordingly. Moreover, because the purpose of
our study was practical, rather than academic in nature,
we decided to use local hectares as the working unit.
3.4
We were able to assemble 26 years (1980-2005) of data
consisting of 35 variables of consumption (Table 3.1).
We also obtained data for agricultural production of the
eight major products in the region but these values were
not necessary to estimate consumption because national
consumption values were used. Appendix 3-B contains a
sample accounting ledger used in EFA and Appendix 3-C
contains the complete data worksheet.
Biocapacity shows a decreasing trend for the 26-year
period, peaking at nearly 36 hectares per capita (ha/ca) in
1980 and decreasing to less than 28 ha/ca in 2005 (Fig.
3. la; Table 3.2). Ecological footprint was relatively
flat, with a low of 5.1 ha/ca in 1997 to a high of 5.5 ha/
ca in 1985 (Fig. 3.la; Table 3.2). Subtracting e/from be
provides the ecological balance and reveals the SLB had
an ecological reserve, although it was declining, during
the period examined. The reserve was at a period high
of 30.54 ha/ca in 1980 and declined steadily though year
2005 when the ecological remainder was 22.34 ha/ca
(Table 3.2).
-------
3.5
3.5.1 (supply)
Although the SLB appears to have an abundance of
biocapacity, this is a result of low population density and
large area of pasture and arable lands. These two land
types typically increase the human carrying capacity
of a system and are heavily weighted in EFA as is
evident in the value the methodology places on these
(human-centric) biologically productive land categories.
Moreover, an abundance of forest is available to
sequester CO2 and provide renewable wood products.
However, it must be remembered the SLB is part of a
larger system and resources produced in the region help
support that larger system, including the rest of the US
and world.
The decrease in be seems to be due, primarily, to
population growth in the region. An increasing
population draws on a relatively fixed supply of
bioproductive land in the SLB, resulting in a decreasing
amount of bioproductive land per person. Specifically,
biocapacity of forest declined from 22.93 ha/ca to 18.56
ha/ca and pasture declined from 12.13 ha/ca to 7.17
ha/ca (Fig. 3.2). Arable land was variable during the
26-year period, starting at 5.67 ha/ca in 1980, peaking
in 1990 at 5.96 ha/ca, declining to 5.13 ha/ca in 2000,
and rising to 5.33 ha/ca in 2005 (Fig. 3.2). During this
period, built land increased slightly during the 26 years
(from 0.12 ha/ca to 0.15 ha/ca; Fig. 3.2). In absolute
terms, the number of hectares of arable land increased,
the hectares of forest remained constant, and the hectares
of pasture decreased in the SLB region during the 26
years. The results suggest the overall decline in EC
for the region was minimal, especially when compared
to the decline in be (Fig. 3. la and b). Hence, as the
population size increased, more people were drawing
on the same number of hectares of resources, resulting
in fewer resources available per person. This indicates
the decrease in biocapacity of forest and pasture land
categories were due to population growth and the be
available from these land categories. Conversely,
although biocapacity of arable land was variable, there
was an overall decrease (Fig. 3.2), indicating an increase
in the supply of total hectares of arable land in the region
that did not quite keep pace with population growth.
Lastly, the amount of built land increased from 4572
ha in 1980 to an estimated 7350 ha in 2005. This is
an increase of more than 60% and represents land that
typically, according to EFA accounting, was agricultural
land. In summary, the overall decrease in biocapacity in
the SLB in these calculations is largely due to population
increase. If the region were a self-reliant, closed system,
the carrying capacity- of the system would be far from
reached and many more individuals could be supported.
However, because the SLB provides resources to support
the larger system, and draws on resources from outside
the region (i.e., it is not a closed system), the excess
biocapacity should be evaluated with that in mind.
3.5.2 Ecological Footprint (demand)
Unlike the US (e.g., Wackernagel and Yount 1998, WWF
2008), the average ecological footprint of individuals in
the SLB decreased slightly (with some variation) during
the time analyzed (Fig 3.la). Ecological footprint (ef)
was 5.41 ha/ca in 1980 and finished the period at 5.13
ha/ca in 2005 (Fig 3. la). However, the high of 5.5 ha/ca
was in 1985 and the low of 5.10 ha/ca occurred in 1997
(Fig 3. la). Although e/showed a general decrease, the
EF for the SLB increased slightly (Fig. 3.1b), likely due
to an increase in population in the SLB.
Examining the individual land categories and the energy
footprint can help identify components responsible for
trends (Fig. 3.3). Overall, demand placed on each of the
land categories remained relatively constant during the
26-year period, except for forest (Fig. 3.3). Per capita
demand on pasture, arable, built, sea, and energy lands
have remained stable, as consumption did not increase
dramatically during the period. However, during the
same period the area of forest remained constant while
forest harvest increased from 23973 million board feet
(MBF) in 1980 to 35251 MBF in 1989 and lias been
declining since to 4444 MBF in 2005. This coupled with
the increase in population resulted in a decrease in area
of forest available per capita in the region.
Focusing on energy- land, there was a general increase
in total energy consumption from fossil fuels for the
region. However, although there was some variability,
per capita energy footprint revealed a slight decline for
coal, natural gas. and petroleum consumption (slope =
-0.0021, -0.002, and -0.004, respectively) in the 26-
year period. Energy consumption from nuclear and
hydroelectric was nonexistent and is a source that could
be expanded to benefit the region. Wood and waste
usage, however, increased during the first 6 years of
the period analyzed and has been decreasing both per
capita and for the population since 1986 (Fig. 3.4). The
US Census Bureau provides a category of number of
homes using wood for heating and this value for the SLB
decreased from 1990 to 2000 census from 3750 to 2327
homes (US Census Bureau 1990, 2000), suggesting that
such a reduction was realistic.
3.5.3 Global Hectares
Converting ha into gha increased e/by less than 10%,
or about 0.5 ha and available be nearly doubled. For
example, in 1980 the e/was 5.41 ha (5.94 gha) and be
-------
(minus 12% for biodiversity) was 35.95 ha (64.39 gha).
The ecological reserve was 30.54 ha (58.46 gha). In
2005 £Fwas 5.13 ha (5.51 gha), be was 27.47 ha (49.31
gha). and the remainder was 22.34 ha (43.80 gha). Of
course, the conversion from ha to gha would be better
using year specific yield and equivalence factors, but
the difference likely would be minor. For example,
Lammers et al. (2008) report an equivalence factor for
energy land of 1.36 in 1980 and 1.38 in 2003 in their
time-series analysis of Ireland. However, our interest
was in trends (i.e., was the system moving toward or
away from sustainability). not necessarily the actual
value and may not be worth the added expense or
difficult}". However, we think using ha is acceptable if
one is not comparing between different regions.
3.5.4 Examination of Sustainability
Whether using ha or gha, our analysis suggests the
SLB was moving away from sustainability, on both a
per capita basis (Fig. 3. la) and a population basis (Fig.
3.1b). Although the e/remained relatively flat during
the 26 years (which by itself would suggest the SLB
was not moving away from sustainability), available
be decreased because of increasing population size.
These trends reveal a decreasing ecological reserve
(the space between the two lines is decreasing; Fig. 3.1
and Table 3.2). The decrease in the ecological reserve
over the 26 years was amplified on a population level
examination. The EF was increasing and BC was
decreasing through time (Figs. 3.1 and 3.5). This was
evident in the individual land categories as well. For
example, per capita food consumption (e.g.. meat,
cereal, fruit and vegetable, and fish) increased during
the period, requiring more bioproductive land to satisfy
the consumption and the amount of available arable and
pasture land per capita decreased. The overall result was
a decreasing surplus of bioproductive land (Fig. 3. la)
that was available to support the population in the SLB.
If the trend continues, e/will eventually overshoot be,
resulting in an ecological deficit.
Although our analysis suggests the efis relatively low,
one potential flaw was the large number of missing
variables thai: represent embodied energy in goods and
services. For example, Utah Vila! Signs reported over
2 gha/ca were required to meet the consumptive levels
for Utah in 1990 and about 1.5 gha/ca in 2003 (Mclntyre
et al. 2007). When compared to the food and energy
consumption for Utah in 1.990 (5.2631 gha/ca) and 2003
(5.8239 gha/ca), it is not an inconsequential quantity.
However, these variables do not represent such a large
part of EFA that they alter Hie trends captured by food
and energy consumption. It is important to note that a
portion of data in that report came from both US and
Utah consumption rates and may or may not accurately
represent consumptive levels for the Utah population.
The trends in the SLB were not unexpected. Because of
the economic, environmental, and social characteristics
of the SLB, we did not expect a substantial amount of
change to occur during the 26 years examined. The SLB
is an agricultural-based region and has many traits that
accompany this type of society (e.g., San Luis Valley
Development Resources Group 2007). Moreover, the
SLB functions as a resource exporting region, providing
much of its bioproductive land and resources to support
larger systems, which include those of Colorado, the US,
and the world.
3.5.5 Strengths and Weaknesses of
Methodology
The simplified EFA we conducted has captured a
trend indicating the SLB was moving away from
sustainability. There are a number of strengths and
weaknesses to the methodology we employed. The
greatest strength of our methodology is the use of
readily available data that makes the methodology
available to any entity interested in computing their
ecological footprint to determine if their system is
moving toward or away from sustainability. The
greatest weakness of our methodology may be the
exclusion of potentially important variables. The lack
of data at sub-national scales resulted in the exclusion
of well over 100 variables that are typically included
in a traditional footprint account. All of the variables
usually included in the embodied energy component,
as well as some other variables, were excluded. In
addition, the footprint methodology in general does
not take into account exports of natural resources that
may lower the biocapacity. Another limitation is the
lack of data for each year of our study. For example,
we had data at five year intervals for agricultural
production from the Census of Agriculture and ten year
intervals for population and housing variables from
the Decennial Census of Population and the census
years did not align (i.e., decennial censuses were 1980,
1990, and 2000 whereas the agricultural censuses were
1978, 1982, 1987, 1992, 1997, and 2002). We used
linear interpolation to estimate the values for missing
years. Because of these imputations, we were limited
in performing statistical analysis on trends in the results
(e.g., a Mann-Kendall test for tends or time-series
analysis). Such interpolation would bias the trend
especially if a measure of uncertainty was available and
we could statistically test the trend. Nonetheless, after
calculating the ecological footprint for the US using the
same variables employed in our study, we are confident
the methodology docs an acceptable job capturing
-------
trends in an EFA for a region, which was a central goal
of this project (see Appendix 3-D for comparison with
US). Although not unimportant, the embodied energy
component of a typical ecological footprint account
generally accounts for less than the food and energy
consumption components. For instance. Lammers et al.
(2008) found that less than 1 ha per capita was from the
embodied energy component. These weaknesses, then,
need to be balanced with the availability of the method
and its relative ease of use. However, there is enough
detail in the account that stakeholders can identify areas
of the system on which to focus attention to improve
sustainability of the system.
Although there were several methodologies used early
on in EFA accounting, recently there lias been a push
to standardize the methods (e.g., Wackernagel 2009).
Specifically, a couple of groups have proposed their
methodology as a standard (e.g., Venetoulis and Talberth
2008 - Redefining Progress; Kitzes et al. 2009 - Global
Footprint Network, Best Foot Forward). Unfortunately,
these groups do not publish their complete
methodologies or data, in part because they provide
a commercial EFA calculating service. Because the
groups do not provide the information free of charge or
make it readily available, it is difficult to conduct an EFA
with limited budgets which is one of our goals for our
methodology. Moreover, this standardization has some
drawbacks that make the calculation more difficult than
it need be. First, the standard recommends starting with
national level data. However, a subsystem or region may
not reflect a national average, or levels of consumption
may not be typical and this could be concealed using
national data. Thus, scaling these variables to the
population of the subsystem may not adequately
represent Hie subsystem. Second, these groups do not
make the factors that are necessary for conversion to a
common normalized unit (global hectare, gha) readily
available. Our goal in this project was not to create a
"better" or more accurate EFA; rather it was to make the
methodology more accessible and usable for the end user
while still being adequate for sustainability analysis.
3.6
Bagliani, M., Bravo, G., Dalmazzone, S. 2008. A
consumption-based approach to environmental
Kuznets curves using the ecological footprint
indicator. Ecological Economics 65, 650-661.
Chambers, N., Simmons, C., Wackernagel, M. 2000.
Sharing nature's interest: ecological footprints as an
indicator of sustainability. Earthscan Publications,
Ltd., London. 185 pp.
Chen, B., Chen, G.Q. 2006. Ecological footprint
accounting based on emergy—a case study of the
Chinese society'. Ecological Modelling 198. 101-104.
Collins, A., Flynn, A., Wiedmann, T., Barrett, J. 2006.
The environmental impacts of consumption at a
subnational level. Journal of Industrial Ecology 10,
9-24.
Eaton, R., L, Hammond, G.P., Laurie, J. 2007.
Footprints on the landscape: an environmental
appraisal of urban and rural living in the developed
world. Landscape and Urban Planning 83, 13-28.
EIA (Energy Information Administration). 2006. Slate
Energy Data 2006: Consumption. State Energy Data
System. Available online at http://www.eia.doe.gov/
emeu/states/_seds.html. Last accessed March 17,
2008.
Erb.K.-H. 2004. Actual land demand of Austria 1926-
2000: a variation on Ecological Footprint assessments.
Land use Policy 21, 247-259.
Global Footprint Network. 2008. The ecological
footprint atlas 2008, Version 1.0. Global Footprint
Network. Available online at http://www.
footprintnetwork.org. Last accessed July 21, 2009.
Graymore, M.L.M.. Sipe, N.G., Rickson, R.E. 2008.
Regional sustainability: how useful are current tools
of sustainability assessment at the regional scale?
Ecological Economics 67, 362-372.
Huang, Q., Wang, R., Ren, Z., Li, J., Zhang, H. 2007.
Regional ecological security assessment based on long
periods of ecological footprint analysis. Resources,
Conservation and Recycling. 51, 24-41.
Kites, J., Galli, A., Bagliani, M., Barrett, J., Dige, G.,
Ede, S., Erb, K., Giljum, S., Haberl, H., Hails, C.,
Jolia-Ferrier, L., Jungwirth, S., Lenzen. M.. Lewis,
K., Loh, J., Marchettini. N., Messinger, H.. Milne.
K., Miles, R., Monfreda, C., Moran, D.. Nakano, K.,
Pyhala, A., Rees, W, Simmons, C., Wackernagel, M.,
Wada, Y., Walsh, C, Wiedmann, T. 2009. A research
agenda for improving national Ecological Footprint
accounts. Ecological Economics 68, 1991-2007.
Lammers, A., Moles, R., Walsh, C., Huijbregts, M.A.J.
2008. Ireland's footprint: a time scries for 1983-2001.
Land Use Policy 25, 53-58.
Lenzen, M., Murray, S.A. 2001. A modified ecological
footprint method and its application to Australia.
Ecological Economics 37, 229-255.
-------
Lenzen, M., Murray, S.A. 2003. The ecological
footprint - issues and trends. ISA research Paper 01-
03, the University of Sydney integrated Sustainability
Analysis. 27 pp. Available online at http://www.isa.
org.usyd.edu.au/publications/documents/Ecological_
Footprint_lssues_and_Trends.pdf. Last accessed July
13, 2009.
Li, L., Chenghua, L., Qi, W., Zhijiang, W., Liyu, D.
2008. Detailed ecological footprint accounting of
Liaoning in 2006. Chinese Journal of Population,
resources and Environment 6, 78-87.
Luck, M.A., Jenerette, G.D., Wu, J., Grimm, N.B.
2001. The urban funnel model and the spatially
heterogeneous ecological footprint. Ecosystems 4,
782-796.
Mclntyre, S., Peters, H.M., Chrislensen, M.H., Emmi,
PC., Martinson, W., Mielkc, M., Senbel, M., Stark,
D.O. 2007. The ecological footprint of Utah: a
Sustainability indicators project of the Utah Population
and Environment Coalition. Utah Vital Signs, 26
June. 59 pp. Available online at http://www.utahpop.
org/vitalsigns/research/report_2007.htm. Last accessed
July 13,2009.
Medved, S. 2006. Present and future ecological
footprint of Slovenia—the influence of energy demand
scenarios. Ecological Modelling 19. 25-36.
Monfreda, C., Wackernagel, M., Deumling, D. 2004.
Establishing national natural capital accounts based on
detailed Ecological Footprint and biological capacity
assessments. Land Use Policy 21, 231-246.
NASS (National Agricultural Statistics Service, United
States Department of Commerce, Bureau of the
Census). 1984. United States Census of Agriculture:
1982. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington. DC:
US Government Printing Office.
NASS (National Agricultural Statistics Sendee. United
States Department of Commerce, Bureau of the
Census). 1989. United States Census of Agriculture:
1987. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Sendee, United
States Department of Commerce, Bureau of the
Census). 1994. United States Census of Agriculture:
1992. Volume 1; Geographic Area Series; Part 6
Colorado State and County Data. Washington. DC:
US Government Printing Office.
NASS (National Agricultural Statistics Sen-ice, United
States Department of Commerce, United States
Department of Agriculture). 2009a. United States
Census of Agriculture: 1997. Available online
at http ://www. nass .usda. gov/Statistics_by_State/
Colorado/index.asp. Last accessed June 26. 2009.
NASS (National Agricultural Statistics Service, United
States Department of Commerce. United States
Department of Agriculture). 2009b. United States
Census of Agriculture: 2002. Available online
at http://www.nass.usda. gov/Statistics_by_State/
Colorado/index.asp. Last accessed June 26, 2009.
Niccolucci, V., Galli, A., Kitzes, J., Pulselli, R.M.,
Borsa, S.. Marchettini. N. 2008. Ecological Footprint
analysis applied to the production of two Italian wines.
Agriculture, Ecosystems and Environment 128, 162-
166.
Noss, R.F., Cooperrider, A.Y. 1994. Saving Natures
Legacy - Protecting and Restoring Biodiversity.
Washington, DC, Island Press.
Patterson, T.M., Niccolucci. V., Bastianoni, S. 2007.
Beyond "more is better": ecological footprint
accounting for tourism and consumption in Val di
Merse, Italy. Ecological Economics 62, 747-756.
Pulselli, F.M., Ciampalini, E, Leipert, C., Tiezzi, E.
2008. Integrating methods for the environmental
Sustainability: the SPIn-Eco Project in the Province of
Siena (Italy). Journal of Environmental Management
86, 332-341.
San Luis Valley Development Resources Group. 2007.
Comprehensive economic development strategy.
Accessed at: http://www.slvdrg.org/ceds.php on June
17, 2009.
Scotti, M., Bondavalli, C., Bodini, A. 2009. Ecological
Footprint as a tool for local Sustainability: the
municipality of Piacenza (Italy) as a case study.
Environmental Impact Assessment Review 29, 39-50.
US Census Bureau. 1990. 1990 Census, Summary
File 1, Table DP-5. Profile of Selected Housing
Characteristics, Colorado. US Department of
Commerce, Washington, DC. http://facrfinder.census.
gov. Last accessed May 15,2009.
US Census Bureau. 2000. 2000 Census, Summary
File 1, Table DP-4. Profile of Selected Housing
Characteristics, Colorado. US Department of
Commerce, Washington, DC.
-------
USDA-ERS (United States Department of Agriculture
Experimental Research Station). 2009. U.S. per
capita food availability' data sets, http://www.ers.usda.
gov/Data/FoodConsumption/FoodAvaillndex. htm Last
accessed April 17, 2009.
Venctoulis, I, Talbcrth, J. 2008. Refining the
ecological footprint. Environment. Development and
Sustainability 10, 441-469.
Wackernagel, M. 2009. Introduction: methodological
advancement in footprint analysis. Ecological
Economics 68, 1925-1927.
Wackernagel, M., Monfreda, C., Erb, K.-H., Haberl,
H., Schulz, N.B. 2004a. Ecological footprint time
series of Austria, the Philippines, and South Korea for
1961-1999: comparing the conventional approach to
an 'actual land area' approach. Land Use Policy 21,
261-269.
Wackernagel, M., Monfreda, C., Schulz, N.B., Erb,
K.-H., Haberl, H., Krausmann, F. 2004b. Calculating
national and global ecological footprint time series:
resolving conceptual challenges. Land Use Policy 21,
271-278.
Wackernagel, M., Moran, D., Wemier, P., Goldfinger, S.,
Deumling, D., Murray, M. 2005. National footprint
and biocapacity accounts 2005: the underlying
calculation method. Global Footprint Network,
Oakland. 33 pp.
Wackernagel, M., Onisto, L., Bello, P., Linares,
A.C., Lopez Falfan, I.S., Mendez Garcia, J., Suarez
Guerrero, A. L, Suarez Guerrero, M.G. 1999.
National natural capital accounting with the ecological
footprint concept. Ecological Economics 29,375-390.
Wackernagel, M., Rees, W. 1996. Our ecological
footprint: reducing human impact on (lie earth. New
Society Publishers, Gabriola Island, BC. 160 pp.
Wackernagel, M., Yount, J.D. 1998. The ecological
footprint: an indicator of progress toward regional
sustainability. Environmental Monitoring and
Assessment 51, 511-529.
Wiedmann, T., Lenzen, M. 2007. On the conversion
between local and global hectares in Ecological
Footprint analysis. Ecological Economics 60, 673-
677.
Wood, R., Garnett, S. 2009. An assessment of
environmental sustainability in Northern Australia
using the ecological footprint and with reference to
Indigenous populations and remoteness. Ecological
Economics 68, 1375-1384.
WWF. 2008. Living Planet Report 2008. Published in
October 2008 by WWF-World Wide Fund For Nature
(formerly World Wildlife Fund), Gland, Switzerland.
Available online at http://assets.panda.org/downloads/
living_planet_report_2008.pdf. Last accessed July 23,
2009.
Zhao, S., Hong, H., Zhang, L. 2008. Linking the
concept of ecological footprint and valuation of
ecosystem services: a case study of economic growth
and natural earning capacity. International Journal
of Sustainable Development & World Ecology 15,
448-456.
-------
Table 3.1- Scale and source of data for the variables used to calculate ecological footprint for the SLB.
Scale
National
State
County
Variable
Consumption of
Animal Products
(Ibs / capita)
Cereal Consumption
(Ibs / capita)
Fruit & Vegetable
Consumption
(Ibs / capita)
Fossil Fuel (Gj / year)
Nuclear Electric
Consumption
(million kWh)
Hydro-electric
Consumption
(million kWh)
Wood and waste
Consumption
(trillion B'l'U)
Population (persons)
Land Area (ha)
Built land (ha)
Arable land (ha)
Forest including
deforestation (ha)
Purpose/Notes
Includes bovine,
buffalo, sheep, goat,
non-bovine, milk,
cheese, butter, eggs,
and fish categories
Includes cereals,
wheat, and maize
categories
Includes vegetables,
fruit, animal feed,
roots, tubers, and
pulses consumption
Includes coal, natural
gas. and petroleum
consumption
Census designated
places based on 1990
and 2000 data
1982, 1987, 1992,
1997, 2002, 2007
1987, 1997, 2002 and
treated as a constant
Source
USDA Economic
Research Service
USDA Economic
Research Service
USDA Economic
Research Service
Energy Information
Administration
Energy Information
Administration
Energy Information
Administration
Energy Information
Administration
Bureau of Economic
Analysis
US Census Bureau
US Census Bureau
USDAAgricultural
Statistics
USDA Forest Service
Website
www.ers.usda.gov
www. ers . usda. gov
www. ers . usda . gov
www.eia.doe.gov
www.eia.doe.gov
www. eia.doe.gov
www.eia.doe.gov
www.bea.gov
www. census, gov
www. census. gov
www.agcensus.
usda.gov
www.fia.fs.fed.us
Specific data link
www.ers.usda.gov/Data/'
FoodConsumption/
FoodAvailQueriable.aspx
www.ers.usda.gov/Data/
FoodConsumption/
FoodAvailQueriable.aspx
www. ers. usda.gov' Data/
FoodConsumption'
FoodAvailQueriable.aspx
www. ei a. doe . gov/
emeu/states/state.
html?q state a=co&q
state=COLORADO
w ww. ei a. doe . gov/
emeu/states/state.
html?q state a=co&q
state=COLORADO
w ww. ei a. doe . gov/
emeu/states/state.
html?q state a=co&q_
state=COLORADO
www. eia . doe . gov/
emeu/states/state.
html?q state a-co&q
state-COLORADO
www. b ea . go v /
regional/reis/defaul t.
clm?selTab1e=CAl-
3§ion~2
quicklacts.census.gov/
qi'd/states/08000.html
and for US
www. census . go v/'geo/
cob/bdy/pl'pl90eOO/
pl08_d90_eOO.zip
http :/'www. agcensus.
usda.gov
www.fia.fs.fed.us/tools-
data/spatial
-------
Scale
Calculated
Variable
Pasture (ha)
Sea (ha)
Forest harvest (MBF)
Energy land (ha)
Non-productive land
(ha)
Purpose/Notes
total farmland minus
arable land; values
reported as "D" in
report are listed as
values from adjacent
survey
No sea category in
biocapacity in SLB
Total land area minus
the sum of EF land
area
Source
USDAAgricultural
Statistics
l.USDA Forest
Service
2. USDA Natural
Resources and
Conservation Service
Website
www.agcensus.
usda.gov
1 . Data generated
by Bruce F.
Short. Biological
Resources Stall"
Officer, San Luis
Valley Public
Lands Center
2. National
Resources
Inventory 2001
Annual NRI,
Urbanization and
Development of
Rural Land, July
2003 (and see
www. nrc s . us d a .
gov/technical/
land)
Specific data link
No sea category in
biocapacity in SLB
1 . Bruce F. Short.
Biological Resources
Staff Officer, San Luis
Valley Public Lands
Center, 1 803 West
Highway 160, Monte
Vista, CO 81 144,
bsriort@is.ied.us, phone
(719)852-6225
2. USDA, Natural
Resources and
Conservation Service.
National Resources
Inventory 2001 Annual
NRI, Urbanization and
Development of Rural
Land, July 2003 (and
see www.nrcs.usda.gov/
technical/land)
Based on CO, production
and subtracted from
forest land. Thus,
biocapacity is zero.
Total land area minus the
sum of EF land area
-------
Table 3.2 - Ecological footprint and biocapacity in the San Luis Basin, Colorado, 1980-2005. Twelve percent of the
biocapacity area was subtracted from the total biocapacity to protect biodiversity.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Ecological Footprint
(hectares per capita)
5.41
539
5.39
5.31
5.45
5.50
5.49
5.37
5.45
5.37
5.42
5.31
5.30
5.31
5.30
5.20
5.22
5.10
5.14
5.15
5.26
5.26
5.14
5.15
5.22
5.13
Biocapacity (hectares
per capita)
35.95
34.91
33.84
33.23
32.85
32.88
32.46
32.15
32.06
32.42
32.65
32.23
32.41
32.10
31.22
30.30
29.73
29.21
28.77
28.43
27.97
28.04
27.74
27.65
27.34
27.47
Ecological Balance
(hectares per capita)
30.54
29.52
28.45
27.92
27.39
27.37
26.97
26.78
26.61
27.04
27.23
26.92
27.11
26.78
25.92
25.10
24.51
24.11
23.64
23.29
22.71
22.78
22.60
22.50
22.12
22.34
-------
•ef
•be
a
03
*—>
'a.
8
ca
"G
CD
40.00
35.00 -
30.00 -
25.00
20.00 -\
15.00
10.00 -\
5.00
0.00
1980 1983 1986 1989
1992 1995
year
1998 2001 2004
•EF
•BC
OT
0)
03
0
CD
t~
O
o
O
1,600 -,
1,400 -
1,200 -
1,000 -
800 -
600 -
400 -
200 -
n
b
.^
1980 1983 1986 1989 1992 1995 1998 2001 2004
year
Figure 3.1- Biocapacity compared to ecological footprint for 26 years in the San Luis Basin. Twelve percent of
the biocapacity area was subtracted from the total BC and be to protect biodiversity, a - Per capita biocapacity (be)
compared to per capita ecological footprint (ef). b - Biocapacity for the population (BC) compared to total ecological
footprint (EF).
-------
45.00 -
. 40.00 •
re
" 35.00 H
re
so.oo -
re 25.00 -\
Q.
re
g 20.00 -
o> 15.00H
.a
= 10.00
5
** 5.00 4
0.00
• Forest HSea • Energy D Pasture O Arable
1980 1983 1986 1989 1992 1995 1998 2001
2004
year
Figure 3.2 - Available biocapacity (be) for the five footprint land categories over a 26-year period in the San Luis
Basin, Colorado. Twelve percent of the biocapacity area was subtracted from the total be to protect biodiversity. Note
that built-up land has no biocapacity and is not included in the figure.
6.00 -i
5.00 -
4.00
3.00 -
2.00 -
Q.
"5
M—
1
g> 1.00
"O
O
LJJ
0.00
I Forest D Built-up BSea • Energy D Pasture D Arable
1980 1983 1986 1989 1992 1995 1998 2001 2004
year
Figure 3.3 - Ecological footprint (ef) and the demand placed on each of the six land categories over a 26-year period
in the San Luis Basin, Colorado.
-------
CD
C
LU
n Hydro-electric
• Natural Gas
n Nuclear
n Petroleum
• Wood and waste
• Coal
o
CO
c
o
"a.
E
w
c
8
>s
3.00 -
2.50 -
2.00 -
1.50 -
1.00 \
0.50
0.00
1980 1983 1986 1989
1992 1995
1998 2001
2004
year
Figure 3.4 - Energy consumption (hectares per person) by major energy source over a 26-year period in the San Luis
Basin. Colorado.
25.00 n
_ 20.00 -
re
o
re
15.00 -
0)
o
c
•5 10.00-
re
re
'I, 5.00 H
0.00
-5.00
•Arable
Pasture
•Sea
Built-up
1980 1983 1986 1989
1992 1995
year
1998 2001
2004
Figure 3.5 - Ecological balance (ha/ca) of the different land categories in the SLB. Remainder equals biocapacity
per capita minus the ecological footprint for the SLB region. Positive values represent a surplus or remainder of
biocapacity and negative values represent a deficit of biocapacity. Twelve percent of the biocapacity area was
subtracted from the total be to protect biodiversity.
-------
Appendix 3-A: of
to EIA
For exposition purposes, this appendix demonstrates
the applicability of using state level data as an estimate
of individual energy consumption in the San Luis
Basin. State level data are easier to obtain, but may not
accurately represent regional level energy consumption.
Energy consumption data for the State of Colorado
were obtained from the Department of Energy- 's
Energy Information Administration (EIA; EIA 2006).
We assumed that EIA data calculated on a per capita
usage and multiplied by the population of the SLB
was sufficient to estimate the energy consumption. We
justified this assumption by comparing coal, natural
gas, and propane consumption in the SLB to their
consumptive use for the State of Colorado.
Xcel Energy and San Luis Valley Rural Electric
Co-Op, the primarj' energy suppliers in the SLB,
provided usage rates for their customers. We estimated
coal consumption by starting with actual electricity
consumption for 2006. The amount of electricity
produced from coal is highly dependent on the type of
coal used (EIA 2009). We assumed it takes 0.521 short-
tons (472.56 kg) of coal to produce 1 MWh of electricity
(Heede 2005), which works out to approximately 1.024
Ibs per kWh. In 2006, the population of the SLB was
nearly 1% of Colorado's population. Therefore, the SLB
would use about 1% of the energy in the state. Using
data provided by the energy suppliers indicates that
residents in the SLB used 236895 short tons (1.18% of
2006 state estimates) of coal and 1.5 billion cubic feet
(0.33% of 2006 state estimates) of natural gas.
We attempted to use the same strategy' for propane
consumption, but were unable to get consumption
rates of propane in the SLB from suppliers. Therefore,
a crude estimate of propane use was generated by
examining the relationship between heating degree days
(HDD) and quantity of propane used in a household. We
averaged the number of HDD from 20 weather stations
located in and around the SLB (NOAA Western Regional
Climate Center, http://www.wrcc.dri.edu/). Based on
an estimate of 1200 gallons of propane used per year
for a 2600 ft2 home in Montrose, CO (USD A 2006) and
6360 mean number of HDD reported at the Montrose 2
weather station, we estimated 0.189 gallons per HDD.
However, the majority (US Census Bureau 2000) of
homes in the SLB were built before 1979 and the median
house size of new construction in the US in 1972 was
around 1,400 ft2 (US Bureau of the Census 1976).
Therefore, we used the mean of 0.102 gallons per HDD
(consumption levels are dependent on numerous factors
including house size, lifestyle, energy efficiency of
appliances, how well a structure is insulated, etc.). The
mean annual number of HDD in the SLB was 9006. The
result, based on HDD, is an estimated average of 914.98
gallons (21.78 barrels) of propane used per household
(approximately 29%, or 5200, of the 17,687 households
are heated with propane; US Census Bureau 2000).
Next, we estimated the per household propane
consumption for Colorado. According to EIA data,
around 40.6% (2632.504 K barrels) of propane was used
by the residential sector in Colorado (EIA 2006) in 2000.
There were 1,808,037 households in the slate in 2000
and 6.2% (112098 households) of them used propane
to heat their home (US Census Bureau 2000). Of the
6484 thousand barrels of propane consumed in the state,
40.6% (2632.504 thousand barrels) were used by the
residential sector (EIA 2006). Thus, the average propane
use for Colorado was 986.32 gallons per household.
Based on HDD (0.102 gallons / HDD / household),
the SLB would consume 113,283 barrels of propane
in a typical year. This is 4.30% of Colorado's total
residential propane use. If we use state level data (986.3
gallons / household), the SLB would consume 122,116
barrels of propane in a typical year. This is 4.64% of
Colorado's total for residential propane use. Propane
consumption is relatively similar for either method and
would not adversely affect the energy' consumption in
the SLB. Of course, this only accounts for residential
consumption, but may adequately represent total
consumption as well.
-------
A summary of these calculations is below:
Data sources (based on 2000 census data, US Census Bureau)
EIA data for Colorado,
1,808,037 households
112,098 use propane (-6.2%)
6,484,000 barrels of propane consumed
40.6% of propane consumed for residential use
San Luis Basin data from energy suppliers:
17,687 households
5200 use propane
9006 average heating degree days (HDD)
1400 sq ft average house size
Montrose. CO (in SLB) special study data:
2600 sq ft house
1200 gallons propane per year
6360 HDD (at Montrose 2 weather station)
SLB propane energy use proportioned from state EIA data:
0.406 residential proportion * 6,484,000 barrels =
2632504 barrels of propane for residential use in CO
2,632,504 barrels / 112,098 households -
23.48395 ban-els of propane per household in CO
23.48395 barrels/household * 5200 households in SLB -
122,116 barrels consumed in SLB per year = 4.64% of CO's propane use (122,116 barrels / 2,632,504 barrels * 100)
SLB propane energy use based on SLB HDD data:
1200 gallons / 42 gallons per barrel - 28.57143 barrels used in Montrose
1400 sq ft SI ,B / 2600 sq ft Montrose - 0.53 84615 SI ,B proporti on of Montrose use
9006 HDD SLB / 6360 HDD Montrose - 1.416038 SLB fraction of Montrose HDD
28.57143 * 0.5384615 * 1.416038 - 21.78520 barrels per household in SLB
21.78520 barrels/household * 5200 households in SLB -
113,283 barrels consumed in SLB per year = 4.30% of CO's propane use (121,375 barrels / 2,632,504 barrels * 100)
Percent difference: (EIA- HDD) / HDD = 7.8%
Because of the ease of getting energy consumption from EIA and because state level data are a consistent source of
energy consumption values, we think they are an adequate surrogate for actual consumption energy data in the region.
Using EIA data for Colorado, consumption was calculated as per capita for the state and included coal, natural gas,
petroleum, nuclear, hydroelectric, and wood and waste categories.
-------
References
Chambers, N., Simmons, C.. Wackernagel. M. 2000.
Sharing nature's interest: ecological footprints as an
indicator of sustainability. Earthscan Publications,
Ltd., London. 185 pp.
El A (Energy Information Administration). 2006. State
Energy Data 2006: Consumption. State Energy Data
System. hltp://www.eia.doe.gov/emeu/slales/_seds.
html. Last accessed March 17, 2008.
El A (Energy Information Administration). 2009.
Energy Kid's Page: coal a fossil fuel, http://www.eia.
doe.gov/kids/energyfacts/sources/non-renewable/coal.
html. Last accessed August 31, 2009.
Heede, R. 2005. Aspen emissions inventor}': electricity
carbon factor. Climate Mitigation Services.
http://aspcnglobalwanning.com/pdf/02Carbon&Notcs.
pdf. Last accessed 23 July 2009.
NOAA (National Oceanic and Atmospheric
Administration) Western Regional Climate Center.
http://www.wrcc.dri.edu/. Last accessed April 23,
2008.
US Bureau of the Census. 1976. Construction Report—
Scries C25 Characteristics of New Housing; 1975.
US Department of Commerce, Washington, DC.
US Census Bureau. 2000. 2000 Census, Summary
File 1, Table DP-4. Profile of Selected Housing
Characteristics, Colorado. US Department of
Commerce, Washington DC.
USD A. 2006. Rural Development Electric Programs
Success Stories, Colorado: Delta-Montrosc Electric
Association Home Energy Makeover Contest. May.
www.usda.gov/rus/electric/success/co-success htm.
Last accessed July 23, 2009.
-------
Appendix 3-B: Footprint
Accounting Ledger for 1980 in the
Luis
This appendix contains the basic tables used in
ecological footprint accounting. Tables list the main
resources produced and consumed and the supply
(biocapacity) and demand (ecological footprint) of land
area in the San Luis Basin region.
Table 3-B. 1 - Ecological footprint accounting ledger for food consumption for 1980 in the San Luis Basin. Local
Yield and Production are italicized because they were not needed to compute apparent consumption. Global yields are
from Chambers et al. (2000).
Resource
Unit
MEAT
.meal (fresh)
. .bovine, buffalo
..sheep, goat
..non-bovine, goat, mutton, buffalo
DAIRY
. .milk
..cheese
..butter
•-eggs
FISH
.marine, inland fish
GRAINS
.cereals
..wheat
. .maize
FRUITS & VEGETABLES
.vegetable & fruit
..vegetable etc.
..fresh fruit
ANIMAL FEED
..animal feed
ROOTS & TUBERS
.roots & tubers
PULSES
.pulses
Global yield
(kg /ha)
32
72
764
458
46
22
573
35
2752
2752
8136
2752
12814
802
Local yield
(kg /ha)
199.243
188.715
7.379
3. 149
0.000
0.000
0.000
0.000
0.000
1229.170
879.013
350.157
0.000
0.000
0.000
0.000
0.000
0.000
1847.935
0.000
2278.070
0.000
0.000
Production
(kg)
92,990,880
88,077,314
3,443,739
1,469,828
268,326,225
191,887,425
76,438,800
403,401,978
497,299,806
Apparent
consumption
(kg)
2.919,258
1,280,774
17,449
1,621.035
4,140,803
305,362
1364,454
218,465
4.590,896
579,315
3,964,468
225,095
10,617,864
10.617,864
5.896,099
4,721,765
1,280,984
93,642
Footprint
component
(ha / capita)
1.0404
0.0063
0.0552
0.2350
0.1726
0.0000
0.0619
0.1623
0.0434
0.0055
0.0339
0.0000
0.0026
0.0030
Land category
pasture
pasture
pasture
sea
sea
arable
arable
arable
arable
arable
arable
-------
Table 3-B.2 - Ecological footprint accounting ledger for energy- consumption for 1980 in the San Luis Basin. Global
average energy to land ratio was used to convert joules to hectares (Chambers et al. 2000).
Primary energy use *
Unit
Coal consumption (short tons = 2000 Ibs)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Total fossil fuel consumption
Nuclear energy consumption
Hydro-electric consumption
Global average
energy to land ratio
(Oj / ha' yr)
55
93
71
71
1000
Energy use
(Gj/SLB/yr)
3880385.04
12625.97
4965333.44
340771652986
31755.94
81746.54
Footprint component
(ha / capita / yr)
0.5468
168.0439
0.4273
66.8133
25.9548
Land type
energy land
energy land
energy land
energy land
energy land
Table 3-B.3 - Ecological footprint accounting ledger for supply and demand of bioproductrve land for 1980 in the
San Luis Basin. Equivalence factors are from Mclntyre et al. (2007) and Chambers et al. (2000) and were treated as
constants for the 26 years.
Demand - Footprint (per capita)
Category
Unit
Fossil energy
Built-up land
Arable land
Pasture
Forest including deforestation
Sea
Total used
Total
(ha / capita)
2.88
0.67
0.09
1.57
0.05
0.16
5.41
Equivalence factor
1.17
2.22
2.22
0.49
1.35
0.36
Equivalence total
(gha / capita)
3.37
1.48
0.20
0.77
0.07
0.06
5.94
Supply - Existing regional capital (per capita)
Category
Unit
Fossil energy-
Built-up land
Arable land
Pasture
Forest including deforestation
Sea
Other
Total existing
Total available (minus 12% for biodiversity)
ECOLOGICAL REMAINDER
Yield factor
1.42
1.42
1.63
1.97
1.28
Regional area
(ha / capita)
0.12
5. 39
12.18
22.93
0.00
14.30
40.61
35.74
30.32
Yield adjusted
equivalent area
(gha / capita)
0.17
7.65
19.85
45.17
0.00
72.84
64.10
58.16
-------
Appendix 3-C: Twenty-six Years (1980-2005) of Data for the San Luis Basin,
Colorado.
Table 3-C. 1 - Twenty-six years (1980-2005) of data for the San Luis Basin, Colorado. Italicized values were
estimated using linear interpolation
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest including deforestation (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (Ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs/capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs/capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
1980
38469
158449.0558
3.385607069
811.3289941
8.821093419
22.70737241
0.141507795
2121718.26
0.018131059
4572.30285
207209.4897
468443.1367
882085.2711
559408.0596
50.8
179.82
167.3
73.4
1
92.9
27.6
17.5
271.1
12.52
263.1
33.2
227.2
12.9
608.5
608.5
337.9
270.6
73.41199886
5.366536979
11.08
3.4
1.506196043
60.1
120.2
23,973.14
92,990,880
88,077,314
3,443,739
1,469,828
268,326,225
191,887,425
76,438,800
403,401,978
497,299,806
1981
38980
176724.9259
2.775030315
742.0195447
9.804234462
18.31258213
0.184565695
2121718.26
0.018371902
4646.10285
214383.1902
438884.2499
882085.2711
581719.4459
52.2
179.69
167
74.1
1
91.9
27.1
18.2
264.4
12.69
262.3
33.98
225.7
13.3
587.3
587.3
333.9
253.4
71.47119517
5.419061948
10.77
3.6
1.391510049
60.2
119.8
14,124.07
99,130,808
94,061,842
3,353,902
1,715,064
306,234,575
208,865,375
97,369,200
299,723,304
524,182,716
1982
39467
178864.4571
2.90050471
750.66351
7.335054132
21.27036787
0.188210528
2121718.26
0.018601433
4719.90285
221556.8907
409325.3631
882085.2711
604030.8322
54.4
174.6
162.1
73.9
1.1
87.1
26.4
19.9
264.1
12.5
262.8
32.32
227.7
13.8
615.1
615.1
336
279.1
71.35533997
6.488707427
10.64
3.7
1.610763691
60.8
117.7
24,001.56
96,894,691
92,394,147
3,000,000
1,500,544
312,838,180
227,228,900
85,609,280
295,581,960
581,850,528
1983
¥0772
166457.3616
2.739301399
767.7980211
9.574754423
23.94968653
0.200967439
2121718.26
0.018905432
4793.70285
225385.0547
402236.8031
882085.2711
607217.4282
55.5
175.5
162.1
75.5
1.1
90.1
26.3
20.6
260.2
13.4
264
32.17
229
14.7
603.4
603.4
339.3
264.1
74.32447044
6.526480216
10.84
4
1.557785299
62.8
119.3
25,115.40
102,167,967
99,536,388
2,631,579
-
347,138,625
221,039,425
126,099,200
304,422,567
631,668,984
-------
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest land (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (Ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs/capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs/capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
1984
40490
188272.4105
2.937764886
767.9700599
0.702508994
27.70440016
0.210752698
2121718.26
0.019083589
4867.50285
229213.2187
395148.2431
882085.2711
610404.0242
56.4
180.9
166.7
75.3
1.1
91.2
26.4
21.5
260.1
14.2
268.7
34.45
231.1
16
626.5
626.5
356.5
270
75.20384805
5.101417896
10.96
4.3
1.709343849
64.6
121.8
30,777.54
106,428,675
103,965,880
2,462,795
-
356,523,415
247,711,175
108,812,240
342,193,593
781,827,618
1985
40369
191747.1074
2.755240242
759.9430665
0.402592181
29.65343036
0.212618996
2121718.26
0.01902656
4941.30285
233041.3827
388059.6831
882085.2711
613590.6202
59.9
185
169.9
76.1
1.1
93.2
26.7
22.5
255.2
15.1
283.6
39.34
241.1
17.2
631.6
631.6
359.6
272
74.2373237
7.117753361
11.25
4.6
1.572551223
68.2
126.2
30,784.99
102,649,819
100,652,087
1,997,731
483,574,415
240,060,065
243,514,350
114,993,440
441,830,724
1986
40804
189421.8273
2.495543403
774.1604165
0.655395237
28.53490033
0.252075091
2121718.26
0.019231583
5015.10285
236869.5467
380971.1231
882085.2711
616777.2162
58.4
184.8
169.3
76
1
92.3
26.5
23.1
253.5
15.5
289.5
43.83
242.3
19.4
637.4
637.4
360
277.4
77.13000354
6.600845207
11.26
4.8
1.678779644
68.6
124.3
34,714.50
-
104,776,134
102,909,528
1,866,606
449,167,880
211,729,155
237,438,725
84,188,080
437,380,938
1987
¥77 04
189189.4125
2.647416314
774.6466201
2.193573517
22.91906123
0.166409025
2121718.26
0.019372977
5088.90285
240697.7122
373882.5899
882085.2711
619963.7839
47.4
184.58
168.38
70.8
0.98
96.6
26.1
24.1
253.8
16.2
301.2
48.77
249
21.7
641.1
641.1
366
275.1
76.89765767
5.157145681
10.94
4.8
1.301461261
66.9
128.8
32,798.34
104790311.1
104582695.1
207616.0228
-
313,170,000
161,090,200
152,079,800
84,673,600
511,067,769
-------
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest land (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (Ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs/capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs/capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
1988
41289
200731.8624
2.885678728
784.1199118
8.353280528
22.08556746
0.178456448
2121718.26
0.019460171
5162.70285
241324.6522
375721.2299
882085.2711
617424.4039
47.5
186.7
171.5
69.7
1
100.8
26.1
23.7
246.6
15.2
307.5
50.12
254.2
21.7
644.3
644.3
365.2
279.1
76.16900377
6.935747548
10.54
4.8
1.29784794
67.7
130.2
33,139.45
104531691.7
104389255.9
142435.831
-
308,737,340
163,127,120
145,610,220
88,291,200
513,245,030
1989
40903
204688.8725
3.084130514
743.9996627
6.605283571
21.87609984
0.141095849
2121718.26
0.019278243
5236. 50285
241951.5922
377559.8699
882085.2711
614885.0239
37.3
184.69
169.09
66.09
1
102
26
23.8
237
15.6
304.7
51.35
250.3
21.8
641.8
641.8
380.1
261.6
78.46913146
5.710816178
10.83
4.9
1.26545431
64.7
128.5
35,251.90
104273072.3
104195816.7
77255.6391
378,839,230
206,442,855
172,396,375
121,690,080
400,367,411
1990
40682
210345.8029
3.037972946
741.978769
0
17.46526957
0.134064393
2121718.26
0.019174082
5310
242578.5322
379398.5099
882085.2711
612345.9467
63.8
183.39
168.39
64.79
1
102.6
25.7
24.6
234.1
15
316.9
53.5
260.4
21.4
659.3
659.3
385.1
274.2
74.99830509
6.716221499
11.03
5.4
1.163385732
64.9
132.4
23,871.55
104014452.9
104002377.5
12075.44724
-
370,231,930
185,472,205
184,759,725
65,962,720
452,367,770
1991
41283
202397.8189
3.266446794
748.0894453
0
21.86569235
0.151134105
2121718.26
0.019457343
5446
243205.4722
381237.1499
882085.2711
609744.3667
51.0
184.51
169.62
63.72
1
104.9
25.5
25
232.9
14.89
318.8
53.93
262
21.7
651.2
651.2
395.2
256
80.97654572
7.347342826
11.09
5.7
1.266307156
67.1
132.9
23,286.68
103808938.3
103808938.3
356,537,105
175,487,080
181,050,025
61,183,680
444,429,827
-------
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest land (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (Ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs/capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs/capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
1992
¥7720
200910.4621
3.05817693
740.125866
0
17.63156623
0.135265518
2121718.26
0.019380519
5582
243832.4072
383075.8336
882085.2711
607142.748
49.2
188.31
173.72
63.23
0.99
109.5
25.1
25.8
233.6
14.59
322.1
56.61
262.9
22.1
677.5
677.5
394.2
283.4
78.46850247
7.816088605
10.86
5.7
1.428604794
69.2
136.1
25,041.38
103615499.1
103615499.1
312,335,335
158,970,960
153,364,375
79,350,560
479,678,842
1993
41 466
200253.9844
3.350570905
778.5602277
0
21.9393547
0.127367593
2121718.26
0.019543594
5718
241884.4072
382653.3536
882085.2711
609377.228
53.8
188.01
173.21
61.75
0.96
110.5
24.4
26
233.7
14.8
331.5
57.63
270.8
23.1
688
688
406.7
281.2
82.22013295
7.230088179
9.89
5.3
1.732915794
71.6
139.1
26,320.28
103422059.9
103422059.9
-
-
263,941,090
132,075,390
131,865,700
59,811,440
529,211,650
1994
42564
204375.7016
3.188727254
790.3129064
0
17.64657663
0.121148777
2121718.26
0.020061099
5854
239936.4072
382230.8736
882085.2711
611611.708
53.7
190.99
176.11
63.65
0.86
111.6
24.3
26.6
235.1
14.88
333.7
57.99
272.2
24
694.8
694.8
412.3
282.6
82.47743694
7.707390748
8.97
4.8
1.617091058
69.4
141.6
36,621.15
103228620.7
103228620.7
270,419,775
136,866,050
133,553,725
74,820,400
556,813,021
1995
43793
198328.0663
3.318819344
811.3139992
0
24.38760008
0.12245299
2121718.26
0.020640346
5990
237988.4072
381808.3936
882085.2711
613846.188
52.6
190.05
175.67
64.32
0.85
110.5
23.9
26.9
232.3
14.38
329.3
58.38
266.9
24.9
690.9
690.9
406.2
284.7
82.41809971
7.507275821
8.74
4.5
1.627044122
67.2
144.1
21,581.56
103035181.5
103035181.5
339,636,530
165,980,305
173,656,225
55,621,280
523,813,843
-------
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest land (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (Ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs'capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs/capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
1996
44566
1999345087
3.581222009
839.2337976
0
20.69150494
0.12392165
2121718.26
0.021004674
6726'
236040.4072
381385.9136
882085.2711
616080.668
46.2
188.56
174.07
64.96
0.81
108.3
23.8
27.3
233.4
14.49
342.9
59
279.3
25.9
703.3
703.3
416.9
286.4
85.54335133
7.432451946
9.51
5.2
1.337575603
65.6
144.4
11,056.57
102841742.3
102841742.3
-
-
375.290,000
200.765,300
174.524,700
119.122,400
461.094,984
1997
45289
206220.1121
3.550272466
789.8510935
0
22.90207509
0.132994334
2121718.26
0.021345435
6262
234092.0285
380963.3746
882085.2711
618315.5858
60.3
186.81
172.54
63.44
0.8
108.3
23.4
27.5
234.3
14.27
343.5
57.25
281.3
26.5
709.7
709.7
415.1
294.6
81.4107545
7.394777992
9.87
5
1.089728557
65.1
147.7
3,871.70
102648303.1
102648303.1
383,370,840
204,269,240
179,101,600
97,997,520
668,035,017
1998
45902
205489.9538
3.679618252
835.9981164
0
16.30182389
0.118193798
2121718.26
0.021634352
6398
236491.2485
375949.7946
882085.2711
620793.9458
50.7
192.19
177.62
64.26
0.86
112.5
23
27.8
239.2
14.57
337.4
57.39
274.6
27.2
696.2
696.2
411
285.2
80.90923623
7.261340726
10.18
5.4
1.320826467
65.5
148.9
4,352.81
102454863.9
102454863.9
-
-
367,873.510
191,675.035
176,198.475
81,382.400
641,613.001
1999
46377
203823.0838
3.654395462
853.5482589
0
17.14163877
0.124008014
2121718.26
0.021858227
6534
238890.4685
370936.2146
882085.2711
623272.3058
51.9
197.17
182.42
64.88
0.84
116.7
22.9
29
249.7
14.75
339.9
56.93
277.3
27.8
705.2
705.2
414.3
291
81.77681111
7.805818174
10.64
5.6
1.544245332
68.2
151.2
6,737.07
1 02261 424. 7
102261424.7
-
-
354.264.550
186.655.875
167.608.675
96.247.200
680,508.942
-------
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest land (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (Ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs'capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs'capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
2000
47074
213776.4764
4.003141834
908.0605101
0
15.81676148
0.125098182
2121718.26
0.022186735
6670
241289.6885
365922.6346
882085.2711
625750.6658
46.7
196.79
181.58
65.05
0.83
115.7
22.5
29.8
251
15.21
345.8
59.61
280
28.4
711.2
711.2
422
289.2
81.17866908
7.641999969
11.14
5.9
1.499920327
84.3
148.8
5,506.58
102067985.5
102067985.5
468,344,800
236,823,800
231,521,000
87,971,600
679,397,630
2001
46863
215523.3959
4.910043486
919.4162247
0
15.82007546
0.071957534
2121718.26
0.022087287
6806
243688.9085
360909.0546
882085.2711
628229.0258
46.9
193.94
179.14
63.59
0.85
114.7
22
30
252.4
14.8
336
60.29
269
29
684.2
684.2
411.9
272.3
81.14909138
6.939987162
10.37
5.6
1.49992465
85.3
147
3,223.76
90,209,619
90,209,619
340,367,365
178,770,515
161,596,850
75,844,480
789,961,911
2002
47304
208941.3594
4.824877192
875.4892085
0
12.70866345
0.067274976
2121718.26
0.022295137
6942
246088.125
355895.5272
882085.2711
630707.3367
30.6
200.49
184.85
64.98
0.87
119
21.9
30.5
254.6
15.64
328.8
61.18
260.6
29.7
684.9
684.9
411.1
273.8
77.07850497
6.744101795
10.03
4.8
1.499942751
90.1
146.1
2,679.38
79,211,661
79.211,661
254,515,900
121,562,200
132,953,700
49,007,600
561,666,515
2003
47376
210025.8599
4.543803648
903.9772299
0
13.15201882
0.069824506
2121718.26
0.022329072
7078
248487.345
350881.9472
882085.2711
633185.6967
43.6
199.04
182.81
62.38
0.83
119.6
21.6
30.5
254.3
16.23
329.9
61.09
261.4
30.3
702
702
420.1
281.8
80.8407184
6.639901043
10.3
5.3
1.499793877
89.2
141.3
3,897.25
64,341,425
64,341,425
-
259,156,120
118,279,070
140,877,050
31,955,920
567,427,189
-------
Variable name
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Nuclear Electric Consumption (million kWh)
Hydro-electric Consumption (million kWh)
Wood and waste Consumption (trillion BTU)
Land Area (hectares)
Population density (individuals per hectare)
Built-up land (ha)
Arable land (ha)
Pasture (ha)
Forest land (ha)
everything else (ha)
mean precipitation (cm)
Meat Consumption (ibs/capita)
meat (fresh; Ibs/capita)
bovine, buffalo (Ibs/capita)
sheep, goat (Ibs/capita)
non-bovine (Ibs/capita)
milk (gal/capita)
cheese (Ibs/capita)
eggs (Ibs/capita)
fish (Ibs/capita)
Cereal Consumption (Ibs/capita)
cereals (Ibs/capita)
wheat (Ibs/capita)
maize (Ibs/capita)
Fruit & Vegetable Consumption (Ibs/capita)
vegetables & fruits (Ibs/capita)
vegetables (Ibs/capita)
fruit (Ibs/capita)
roots and tubers (Ibs/capita)
pulses (Ibs/capita)
coffee and tea (Ibs/capita)
cocoa (Ibs/capita)
oil seed (Ibs/capita)
fats (Ibs/capita)
sweetener consumption (Ibs/capita)
forest harvest (MBF)
meat production (fresh; kg)
bovine, buffalo production (kg)
sheep, goat (kg)
non-bovine, goat, mutton, buffalo (pork; kg)
grain production (kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
2004
47831
205594.4562
4.576624544
974.5921966
0
12.4296962
0.076970504
2121718.26
0.022543521
72 14
250886.565
345868.3672
882085.2711
635664.0567
49.1
201.45
184.85
63.31
0.83
120.4
21.2
31.2
256
16.6
326.6
60.58
258.1
30.9
693.8
693.8
421.7
272.1
79.35153158
5.966688317
10.45
6
1.488811125
88.4
141.6
6,280.86
69.523,593
69.523,593
265,175,950
114,660,030
150,515,920
27,140,160
569,717,854
2005
47530
198190.5176
4.790411072
940.204808
0
14.42219504
0.070327311
2121718.26
0.022401655
7350
253285. 785
340854. 7872
882085.2711
638142.4167
47.0
199.76
183.56
62.78
0.78
120
21
31.4
253.9
16.2
326.3
61.89
256.4
31.4
688.6
688.6
415.4
273.2
74.98827017
6.148759756
10.35
6.5
1.478475521
87.2
141.6
4.444.48
-
68,018,376
68,018,376
-
241,908,150
110,377,850
131,530,300
31,252,800
655,502,132
-------
Appendix 3-D: Footprint
for the
1980-2005
This appendix provides results of an Ecological
Footprint Analysis (EFA) for the United States from
1980 to 2005 for comparative purposes with the San
Luis Basin EFA.
Because EFAs usually reveal patterns for the US where
ecological deficit (ef> be or EF > BC) is the rule, as
demonstrated by the analysis on Utah (Mclntyre et al.
2007), and the analysis for the SLB reveals an ecological
reserve, we decided to conduct an EFA for the US to
test the methodology employed for SLB, using the
same variables and data sources (see Table 3.1). For
comparison purposes, the Global Footprint Network
(2009) graph for the US shows a general increase in ef
for the US from 1980 (~6 gha) to 2005 (~5 gha) and a
general decrease in be from 1980 (~7.5 gha) to 2005
(~9 gha). These results indicate the ecological balance
for the US is a deficit and US was moving away from
sustainability for the period (an increasing ecological
deficit from ~1.5 to 4 gha). Our methodology reveals
similar trends for the US, revealing the US had an
ecological deficit because EF had already exceeded BC
by 1980 (whether reported ha or gha or per capita or for
the entire population; Figs. 3-C.l and 3-C.2). EFAs for
the US usually identify an ecological deficit occurring
around 1970 (e.g.. Global Footprint Network 2009). The
differences in the results between U S and SLB increased
our confidence that our methodology captured trends in
the SLB, rather than merely duplicating that of the larger
svstem.
References
Global Footprint Network. 2009. Footprint for Nations.
http://www.footprintnetwork.org/en/index.php/GFN/
page/trends/us/. Last accessed July 27, 2009.
Mclntyre, S., Peters, H.M.. Christcnsen, M.H., Enimi,
P.C., Martinson, W., Mielke, M., SenbeL M., Stark,
D.O. 2007. The ecological footprint of Utah: a
sustainability indicators project of the Utah Population
and Environment Coalition. Utah Vital Signs, 26 June.
59 pp. Available online at http://www.utahpop.org/
vitalsigns/research/report_2007 htm. Last accessed
July 13, 2009.
-------
•ef
•be
6.00
0.00
1980 1983 1986 1989 1992 1995 1998 2001 2004
year
Figure 3-D. 1 - Biocapacity (be) compared to ecological footprint (ef) per capita for 26 years in the US. Twelve
percent of the biocapacity area was subtracted from the total be to protect biodiversity.
to 1
I
O
O
O
,800,000
,600,000
,400,000 -
,200,000 -
,000,000 -
800,000
600,000
400,000
200,000
•EF
•BC
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
year
Figure 3-D.2 - Total ecological footprint (EF) and biocapacity (BC) for the US population. Twelve percent of the
biocapacity area was subtracted from the total BCto protect biodiversity.
-------
4.0
Green Net Regional Product
4.1 Introduction
This chapter describes the data sources and methodology
used to estimate Green Net Regional Product (GNRP)
for the San Luis Basin (SLB). GNRP is similar to Green
Net Domestic Product (GNDP; defined in Chapter 2),
but applied to a region on a smaller spatial scale. The
boundary of this metric was delineated by the seven
counties that we define as the SLB (Fig. 1.1). The
chapter is outlined as follows. First, we describe the
basic model of estimating GNRP and the components
of the depreciation of natural capital. Next, we
highlight the public data sources used to estimate Net
Regional Product (NRP), and the three components
of natural capital depreciation. Then we present the
methodology used to estimate GNRP, followed by the
estimates themselves. Finally, we point out some of our
assumptions when calculating GNRP with public data
sources and our conclusion.
4.2 Model
Using the information describing GNDP in Section 2.2.2,
the model for the SLB GNRP is
GNRP(t) = NRP(t) + I „ (t)K"(t) (4.1)
Net Regional Product (NRP) is equal to Gross Regional
Product (GRP) minus the depreciation of human-made
capital. It does not include the depreciation of natural
capital, which is represented by the second term on the
right of Equation 4.1, where In(t) is the shadow price of
natural capital K"(f) and is the change in natural capital
over time, t. This component still has to be estimated in
order to estimate GNRP.
Natural capital, when incorporated in NRP calculations,
is typically included as the extraction of natural resource
stocks. For example, a resource stock that might be
important is forests. Although timber is considered a
valued output, GNRP captures the depreciation of natural
capital through the value of decreases in the forest stock
(Seroa da Motta and Ferraz do Amaral 2000). Given the
contribution of agribusiness to the SLB, we included the
depletion of both groundwater and soil as components
in the depreciation of natural capital. In addition, we
captured the effect of the consumption of energy on
future generations through carbon dioxide emissions
(CO2). When stocks of capital are not changing over
time or the change is minimal, an advantage of using
the change in GNRP is that those stocks of capital
can be dropped as explained in Chapter 2. Based on
our assumptions, mining activities and forestry were
considered constant. The assumption for forestry was
consistent with Ecological Footprint Analysis. The
decision to consider mining constant could be debated
given the export data used in Emergy Analysis (Chapter
5; e.g., sand and gravel was decreasing, broken stone
or rip rap was marginally decreasing). We did not
explicitly include the depreciation of human capital in
our estimate of GNRP. Aggregate consumption in NRP
(Eq. 4.1) already incorporates education expenditures.
If these expenditures are the best proxy for investment
in human capital, the expenditures are already included
in the equation (see Pezzey et al. 2006 for discussion).
We are aware, however, that education expenditures
are typically considered a poor measurement of the
depreciation of human capital (e.g., Ferreira and Vincent
2005). Because sustainability relates to the potential
for future increases or decreases in welfare rather than
any absolute level of welfare, the absolute estimates of
GNRP do not relate to sustainability. Sustainability is
measured by changes in GNRP over time.
4.3 Data and Sources
One approach to calculate NRP, Equation 4.1, would
begin with collecting data to estimate GRP and then
adjust GRP by the depreciation of man-made capital
stock (i.e., top-down approach). This nominal estimate
is then converted into real NRP using the Gross
Domestic Product (GDP) deflator (Banzhaf and Smith
2002, Landefeld et al. 2008). Unfortunately, the lack of
economic data for the counties in the SLB hampers the
implementation of this approach.
The data we used to estimate real NRP, described
in the next section, are available from the Bureau of
Economic Analysis (BEA) in the US Department of
Commerce. The BEA is responsible for calculating
the economic accounts in order to understand how the
economy is performing (BEA 2006a). This Agency
provides economic accounts data for the US, states,
and regions, including counties (BEA 2006a). Our
approach for estimating NRP included data from all
three scales. The economic measures required include
GDP and Consumption of Fixed Capital (CFC) for the
US. Personal Income and GDP for Colorado and New
Mexico were needed and available from the BEA. For
the seven counties in the SLB, we used Personal Income
from the BEA (Table 4.1).
-------
For each component of the depreciation of natural
capital, we needed an estimate of the change in the
stock of the capital and a "'price." Goods and services
provided by the environment and not traded in markets
are not considered in the definition of natural capital
depreciation above. A number of articles, however,
have been published on how to proceed with calculating
GNRP when some of the goods and services cannot be
valued with existing prices (e.g., Banzhaf and Smith
2002, Cairns 2001). Banzhaf and Smith (2002) stated
that incorporating the depreciation of nomnarket goods
and sendees is difficult because of the lack of data and
because these goods and sen-ices are not measured in
standard, well-defined units like market goods.
Cairns (2001) suggested using shadow prices of the
local amenities that do not have market prices. For
environmental spillovers or externalities that extend
beyond the local economy (e.g., reducing CO,
emissions), he recognized the global shadow value
overstates the benefits to the local economy. However,
he pointed out the inadequacy (for green accounting
purposes) of imposing local value (as measured by
willingness to pay) as the price of benefits that extend
globally. His proposed solution was to count the
difference between the shadow value and the domestic
willingness to pay for mitigation of the spillover
pollution as an export from the pollution emitter to
the recipient of pollution. He provided a theoretical
approach for incorporating these nonmarket values, but
did not implement his approach.
In the case of carbon emissions, we have opted to
follow the example of Hamilton and Clemens (1999)
who assumed the property right to a clean environment
lies with the victim of pollution and thus charge global
damages to emitting countries.8 Carbon emissions are
counted as an environmental cost equal to the global
social cost of carbon.
Next, we describe the data needed to calculate the
change in the stock of groundwater, soil erosion, and
CO2 emissions and the sources of these data. Additional
details of the methodology' are described in Section
4.4. The Colorado Water Conservation Board and the
Colorado Division of Water Resources have created
Colorado's Decision Support Systems that provide
information about the water balance of many of its river
basins (CWCB and CDWR 2008). Its purpose is to
help water managers make better decisions related to
water use. The Upper Rio Grande is one of the basins
that is available. Colorado Division of Water Resources
provided the updated calculations for the groundwater
basin in the Upper Rio Grande and data required to
include the years 1970-2005 (personal communication
with R. Bennett, February 2008).
Wind erosion is the main cause of soil erosion in the
SLB (personal communication, R. Sparks, USD A,
May 2008). Finding reliable, county-level soil-erosion
data is challenging. No estimates have been attempted
for the SLB. The National Resource Information
(NRI) data on wind erosion at first appeared to be a
promising data source (e.g., NRCS 2007); however,
NRI data are not reliable at the county level (personal
communication, R. Sparks, USD A, May 2008).
Following the recommended approach described by
Richard Sparks, an agronomist in the USD A Natural
Resources Conservation Sereice (NRCS) stationed in the
region (see Appendix 4-A), a number of data sets were
required. First, the Agricultural Statistics for Colorado
provides annual estimates of planted acres of different
crops by county. The US Census of Agriculture (NASS
1984, 1989, 1994, 2009a, 2009b) provides data on total
crop land and total harvested land. We used reports
beginning in 1982 and continuing every five years until
2002. Through numerous communications. R. Sparks
provided information about crop rotations, soil type, and
residue management by year (personal communication,
May-November 2008). The Wind Erosion Equation
Technology from 1980-2005 (Woodruff and Siddoway
1965) was then applied to this calculation of acreage by
crop rotation, soil type, and residue management.
Carbon dioxide emission estimates do not exist at the
county level. Therefore, we needed to find an alternative
approach. Required data included state-level data
for Colorado and New Mexico and the population of
each state for the same years (1980-2005). State-level
emissions can be found at the Energy Information
Agency (EIA 2007). In addition, we needed the SLB
population for the same years. Population data were
acquired through the Bureau of Economic Analysis
(BEA 2007). We used the average per capita CO2
emissions for Colorado and New Mexico (EIA 2008) and
applied this to the SLB population estimates. Table 4.1
presents the sources and the Websites where these data
can be accessed.
8 Although we focus on determining whether the region is moving to-
wards or away from sustainability, another possibility would have been
to assess the region's contribution towards global sustainability. If this
would have been our objective, global emissions of carbon would have
been the correct estimate. The approach used in this report is similar to
that of Pezzey et al. (2006).
-------
4.4 Calculation Methodology
Due to the lack of data for the top-down approach,
the alternative, bottom-up approach adjusts Personal
Income, which is available for all counties in the SLB
(e.g., BEA [2008] presents the relationship between
Personal Income and GDP). Personal Income is defined
as(BEA2007:XI-6):
income received by persons from
participation in production, plus transfer
receipts from government and business, plus
government interest (which is treated like a
transfer receipt).
Personal Income for the region was converted to
Net Regional Product as follows. We estimated Net
Domestic Product by state (NDPstate) using Gross
Domestic Product by state (GDPstote) and the ratio of
Consumption of Fixed Capital (CFCUS) to GDP for
the US (GDPUS) (4.2). This was proposed because
BEA does not estimate CFC for states (personal
communication C. Woodruff, 25 January 2008). It is
somewhat similar to the approach used by Pezzey et al.
(2006) where United Kingdom estimates were used to
convert Scottish data. Once NDPstate was calculated, we
estimated the ratio of NDPsMe to Personal Income by
state. We used this process for both Colorado and New
Mexico (i.e., represented by subscript state in Equations
4.2 and 4.3) across time and multiplied by the Personal
Income for the counties in the SLB, as presented in
Equation 4.3. The estimate of NRP for the SLB was
the average of both the Colorado and New Mexico
calculations because we believed the GDPstate and
Personal Income for Colorado and New Mexico bound
the calculations for the SLB.
\GDP«.
NRP i
,te
(4.2)
(4.3)
Without data to support this approach, the alternative
would have been to use the ratio of Net Domestic
Product for the US to Personal Income for the US,
but we do not believe the Net Domestic Product is
representative of the economy of the SLB. We are
aware of another complication described by BEA related
to GDP by state time series. Because BEA switched
from Standard Industrial Classification (SIC) to North
American Industry Classification System (NAICS)-based
GDP in 1997, the time series results in a discontinuity.
A SIC-based GDP is closer to Gross Domestic Income
(GDI) and a NAICS-based GDP is closer to US GDP
(BEA 2006b). Although BEA strongly cautions against
combining these two series to have a complete time
series, we appended the two by adjusting the SIC-based
GDP by state (prior to 1997) using the ratio GDPUS/
GDIUS to make years relatively consistent with post-
1997. The estimate for real NRP, which has been
adjusted using the GDP deflator, is in the second column
of Table 4.2.
As stated in the previous section, we used data to
estimate the change in the stock of natural capital and
existing studies to estimate the shadow price. The
previous section described the change in the stock of
natural capital and data needed to estimate NRP. We
provide more detail here on how we combined shadow
price and the change in natural capital stock. In order to
estimate the value of the change in the natural capital, we
used benefit transfer, an economic valuation approach.
An alternative to collecting primary economic value
data via observation or survey is to use benefit transfer
based on secondary data. Benefit transfer applies
economic value (e.g., willingness to pay for a change
in environmental quality) from one study to another
location or context (Desvousges et al. 1992). The
accuracy of benefit transfer depends on the existence and
quality of applicable studies. The advantages of benefit
transfer approaches include saving the time and cost of
developing and implementing new studies. This is a
standard approach for the USEPA because of limitations
on primary data collection using surveys (see The Paper
Reduction Act of 1995). Of course, the disadvantages
are obvious. Benefit transfer is not usually as accurate
as analysis done with primary data. This means the
values estimated for the policy site will depend on
differences in the resources being valued and the relevant
populations.
According to Wilson and Hoehn (2006: 336), "few
benefit transfer practitioners seem fully satisfied with
the state of the science." However, certain steps need
to be taken to ensure a high quality transfer. A few
authors have listed these steps. For example, the USEPA
(2000: 86-87) describes the steps as: 1) understanding
the characteristics and consequences of the policy
case along with the effected population; 2) identifying
existing studies or values that are relevant for the policy
case through a literature search; 3) reviewing available
studies for quality and applicability and to determine
their transferability; 4) transferring the benefit estimates;
and 5) addressing uncertainty because judgments and
assumptions are involved.
We used a water balance model from the Colorado's
Decision Support Systems for the depletion of
groundwater. This model was used to estimate the water
-------
balance for many of Colorado's basins, including the Rio
Grande River Basin (CWCB and CDWR 2008). Over
the 26 years of this study, the model showed, on average,
that ground-water storage has been declining in the region
(outflow has been greater than inflow; Table 4.2, column
3). We used the shadow price of water for agriculture
from Kurd and Coonrod (2007), which was based on a
hydro-economic model for the Rio Grande watershed.
The purpose of their study was to examine the impact of
climate change on water supplies in the arid southwest.
We used $76/acre-foot (2005$) from their baseline
estimate in 2000. It is not reflective of the market price
of water, but rather the marginal value of an additional
acre-foot of water for agricultural purposes. The
authors pointed out that this estimate was relatively high
compared to the estimate for New Mexico because of
the Rio Grande Compact (CWCB 2000), which requires
Colorado to deliver a certain amount of water to New
Mexico and Texas, and because of the high marginal
value of agricultural production in the SLB (third and
Coonrod 2007). The depreciation of groundwater is
presented in column 4 of Table 4.2.
Soil erosion in the SLB is primarily from wind, and it
had not been estimated over time prior to this study.
To produce a credible estimate, we needed data on soil
credibility, crop rotations, tillage management, crop
growth stages, and winterkill. Because agriculture is not
a major activity in all of the SLB, we focus on Alamosa,
Conejos, Costilla, Rio Grande, and Saguachc counties.
We collaborated with R. Sparks, who is familiar with the
farming practices in the SLB (more details of this effort
are described in Appendix 4-A). The estimate for wind
erosion, in tons/year, is in column 5 of Table 4.2. The
damage estimate was based on Huszar and Piper (1986)
who calculated off-site damages due to wind erosion in
New Mexico. Ringquist et al. (1995) revised the Huszar
and Piper estimate to focus on agricultural sources only.
The value we used was $2.42/ton (2005$) and the total
estimate is in column 6 of Table 4.2. Therefore, rather
than capturing the quality-adjusted stock of agricultural
land as described by Hrubovcak et al. (2000), we present
the off-site damages caused by wind erosion. Other
papers have derived estimates from Huszar and Piper
(1986). For example, Bunn (1998) used one particular
Major Land Resource Area (MLRA) that is part of the
study area and determined the off-site cost per ton of
eroded soil per acre was approximately $10.12 (2005$).
Williams and Young (1999) used Huszar and Piper to
estimate off-site costs in South Australia. Additional
studies, such as a hedonic analysis, would be necessary
in order to determine the foregone production rents due
to the soil erosion on farmland values (Hrubovcak et
al. 2000). We are currently attempting to determine the
feasibility of such a study in the SLB.
Per capita CO2 emissions for the SLB were calculated
as the average of the EIA estimates for Colorado and
New Mexico. This per capita emission estimate was
multiplied by the population of the SLB to calculate total
emissions (Table 4.2, column 7). Although we estimated
carbon flows (i.e. emissions) to estimate the annual
social cost of carbon emissions, global temperature is
influenced by the absolute concentration of CO2 in the
atmosphere. CO2 emissions remain in the atmosphere
for many years. Pezzey et al. (2006) noted the social
cost of carbon emission estimates should incorporate the
atmospheric lifetime of emissions. Following Tol (2007)
who calculated the social cost of carbon based on a
meta-analysis of 125 peer-reviewed estimates, we used a
value approximately equal to $24/ton CO2 (2005$). The
damage cost caused by these emissions is presented in
column 8 of Table 4.2. Lifetime costs of emissions were
included in this estimate.
Because we did not know the right prices and incorrectly
estimating the prices could lead to a misinterpretation
of the results, we conducted a sensitivity analysis. We
calculated low and high estimates of the prices and
marginal damages of natural capital depreciation. For
the depreciation of groundwater, the low value was
$66.05/acre-foot and the high value was approximately
$85/acre-foot. These values were based on the
variability of the 30 years of simulation (personal
communication B. Hurd, May 2010). The low and
high estimates of damages caused by wind erosion
were $0.27/tonand $9.57/ton, respectively (Huszar
and Piper 1986). These costs were based on specific
regions (MLRA 48/51 and 77) rather than for the
entire state. The social cost of carbon ranged from $07
ton CO, to $77.29/ton CO2. Tol (2007) calculated the
95-percentile for peer-reviewed studies. The estimate
means the author is 95% confident the expected social
cost of carbon is equal to or less than $77.29/ton CO2.
Tol (2007) did not provide a lower bound estimate, so we
chose $0.
4.5
Following the methodology presented in the previous
section and using the 26 years of data consisting of the
variables described in Table 4.1, we estimated the SLB
GNRP. Aggregating the results from NRP, Groundwater
Storage Depreciation, Soil Erosion Damage, and CO,
Damage Costs, we estimated GNRP in column 9 of
Table 4.2. The final column shows the change in GNRP
from one year to the next in between the rows. The
-------
NRP, GNRP, and the change in GNRP are all presented
in Figure 4.1; components of natural capital depreciation
are presented in Figure 4.2.
Although there were peaks and troughs, NRP and GNRP
both had upward trends suggesting the components
of natural capital depreciation have a minimal effect.
Natural capital depreciation ranges from -8.6 to 0.11
percent of NRP. However, caution should be taken
in interpreting the actual values in those two columns
because we dropped the components that do not vary.
Through the 26 years of data, CO2 damages and soil
erosion changes have remained fairly constant; however,
groundwater storage fluctuated over this time (Fig.
4.2). Overall, it appears that natural capital stocks have
declined.
If we strictly follow Pezzey et al. (2006) and Hanley
et al. (2002), originally summarized in Chapter 2, who
suggested the change in GNRP is a one-sided test of
weak sustainability (e.g., when man-made capital can be
substituted for natural capital), the calculation revealed
the SLB was moving away from sustainability during the
time period 1985-86 to 1987-88. This appears to be the
longest negative range. Based on the results, there are
other periods of negative change, but not since 1999-
2000. As we discuss below, this strict interpretation,
based on the theoretical literature, may have some
problems.
The sensitivity analysis, when using the lower bound
estimates, revealed that GNRP is about 3% to 6% higher
than our GNRP estimate in Table 4.2 (see Fig. 4.3).
When we calculated GNRP with the high values, our
estimates ranged from 7% to 14% lower than GNRP in
Table 4.2.
4.6 Discussion
The previous section presented the results of estimating
GNRP for the SLB over a 26-year period. The
methodology satisfies the objective of estimating
GNRP using an approach that others can follow using
publicly available data. With that said, the data were
not perfect. This section discusses the assumptions
needed in order to estimate GNRP of the SLB and some
of the limitations we found while developing the GNRP
methodology.
Before we present some of the specific assumptions
for our calculations, we want to state that there is
uncertainty surrounding the application of this method.
Although the calculation is theoretically founded, the
approach requires many assumptions (e.g., Stern 1997).
Although these assumptions are very important, we
briefly mention them because they can be found in the
literature (e.g., Dasgupta 2001, Neumayer 2010, Stern
1997). Assumptions related to weak sustainability and
using the right prices (i.e., shadow prices) are important
and do lead to uncertainty in the calculations. It is not
our intent to minimize this first group of assumptions
underlying the theory, but this section focuses on those
assumptions related to our specific calculations and
interpretation.
Calculating the NRP of the SLB required a number of
simplifying assumptions. We already stated the type of
data necessary to use a top-down approach would have
been prohibitive; therefore, we used US and state data to
convert Personal Income in the SLB to NRP. To do so,
we assumed the ratio of US GDP to US CFC was similar
to that in both Colorado and New Mexico. Taking the
Net Domestic Product by state and estimating the ratio to
Personal Income to the state required another assumption
of similarity. If neither of these are sound assumptions,
we may have over or underestimated NRP in the SLB.
Because of a change in the BEA approach to estimating
state GDP, we assumed the ratio of US GDI to US GDP
is similar to that of Colorado and New Mexico prior
to 1997. This is necessary because of the switch from
the SIC to the NAICS-based GDP. The BEA cautions
against completing the time series by combining the two;
it is not clear that adjusting the SIC-based GDP with the
US ratio is acceptable either. For studies that may follow
this methodology, estimating GNRP after 1997 avoids
the need for this assumption and transformation.
Based on existing literature, preliminary examination
of data, and the methodology for the other metrics,
human capital and certain components of natural capital
depreciation were assumed constant (e.g., forestry). The
decision to do this removed the need to include those
data in the calculations. Although we were interested
in the change in GNRP, no additional information can
be gleaned from the calculations because the annual
calculations do not capture these particular components.
Business cycles complicate the use of GNRP to analyze
sustainability. Sustainability is concerned with the
long-term and thus short-term business cycles may
obscure the interpretation of the metric. The theoretical
literature on GNRP assumes the economy is achieving
its potential output. GNRP calculated with actual data
reflects recessions that prevent GNRP from reaching its
potential output. Temporary dips in GNRP due to the
business cycle may not represent movement away from
sustainability if potential GNRP is steadily increasing.
BEA (2003) determined that recessions occurred in the
area from 1981-82 (most severe), 1990-91, and in 2001.
GNRP follows the decline in 1981-82, but not 1990-91
-------
(although it slows). There was a fall from 1999 to 2001
in Real NRP, but not GNRP. This introduced some
doubt as to whether the decline in GNRP 1981-1982
indicates movement away from sustainability or simply
reflects the temporary economic downturn. There is no
generally accepted method of controlling for the business
cycle in sustainability analysis, but Pezzey et al. (2006:
71) attempted "an ad hoc allowance for the Scottish
business cycle during 1992-1999."
There is another reason to be cautious of concluding the
region was moving away from sustainability based on
periodic decreases in GNRP interspersed throughout the
overall growth in GNRP during the study period. GNRP
does not explicitly consider technological progress,
but the overall growth or decline in GNRP is largely
determined by a competition between technological
progress that allows GNRP to rise and natural capital
depreciation that pulls down GNRP (Weitzman and
Lofgren 1997). Technological progress does not proceed
smoothly, but rather comes in fits and starts. Because
consumers typically wish to smooth their consumption,
consuming capital (resulting in negative growth in
GNRP) during years of little technology growth is
rational behavior.
Over the 26-year period, it appears that technological
progress was greater than net depreciation for the region,
since GNRP rose over the entire period.9 The years of
declining GNRP when GNRP was moving away from
sustainability should provoke legitimate concern the
region might have been consuming its capital at too rapid
a rate and technology was not keeping up. However, the
ultimate growth in GNRP suggests that isolated declines
in GNRP were only signs the region was moving away
from sustainability, rather than a definitive judgment as
the theoretical literature would suggest.
We used a number of approaches to estimate the change
in natural capital stock for CO2 emissions, soil erosion,
and groundwater. Colorado's Decision Support Systems,
which is used to estimate the change in groundwater
stock, has its own set of assumptions, but given their
approach, we feel the calculation was accurate (e.g.,
CWCB and CDWR 2008). Soil erosion was developed
with R. Sparks and required a number of simplifying
assumptions in order to calculate crop rotations by
soil type and wind erosion based on the Wind Erosion
Equation. More details on the approach can be found in
Appendix 4-A. Finally, we assumed that Colorado and
New Mexico per capita CO2 emissions bound the SLB
per capita CO2 emissions. If the residents of the SLB
had different habits in terms of energy consumption
that differs from the two states, the results may not be
accurate. However, the research for ecological footprint
determined that per capita energy consumption in the
SLB was not significantly different than per capita state
consumption (see Appendix 3-A).
We did not know the right prices for the different capital
stocks and we used existing literature for those estimates.
Because we used benefit transfer to estimate the shadow
price and marginal damage costs for natural capital
depreciation, we need to consider the quality of the
studies, the relevant populations, and the uncertainty of
the results. We have the most confidence in estimating
the depreciation of water because the study of Kurd
and Coonrod (2005) included the Upper Rio Grande
River Basin, which is part of the SLB. They discussed
the value of agriculture and the Rio Grande Compact,
which both play an important role in the shadow price
of groundwater in the SLB. We were fairly confident in
the marginal damage cost of carbon emissions that was
based on Tol (2005, 2007). This confidence was based
on their use of 125 peer-reviewed studies. Nonetheless,
there were many different statistical distributions that
could have been used and many studies in the gray
literature that were not included. The studies used in
the meta-analysis may not be representative, but we
believe it is the best information available. We had the
least confidence in our estimate of damages due to wind
erosion. The study we used was for New Mexico, but it
had a very low response rate for the survey suggesting
that it may not even be representative of New Mexico,
let alone the SLB. We only used off-site costs of
wind erosion and not on-site agricultural damages. In
addition, the results were presented in total dollars
and we revised the estimates to dollars per ton. The
assumptions required to use the value from the Huszar
and Piper study (1986) were quite simplifying.
Natural capital was mostly declining (Fig. 4.2) and, if
the weak sustainability assumption was incorrect or if we
underestimated the prices for natural capital, the results
could reveal a movement away from sustainability.
Based on the sensitivity analysis, we believe other
components, such as NRP, are relatively more important
than the shadow prices and damage costs (see Fig. 4.3).
Because the trends in GNRP did not differ substantially
throughout the range of the sensitivity analysis, the
interpretation remains consistent with the results in Table
4.2. That is, given the results of the weak sustainability
metric, the system appears not to be moving away from
sustainability.
9 Error in our estimate of human capital may lead to a misleading
result, but we think that our interpretation regarding technological
progress was likely correct.
-------
4.7 Conclusion
We estimated GNRP, a metric based on the concept of
weak sustainability, from 1980-2005 for the SLB. The
results show that GNRP had an increasing trend over
the 26-year study period with a few negative periods.
Given this fairly short time horizon, we conclude that
there was no definitive evidence of movement away
from sustainability. However, if the assumptions behind
GNRP, like weak sustainability, are wrong, along with
the decline in natural capital stocks, then the results
could indicate problems in the longer term.
4.8
Banzhaf, H., Smith, V. 2002. Adding environmental
sendees to an economy's scorecard. August, Issue
Brief 02-09. Resources for the Future, Washington,
DC.
BEA (Bureau of Economic Analyses). 2003. Gross
State Product by Industry- for 2001: U.S. Economic
Slowdown was Widespread. News Release. US
Department of Commerce, Washington. DC. Available
online at http://www.bea.gov/newsreleases/regional/
gdp_statc/2003/pdf/gsp0503.pdf. Last accessed June
26, 2009.
BEA (Bureau of Economic Analyses). 2006a. Final
Strategic Plan for FY 2006-FY 2010. US Department
of Commerce, Washington. DC.
BEA (Bureau of Economic Analyses). 2006b. Gross
Domestic Product by State Estimation Methodology.
Regional Economic Accounts, US Department of
Commerce, Washington, DC. October 2006.
BEA (Bureau of Economic Analyses). 2007. Local
Area Personal Income and Employment Methodology.
Regional Economic Accounts, US Department of
Commerce, Washington DC.
BEA (Bureau of Economic Analyses). 2008. National
Income and Products Accounts Tables. Table 1.7.5
Relation of Gross Domestic Product, Gross National
Product, Net National Product. National Income, and
Personal Income. US Department of Commerce,
Washington, DC.
Burin. J. 1998. Government Policy, Wind Erosion, and
Economic Viability in Semi-Arid Agriculture: The
Case of the Southern Texas High Plains. Journal of
Agricultural and Applied Economics 30, 339-351.
Cairns, R. 2001. Seeing the trees as a forest: What
counts in green accounting. Ecological Economics 36,
61-69.
CWCB and CDWR (Colorado Water Conservation
Board and the Colorado Division of Water Resources).
2008. Colorado Decision Support Systems and Rio
Grande Decision Support System. Available online at
http://cdss.state.co.us/DNN/Home/tabid/36/Default.
aspx. Last accessed June 26, 2009.
CWCB (Colorado Water Conservation Board). 2000.
Rio Grande Decision Support System: An Overview.
Section 4.0 Water Administration Issues. CWCB,
State of Colorado. Available online at http://cdss.state.
co.us/DNN/RioGrande/tabid/57/Default.aspx. Last
accessed June 26, 2009.
Dasgupta, P. 2001. Human Well-Being and the Natural
Environment. Oxford University Press, New York.
Desvousges, W., M. Na.ught.on, Parsons. G. 1992.
Benefit transfer: Conceptual problems in estimating
water quality benefits using existing studies. Water
Resources Research 28, 675-683.
El A (Energy Information Administration). 2007. State
Energy Consumption Estimates. 1960-2004. Table 7.
Energy Consumption Estimates by Source, Selected
Years, 1960-2004, Colorado, DOE/EIA-0214(2004),
US Department of Commerce, Washington, DC, p. 55.
EIA (Energy Information Administration). 2008. State
Energy Related Carbon Dioxide Emissions Estimates.
Department of Energy. Washington, DC.
Ferreira, S., Vincent, J. 2005. Genuine savings: Leading
Indicator of Sustainable Development? Economic
Development and Cultural Change 53: 737-754.
Hamilton, K., Clemens, M.A. 1999. Genuine savings
rates in developing countries. World Bank Economic
Review 13, 333-356.
Hanley, N., McGregor, P., Turner, K., Wilson, M.
2002. Greening the National Accounts for Scotland.
Scottish Economic Policy Network (scotecon). ISBN
1-904365-03-5. University of Stirling, Stirling.
Hrubovcak, J., LeBlanc, M., Eakin, B. 2000.
Agriculture, natural resources, and environmental
accounting. Environmental and Resource Economics
17, 145-162.
Hurd, B.. Coonrod, J. 2007. Climate Change and Its
Implications for New Mexico's Water Resources and
Economic Opportunities. National Commission on
Energy Policy. Las Cruces, NM: New Mexico State
University.
Huszar, P., Piper, S. 1986. Estimating the off-site costs
of wind erosion in New Mexico. Journal of Soil and
Water Conservation 41. 414-416.
-------
Landefeld I, Seskin, E., Fraumeni, B. 2008. Taking the
Pulse of the Economy: Measuring GDP. Journal of
Economic Perspectives 22, 193-216.
NASS (National Agricultural Statistics Service, United
States Department of Commerce, Bureau of the
Census). 1984. United States Census of Agriculture:
1982. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Service, United
States Department of Commerce, Bureau of the
Census). 1989. United States Census of Agriculture:
1987. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Service, United
States Department of Commerce, Bureau of the
Census). 1994. United States Census of Agriculture:
1992. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Service, United
States Department of Agriculture). 2009a. United
States Census of Agriculture: 1997. Available online
at http://www nass.usda.gov/Statistics_by_State/
Colorado/index.asp. Last accessed June 26, 2009.
NASS (National Agricultural Statistics Service, United
States Department of Agriculture). 2009b. United
States Census of Agriculture: 2002. Available online
at http://www nass.usda.gov/Statistics_by_State/
Colorado/index.asp. Last accessed June 26, 2009.
NRCS (Natural Resources Conservation Service).
2007. 2003 Annual NRI: Soil Erosion. February
2007. Available online at http://www nrcs.usda.
gov/technical/NRI/2003/SoilErosion-mrb.pdf. Last
accessed June 26, 2009.
Neumayer, E. 2010. Weak versus Strong Sustainability:
Exploring the Limits of Two Opposing Paradigms.
Edward Elgar, Northampton, MA.
Pearce, D., Atkinson, G. 1995. Measuring sustainable
development, in: Bromley, D. (Ed.), The Handbook
of Environmental Economics. Blackwell Publishers,
Cambridge, MA.
Pezzey, J., Hanley, N., Turner, K., Tinch, D. 2006.
Comparing augmented Sustainability measures for
Scotland: Is there a mismatch? Ecological Economics
57, 60-74.
Ringquist E., J. Lee, Ervin, R. 1995. Evaluating the
environmental effects of agricultural policy: the soil
bank, the CRP and airborne paniculate concentrations.
Policy Studies Journal 23, 519-533.
Seroa da Motta, R., C. Ferraz do Amaral. 2000.
Estimating timber depreciation in the Brazilian
Amazon. Environment and Development Economics
5, 129-142.
Stern, D. I. 1997. The capital theory approach to
Sustainability: a critical appraisal. Journal of Economic
Issues 31, 145-173.
Tol, R. 2005. The marginal damage costs of carbon
dioxide emissions: An assessment of the uncertainties.
Energy Policy 33, 2064-2074.
Tol, R. 2007. The social cost of carbon: trends,
outliers, and catastrophes. Economics Discussion
Papers No 2007-44; 2007. Available online at
http ://www. economics-ej ournal. org/economics/
discussionpapers/2007-44. Last accessed June 26,
2009.
USEPA. 2000. Guidelines for Preparing Economic
Analyses. U.S. Environmental Protection Agency,
Office of Policy, Economics and Innovation.
EPA/240/R-00/003.
Weitzman, M.L., Lofgren, K.-G. 1997. On the Welfare
Significance of Green Accounting as Taught by
Parable Journal of Environmental Economics and
Management 32, 139-153.
Williams, P., Young, M. 1999. Costing Dust: How
much does wind erosion cost the people of South
Australia? Final Report. Policy and Economic
Research Unit, CSIRO Land and Water, August.
Wilson, W, Hoehn, J. 2006. Valuing environmental
goods and services using benefit transfer: The state-of-
the art and science. Ecological Economics 60, 335-
342.
Woodruff, N., Siddoway, F. 1965. A wind erosion
equation. Soil Science Society American Proceedings
29, 602-608.
-------
Table 4.1- Green Net Regional Product (GNRP): Variables, Purpose, and Data Sources.
Scale
Variable name
Purpose
Source*
Home Website
Specific Data Link
US
Gross Domestic
Product (GDP)
Consumption of
Fixed Capital
(CFC)
Ratio: CFC/GDP
Gross Domestic
Income (GDI)
Ratio: GDP/GDI
Ratio needed to
convert GDP by state
to NDP by state
For years prior to
1997, GDP by slate
was calculated
differently. This ratio
makes GDP by state
relatively consistent.
BEA
BEA
EBT
BEA
EBT
http : //www. bea.gov/
http://www.bea.gov/
http : //www. bea.gov/
http://www.bea.gov/nationaVnipaweb/
Table View.asp?SelectedTable=43&V
iewSeries=NO&.Tava=no&Request3P
lace=N&3Place=N&FromView=YES
&Freq= Year&FirstYear= 1 980&LastY
ear=2005&3Place=N&Update=Updat
e&JavaBox=no
http://www.bea.gov/national/nipaweb/
Table View.asp?SelectedTable=43&V
iewSeries=NO&Java=no&Request3P
lace=N&3Placc=N&FromView=YES
&Freq= Year&FirstYear= 1 980&LastY
ear=2005&3Place=N&Update=Updat
e&JavaBox=no
http://www.bea.gov/natkmal/nipaweb/
Table View.asp?SelectedTable=43&V
iewSeries=NO&.Tava=no&Request3P
lace=N&3Place=N&FromView=YES
&Freq= Year&FirstYear= 1 980&LastY
ear=2005&3Place=N&Update=Updat
e&JavaBox=no
-------
Scale
Variable name
Purpose
Source*
Home Website
Specific Data Link
Slate
Personal Income
(PI) for Colorado
(thousands of
dollars)
GDP for CO
(thousands of
dollars)
Net Domestic-
Product (NDP) for
CO (thousands of
dollars)
Ratio: NDP/PI
for CO
CO, emissions for
Colorado
Population for
Colorado
Personal
Income (PI) for
New Mexico
(thousands of
dollars)
GDP for NM
(thousands of
dollars)
Net Domestic
Product (NDP) lor
NM (thousands of
dollars)
Ratio: NDP/PI for
NM
CO emissions for
New Mexico
Population for New
Mexico
Ratio multiplied by
PI for SLB provides
estimate of Net
Regional Product
(NRP)
Used to estimate per
capita CO, emissions
for CO and total
emissions for SLB
Ratio multiplied by
PI for SLB provides
estimate of Net Regional
Product (NRP)
Used to estimate per
capita CO2 emissions for
NM and total emissions
for SLB
BEA
BEA
EBT
EBT
EIA
BEA
BEA
BEA
EBT
EBT
EIA
BEA
http://www.bea.gov/
http://www.bea.gov/
http://www.eia.doe.gov/
http://www.bea.gov/
http://www.bea.gov/
http://www.bea.gov/
http://www.eia.doe.gov/
http://www.bea.gov/
http://www.bea.gov/regional/spi/
default . cfm? selTabl e= summary
http://www.bea.gov/regional/gsp
http://www.eia.doe.gov/oiaf/1605/
state/state__emissions.htnil
http://www.bea.gov/regional/spi/
dcfaull.cfm?sclTablc=summary
http://www.bea.gov/regional/spi/
default . cfin? sel Table= summary
http://www.bea.gov/regional/gsp
http://www.eia.doe.gov/oiai71605/state/
state emissions.html
http://www.bea.gov/regional/spi/default.
cfm?selTable=summary
-------
Scale
Variable name
Purpose
Source*
Home Website
Specific Data Link
SLB
Personal Income
(P!) for Colorado
counties
Water Budget for
the Upper Rio
Grande River
Crop acres planted
Total crop land,
Total harvested land
Crop rotation acres.
Soil type, Tillage/
residue management
Wind Erosion
(tons/year)
Population
Used to estimate SLB
Net Regional Product
Change in ground-water
levels
Used to estimate total
crop rotation acres
and tillage/residue
management practices
needed for wind erosion
Used to estimate total
crop rotation acres
and tillage/residue
management practices
needed for wind erosion
Used to estimate total
crop rotation acres
and tillage/residue
management practices
needed for wind erosion
Used for CO, emissions
BEA
CDSS
NASS
NASS
EBT
WEQ,
EBT
BEA
http://www.bea.gov/
http://cdss.state.co.us/
DNN/default.aspx
http://www.nass.usda.gov/
index. asp
http://www.agcensus.
usda.gov/
http://www.woru.ksu.odu/
nrcs.'wcq.html
http://www.bea.gov/
http://www.bea.gov/regional/reis/default.
cfm?selTable=CAl-3§ion=2
http://cdss. state, co. us/DNN/RioGrande/
tabid/57/Default. aspx
http://www.nass.usda.gov/Statistics by
State/Colorado/index. asp
http://www.agcensus.usda.gov/
Publications/2007/Full Report/Census
by State/Colorado/index.asp
http:HWWw.bea.gov/regional/reis/default.
cfm?selTable=CAl-3§ion=2
"BEA-Biircau of Economic Analysis
F.BT-F.stimated by Team
EIA-Energy Information Agency
CDSS-Colorado's Decision Support Systems
NASS-National Agricultural Statistics Service
WEQ-Wind Erosion Equation
-------
Table 4.2 - Estimates of all components of Green Net Regional Product (GNRP).
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
NRP*
$775,613
$727,258
$661,019
$761,240
$759,773
$771,614
$737,433
$719,223
$702,787
$754,919
$793,954
$786,673
$794,303
$894,289
$953,794
$969,925
$1,001,422
$1,057,552
$1,061,384
$1,135,535
$1,106,954
$1,086,389
$1,168,512
$1,186,687
$1,176,482
$1,190,061
Ground-
water
Storage
(acre-feet)
81,972
-287,123
222,041
250,925
157,077
306,292
258,278
63,079
-300,027
-278,902
-39,154
42,696
50,215
256,157
-28,618
445,193
-435,869
210,103
-110,554
252,837
-504,537
24,917
-860,888
-302,288
-21,781
125,141
Ground-
water
Storage
Dep.*
$6,190
-$21,681
$16,766
$18,947
$11,861
$23,128
$19,503
$4,763
-$22,655
-$21,060
-$2,957
$3,224
$3,792
$19,342
-$2,161
$33,616
-$32,912
$15,865
-$8,348
$19,092
-$38,098
$1,881
-$65,006
-$22,826
-$1,645
$9,449
Total
Erosion
(tons/year)
4506372
4858200
4703520
4739397
4838600
4480889
4114138
3443462
3402277
3457701
3254087
3216740
3082521
3047803
2965568
2984107
3239852
3228373
3498540
3236475
3091684
2869825
2907283
3037519
3051094
3028607
Soil
Erosion
Damage*
$10,913
$11,765
$11,390
$11,477
$11,717
$10,851
$9,963
$8,339
$8,239
$8,373
$7,880
$7,790
$7,465
$7,381
$7,182
$7,226
$7,846
$7,818
$8,472
$7,838
$7,487
$6,950
$7,040
$7,356
$7,389
$7,334
C02
Emissions
(metric
tons/year)
1045868
1016955
1028255
1060526
1057820
1034373
976931
1023873
1062133
1087591
1117838
1062657
1067248
1080506
1079803
1066590
1109930
1153430
1157612
1173798
1227015
1242678
1194397
1199228
1230885
1228372
Damage
Cost*
$24,845
$24,158
$24,426
$25,193
$25,128
$24,571
$23,207
$24,322
$25,231
$25,836
$26,554
$25,243
$25,352
$25,667
$25,651
$25,337
$26,366
$27,400
$27,499
$27,884
$29,148
$29,520
$28,373
$28,488
$29,240
$29,180
GNRP*
$746,045
$669,655
$641,969
$743,518
$734,789
$759,320
$723,766
$691,326
$646,662
$699,650
$756,563
$756,864
$765,278
$880,583
$918,801
$970,978
$934,298
$1,038,199
$1,017,065
$1,118,905
$1,032,222
$1,051,801
$1,068,093
$1,128,018
$1,138,209
$1,162,997
Change
in
GNRP*
-$76,390
-$27,686
$101,549
-$8,729
$24,531
-$35,554
-$32,440
-$44;664
$52,988
$56,913
$301
$8,414
$115,305
$38,218
$52,177
-$36,680
$103,902
-$21,134
$101,840
-$86,683
$19,579
$16,292
$59,925
$10,191
$24,787
* In thousands of 2005 dollars
-------
•Real NRP
•Real GNRP
Change in GNRP
1,400,000
1,200,000
COCOCOCOCOCOCOCOOTOTOTOTOTOTOTOTOTOTOOOOOO
O)O)O)O)O)O)O)O)O)O)O)O)O)O)O)O)O)O)O)O)OOOOOO
-------
•Lower Bound Upper Bound ^—Real GNRP
$1,400,000
Year
Figure 4.3 - Sensitivity Analysis Using Upper and Lower Bounds on the Prices for Natural Capital Depreciation.
-------
Appendix 4-A: Wind
Calculations
Green Net Regional Product, Emergy Analysis, and
Fisher information all included wind erosion in the
calculations. This appendix describes details of the
approach for estimating wind erosion in the San Luis
Basin (SLB). Agriculture was not a major activity in all
SLB counties, so we only focused on Alamosa, Conejos,
Costilla, Rio Grande, and Saguache counties. The
estimates were based primarily on the following data: 1)
National Agricultural Statistics Sen-ice (Colorado profile
and Census of Agriculture); 2) farming practices and
trends in the SLB based on the knowledge, experience,
and expert opinion of Richard Sparks, an agronomist
and irrigation specialist in the US Department of
Agriculture (USDA), Natural Resource Conservation
Sendee (NRCS), Center, CO; and 3) Wind Erosion
Equation (Woodruff and Siddoway 1965). Each plays an
important role in the general approach described below.
This work had never been attempted for the SLB over
the 26-year study period. Figure 4-A. I identifies the
basic steps we used in the process.
The general approach for collecting initial data and
estimating wind erosion was as follows:
1. We first needed the individual county acreages of
spring barley, spring wheat, winter wheat, potatoes.
alfalfa, oats, and other hay (only for Conejos,
Costilla. and Saguache counties). The National
Agricultural Statistics Sen-ice (NASS) provides
annual estimates of planted acres of different crops
by county and data on total crop land and total
harvested land beginning in 1982 and continuing
even-1 five years until 2002 through the Census of
Agriculture (NASS 1984, 1989, 1994, 2009a, 2009b).
2. The soil credibility index was estimated for county-
acreage for two major soil types (soil credibility
index 1=134 tons per acre per year and 1=86 tons per
acre per year) based on expert opinion and Web Soil
Survey (Soil Survey Staff 2009). Index I represents
potential for soil loss measured in tons per acre per
year. It assumes the field is isolated, level, smooth,
unsheltered, bare, wide, and loose (NRCS 2002).
Table 4-A. 1 presents the approximate distribution of
crops between the two soil types.
3. We estimated the acres of different crop rotations by
the two soil types (based on the soil credibility index)
for each county. Seven crop rotations for the counties
were included in the calculation (Table 4-A.2). A
crop rotation is a series of different crops in the same
field in succeeding seasons. It could occur over a
couple of years to a much longer period. There are
many benefits of rotating crops including preventing
diseases, reducing soil erosion, controlling pests, and
improving the soil fertility (NRCS 2002).
Some rotations were included to factor in Hie
drought and aquifer decline years (2001 - 2007).
For example, cropland acreages were idled or
planted to Sorghum /Sudan cover crops instead of
spring grain. Other trends were included such as
introducing Winter Rye cover crops between years
of back-to-back potatoes from 1980 through 2000
and introducing Winter Wheat into the Potato - Grain
rotations around the late 1990s depending on the
county.
Total acres of irrigated cropland were based on
Agricultural Statistics for Colorado and NASS.
Because NASS reports census data even- five years,
the estimates for the other years were based on
trends and expert opinion (total acres of planted
crops gleaned from the agricultural statistics for the
counties provided the initial minimum total acres of
irrigated cropland). Total irrigated cropland acres
minus the total crops planted (based on Agricultural
Statistics for Colorado) equaled unreported crops
(e.g.. canola, sorghum, and vegetables) and implanted
acres. We assumed that vegetables and unplanted
acres had similar wind erosion rates (therefore, we
combined them into one rotation). Because of some
large numbers of unaccounted acres, we included
Colorado estimates for Other Hay in Conejos,
Costilla, and Saguache counties.
The first step was to estimate continuous potatoes.
Alamosa and Rio Grande counties planted about
20% of their total potatoes as continuous in the
1980's and grew to about 40% in the 1990's. The
number of acres of continuous potatoes decreased
around 2000 to about 20%, which corresponded to
the decline of the aquifer and an increase in potato
pathogens. Once we estimated continuous potatoes,
we combined the remaining potatoes with small
grains. The Potato-Small Grain rotation was a
two-year rotation. We estimated the Alfalfa - Small
Grain rotation next and the remaining small grains
were assumed to be planted as continuous grains.
We assumed the Alfalfa - Small Grain rotation was
a 7-year rotation. For each count}' and ever}- year,
we compared total irrigated cropland acres based
on Agricultural Statistics for Colorado and NASS
to our estimate of total irrigated cropland acres
based on crop rotation acres to make sure the data
and the approach were consistent. We accounted
for evety acre of crops reported in the Agricultural
Statistics for Colorado. Once this set approach was
-------
completed, each year was examined for accuracy
and the time series was examined for trends. Minor
revisions were made depending on the year and crop
rotation based on known trends. For example, certain
years had very high estimates of continuous small
grains and the values were not realistic with known
quantities; therefore, we adjusted that year to have
more Potato - Small Grain acres in order to reduce
continuous small grains. This required reducing the
percent of continuous potato acres as well.
4. After all of the calculations for Hie individual counties
were reviewed and determined to be accurate, a total
estimate of crop acreage for each rotation in the SLB
was then calculated. Table 4-A.3 displays total crop
acreage for the SLB for 1980-2005. Rows 2-7 are
based on the Agricultural Statistics for Colorado.
5. We accounted for the general trend of improving
residue management and changes in tillage practices
with the different crop rotations. The two practices
were moldboard plow and noninversion tillage
and we divided the rotation acreage between
the two practices based on the trends for residue
management. Using expert opinion, we assumed
that 10% of continuous small grain acres were
noninversion tillage in 1980 for soil type 1=134. That
percent grew to a high of 80% in 1997 and continued
through 2005.
6. We used the Wind Erosion Equation (WEQ) to
estimate average annual wind erosion for each of
the different crop rotations and different tillage
practices on the two soil types. WEQ is based on
Woodruff and Siddoway (1965) where Erosion (tons/
acre/year) is a function of soil credibility index, soil
ridge roughness factor, climatic factor, field length
along the prevailing wind erosion direction, and
equivalent quantity of vegetative cover. WEQ is
used by the NRCS to estimate average wind erosion.
The Wind Erosion Research Unit (WERU) of the
NRCS has developed the "Excel Spreadsheet and
Guidance Document" on which the wind erosion
estimates are based (NRCS 2003). Table 4-A.4
presents the average annual wind erosion (tons/acre)
for noninversion tillage and Table 4-A. 5 presents
the average annual wind erosion for conventional,
moldboard plow. Continuous potatoes show no
difference in average annual wind erosion, but other
crop rotations reveal significant changes in the
amount of erosion with the noninversion tillage.
7. Combining crop rotation acres and soil types with
the WEQ average annual wind erosion, we estimated
total erosion for the SLB in tons/year (Table 4-A.6).
We divided this estimate by total acres of irrigated
cropland to estimate tons/acre/year, which allows the
NRCS to examine the trends of erosion management.
From 1980 to 2005, there was a general trend of tons/
acre/year decreasing, revealing an improvement in
practices to reduce wind erosion (Table 4-A.6).
8. The final step was to have experts from the region
review the results for wind erosion to assess if the
results seem reasonable.
References
NASS (National Agricultural Statistics Service, United
States Department of Commerce, Bureau of the
Census). 1984. United States Census of Agriculture:
1982. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Sendee. United
States Department of Commerce, Bureau of the
Census). 1989. United States Census of Agriculture:
1987. Volume I; Geographic Area Series; Part 6
Colorado State and County' Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Sendee, United
States Department of Commerce, Bureau of the
Census). 1994. United States Census of Agriculture:
1992. Volume I; Geographic Area Series; Part 6
Colorado State and County Data. Washington, DC:
US Government Printing Office.
NASS (National Agricultural Statistics Sendee, United
States Department of Agriculture). 2009a. United
States Census of Agriculture: 1997. Available online
at http://www.nass.usda.gov/Statistics_by_State/
Colorado/index.asp. Last accessed June 26, 2009.
NASS (National Agricultural Statistics Sendee, United
States Department of Agriculture). 2009b. United
States Census of Agriculture: 2002. Available online
alhtip://www.nass.usda.gov/Staiistics_by_State/
Colorado/index.asp. Last accessed June 26, 2009.
NRCS (Natural Resources Conservation Sendee).
2002. National Agronomy Manual. United States
Department of Agriculture, Natural Resources
Conservation Sendee. October 2002. Accessed at:
http://www.weru ksu.edu/nrcs/nam/M_190.pdf.
NRCS (Natural Resources Conservation Service). 2003.
Wind Erosion Equation Guidance Document. USD A.
NRCS. Accessed at: http://www.weni.ksu.edu/nrcs/
weqguidance062503 .doc.
-------
Soil Survey Staff. 2009. Natural Resources
Conservation Service, United States Department of
Agriculture. Web Soil Survey. Accessed at http://
websoilsurvey.nrcs.usda.gov/
Woodruff, N., Siddoway, F. 1965. A wind erosion
equation. Soil Science Society American Proceedings
29, 602-608.
Table 4-A. 1 - Approximate distribution of crops between the two soil types (Based on Soil Erodibility Index, I).
County
Alamosa
Conejos
Costilla
Rio Grande
Saguache
1=134
50%
0%
80%
10%
30%
1=86
50%
100%
20%
90%
70%
Table 4-A.2 - Crop rotations included in soil erosion calculation.
Crop Rotation
Continuous Potatoes, Vegetables, and Unplanted Acres
Potatoes-Winter Wheat
Potatoes-Winter Rye
Potatoes-Small Grain
Potatoes-Canola and/or Sorghum Cover
Alfalfa-Small Grain
Continuous Small Grain
Description
Potatoes planted every season, vegetables and unplanted acres assumed to have the
same erosion rate as continuous potatoes
2-year rotation: one year potatoes, one year winter wheat
1-year rotation with rye planted after potato harvest
2-year rotation: one year potatoes, one year small grain (e.g., barley, spring wheat)
2-year rotation: one year potato, one year sorghum sudan
7-year rotation: six years alfalfa, one year small grain (e.g., barley, spring wheat)
Small grain planted every season
-------
Table 4-A.3 - Total acreage by crop rotation and Soil Erodibility Index.
Variable name
Potatoes All (Planted)
Wheat Other Spring (Planted)
Barley All (Planted)
Hay Alfalfa (Dry) (Harvested)
Wheat Winter All (Planted)
Oats
Unreported crops (Sorghum, carrots/lettuce/canola) +
implanted acres
Vegetables & implanted acres (100% to 30% of unreported
crops)
Canola and sorghum cover ( 0% to 70% of unreported crops)
Other Hay (Wild dry hay, other haylage, etc.)
Total irrigated cropland acres
134: Potatoes
86: Potatoes
134: Continuous Potatoes + Vegetables and unplanted acres
86: Continuous Potatoes + Vegetables and unplanted acres
134: Potatoes- Winter Wheat
86: Potatoes-Winter Wheat
134: Potatoes- Winter Rye (only potato acreage)
86: Potatoes-Winter Rye (only potato acreage)
134: Potatoes-Small Grain (1:1 acreage)
86: Potatoes-Small Grain (1:1 acreage)
134: Potatoes-Canola and/or sorghum cover (1:1 acreage)
86: Potatoes-Canola and/or sorghum cover (1:1 acreage)
134: Alfalfa-small grain (7 year rotation)
86: Alfalfa-small grain (7 year rotation)
Remaining Small Grain
134: Continuous Small Grain
86: Continuous Small Grain
1980
37000
36000
107000
122900
0
16000
56550
11842
0
44708
375450
8090
28910
5492
12630
0
0
0
0
13604
47836
0
0
24547
118837
1981
40500
43000
118000
99800
0
17000
57150
13468
0
43682
375450
9310
31190
7796
15682
0
0
0
0
14318
46662
0
0
19460
96973
1982
45500
38000
114000
92800
0
24000
61150
15468
0
45682
375450
11530
33970
5612
20646
0
0
0
0
18326
51094
0
0
18468
89798
1983
47000
50000
100000
89700
0
39000
53537
14555
0
38982
379237
11100
35900
4946
20710
0
0
0
0
17520
54280
0
0
18597
86053
1984
53500
49000
117000
108600
0
32000
39755
6700
0
33055
399855
14100
39400
4720
13880
0
0
0
0
22860
60340
0
0
21187
105513
1985
56500
50500
115000
129600
0
26000
28055
800
0
27255
405655
14680
41820
3298
9862
0
0
0
0
23864
64416
0
0
25737
125463
1986
57000
40000
120000
110000
0
15000
47908
4834
0
43074
389908
17440
39560
5760
12314
0
0
0
0
26948
60572
0
0
21467
106867
1987
61000
38000
77000
118000
0
14000
62208
8123
0
54085
370208
18040
42960
5246
16837
0
0
0
0
28220
65860
0
0
23217
114450
1988
60000
41000
65000
115000
0
24000
70574
10923
0
59651
375574
17660
42340
6775
18168
0
0
0
0
27292
64668
0
0
22108
112058
1989
62000
55000
79000
105000
0
15500
65910
7450
0
58460
382410
17930
44070
7563
14157
0
0
0
0
27826
67634
0
0
19833
102667
1990
65500
29000
84000
105000
0
16700
82210
11413
0
70797
382410
19685
45815
7051
18329
0
0
669
669
30409
69981
0
0
19833
102667
1991
71000
24900
89500
105000
0
11800
80999
11688
0
69311
383199
20870
50130
7411
16764
0
0
3482
3645
30771
71999
0
0
19717
102783
107797
130877
125823
138150
138300
125760
112907
62293
64853
84270
62005
26735
81062
31158
99719
31795
94028
34287
103863
34990
103310
32605
93155
27876
85031
14003
48290
15342
49511
18967
65303
13722
48283
57315
12861
44454
-------
Variable name
Potatoes All (Planted)
Wheat Other Spring (Planted)
Barley All (Planted)
Hay Alfalfa (Dry) (Harvested)
Wheat Winter All (Planted)
Oats
Unreported crops (Sorghum, carrots/lettuce/
canola) + implanted acres
Vegetables & implanted acres (100% to 30% of
unreported crops)
Canola and sorghum cover ( 0% to 70% of
unreported crops)
Other Hay (Wild dry hay, other haylage, etc.)
Total irrigated cropland acres
134: Potatoes
86: Potatoes
134: Continuous Potatoes + Vegetables and
unplanted acres
86: Continuous Potatoes + Vegetables and
unplanted acres
134: Potatoes-Winter Wheat
86: Potatoes- Winter Wheat
134: Potatoes-Winter Rye (only potato acreage)
86: Potatoes- Winter Rye (only potato acreage)
134: Potatoes-Small Grain (1:1 acreage)
86: Potatoes-Small Grain (1:1 acreage)
134: Potatoes-Canola and/or sorghum cover
(1:1 acreage)
86: Potatoes-Canola and/or sorghum cover
(1:1 acreage)
134: Alfalfa-small grain (7 year rotation)
86: Alfalfa-small grain (7 year rotation)
Remaining Small Grain
134: Continuous Small Grain
86: Continuous Small Grain
1992
66500
32000
77500
115000
0
13200
87166
13569
0
73597
391366
19680
46820
5746
18084
0
0
4019
5750
27680
65258
0
0
21058
113108
1993
72500
23500
66000
129000
0
17900
78529
18186
0
60343
387429
22510
49990
6280
16682
0
0
6147
12656
29740
68104
0
0
23275
127225
1994
74000
30500
58000
135000
0
18000
73929
13648
0
60281
389429
22860
51140
4736
13616
0
0
5905
12486
31018
70796
0
0
24267
133233
1995
77000
24000
71000
140000
0
24000
64168
6105
540
57523
400168
23490
53510
3202
7694
0
0
5820
13083
32387
73147
156
924
26133
137200
1996
78000
43300
61000
130000
1500
21000
71101
11019
4824
55258
405901
25140
52860
4512
11289
300
2700
6287
12727
31251
64509
3170
6478
22750
128917
1997
77000
34500
63500
130000
5000
21000
101506
15191
6312
80003
432506
24660
52340
4749
15105
3760
6240
6074
12245
26996
58415
3483
9141
23158
128508
1998
75800
25000
60000
135000
3000
15000
113768
31873
9276
72619
427568
24740
51060
10556
25863
2320
3680
6089
11921
26094
55842
5995
12557
24150
133350
1999
77200
34000
65000
160000
3000
17000
80088
12918
7083
60087
436288
24720
52480
4981
12517
2300
3700
5879
12093
27608
61525
4960
9205
30683
155983
2000
75800
32200
77000
165000
5000
30000
51688
1920
2343
47425
436688
24010
51790
1637
3807
4600
5400
4505
9255
30715
71633
1562
3123
31033
161467
2001
68100
25000
59000
167000
5000
18000
75495
8282
6317
60897
417595
21630
46470
2840
8576
4040
5960
4101
8369
24948
57409
4189
8444
31442
163392
2002
71600
14500
49000
135000
4000
17000
80293
13888
13735
52670
371393
23110
48490
4332
12359
3400
4600
3391
7686
23481
56491
11025
16445
25608
131892
2003
66300
11000
53000
130000
4000
20617
99406
22924
12819
63664
384323
20980
45320
8060
17457
3120
4880
3045
7126
21911
51522
9452
16186
21642
130025
2004
64083
15067
52000
132833
4150
20617
95573
22747
13387
59439
384323
20310
43773
8880
16368
3220
5080
3215
7102
18945
48515
10698
16075
21039
133933
2005
58883
15067
38000
147000
4150
20617
100606
27462
16370
56774
384323
18190
40693
10216
19376
3220
5080
2813
6512
13142
40675
13397
19342
22750
148750
57064
36978
33093
42900
55753
54627
36532
44767
60526
32988
18014
26233
31815
12206
44858
6193
30785
6371
26722
8476
34423
10126
45628
9271
45357
4619
31913
5675
39092
8855
51671
5435
27553
2164
15850
4906
21327
6802
25013
22275
5693
16582
-------
Table 4-A.4 - Average annual wind erosion by crop rotation: practicing mulch or noninversion tillage.
Soil type: Crop rotation
134: Continuous Potatoes + Vegetables and unplanted acres
86: Continuous Potatoes + Vegetables and unplanted acres
134: Potatoes-Winter Wheat
86: Potatoes-Winter Wheat
134: Potatoes- Winter Rye (only potato acreage)
86: Potatoes-Winter Rye (only potato acreage)
134: Potatoes- Small Grain
86: Potatoes-Small Grain
134: Potatoes-Canola and/or sorghum cover (1:1 acreage)
86: Potatoes-Canola and/or sorghum cover (1:1 acreage)
134: Alfalfa-small grain (7 year rotation)
86: Alfalfa-small grain (7 year rotation)
134: Continuous Small Grain
86: Continuous Small Grain
Average Annual Wind
Erosion (t/ac)
39.8
16.1
7.6
2.4
10.7
3.4
19.5
7.4
20.9
8.5
5.7
2.2
9.8
2.9
Table 4-A.5 - Average annual wind erosion by crop rotation: conventional, moldboard plow.
Soil type: Crop rotation
134: Continuous Potatoes + Vegetables and unplanted acres
86: Continuous Potatoes + Vegetables and unplanted acres
134: Potatoes-Winter Wheat
86: Potatoes-Winter Wheat
134: Potatoes-Winter Rye (only potato acreage)
86: Potatoes- Winter Rye (only potato acreage)
134: Potatoes-Small Grain
86: Potatoes-Small Grain
134: Potatoes-Canola and/or sorghum cover (1:1 acreage)
86: Potatoes-Canola and/or sorghum cover (1: 1 acreage)
134: Alfalfa-small grain (7 year rotation)
86: Alfalfa-small grain (7 year rotation)
134: Continuous Small Grain
86: Continuous Small Grain
Average Annual Wind
Erosion (t/ac)
39.8
16.1
24.9
10.1
15.5
5.2
33.2
13.3
44.2
18.9
12.5
5.4
35.6
14.6
-------
Table 4-A.6 - Total Erosion (tons/year) for the SLB and Tons per Acre per Year.
Variable name
134: Continuous Potatoes + Vegetables and unplanted acres
86: Continuous Potatoes + Vegetables and unplanted acres
134: Potatoes-Winter Wheat
86: Potatoes- Winter Wheat
134: Potatoes-Winter Rye (only potato acreage)
86: Potatoes- Winter Rye (only potato acreage)
134: Potatoes-Small Grain (1:1 acreage)
86: Potatoes-Small Grain (1:1 acreage)
134: Potatoes-Canola and/or sorghum cover (1:1 acreage)
86: Potatoes-Canola and/or sorghum cover (1:1 acreage)
134: Alfalfa-small grain (7 year rotation)
86: Alfalfa-small grain (7 year rotation)
134: Continuous Small Grain
86: Continuous Small Grain
Total Erosion (tons/year)
Tons/Acre/Year
1980
218471
203640
0
0
0
0
442140
637745
0
0
297658
639097
883472
1184147
1981
310112
252854
0
0
0
0
435858
622093
0
0
229443
521517
1029634
1456687
1982
223236
332892
0
0
0
0
520125
666132
0
0
205352
468873
968494
1318415
1983
196724
333916
0
0
0
0
485222
691682
0
0
200536
435848
1000054
1395415
1984
187754
223798
0
0
0
0
617422
751132
0
0
221352
534410
975332
1327400
1985
131189
159013
0
0
0
0
628156
782899
0
0
251610
615813
824536
1087672
1986
229104
198550
0
0
0
0
672335
718340
0
0
195453
524535
632875
942946
1987
208670
271475
0
0
0
0
665325
761654
0
0
203592
543840
281713
507194
1988
269491
292941
0
0
0
0
624711
728822
0
0
193873
532475
268980
490984
1989
300858
228262
0
0
0
0
598729
742330
0
0
173923
471776
332527
609294
1990
280494
295533
0
0
7969
3193
654307
726867
0
0
167265
455704
240576
422178
1991
294794
270302
0
0
41480
17400
640972
747827
0
0
166281
440132
208853
388698
4506372
4858200
4703520
4739397
4838600
4480889
4114138
3443462
3402277
3457701
3254087
3216740
12.00
12.94
12.53
12.50
12.10
11.05
10.55
9.30
9.06
9.04
8.51
8.39
-------
Variable name
134: Continuous Potatoes
+ Vegetables and unplanted
acres
86: Continuous Potatoes +
Vegetables and unplanted
acres
134: Potatoes-Winter Wheat
86: Potatoes- Winter Wheat
134: Potatoes-Winter Rye
(only potato acreage)
86: Potatoes- Winter Rye
(only potato acreage)
134: Potatoes-Small Grain
(1:1 acreage)
86: Potatoes- Small Grain
(1:1 acreage)
134: Potatoes-Canola and/or
sorghum cover (1:1 acreage)
86: Potatoes-Canola and/or
sorghum cover (1:1 acreage)
134: Alfalfa-small grain (7
year rotation)
86: Alfalfa-small grain (7
year rotation)
134: Continuous Small
Grain
86: Continuous Small Grain
Total Erosion (tons/year)
Tons/Acre/Year
1992
228583
291580
0
0
47876
26917
576581
658593
0
0
177596
484345
198214
392235
1993
249800
268972
0
0
73225
58076
619505
687309
0
0
196291
524878
100567
269180
1994
188374
219536
0
0
70334
56144
646113
693632
0
0
204654
549666
103465
233650
1995
127358
124050
0
0
69321
57625
674643
716660
5072
14561
211624
544552
137649
300992
1996
179485
182021
3320
16808
74894
54883
650968
613033
103054
102089
184226
511675
164431
398963
1997
188911
243547
41613
38846
72353
52806
562347
537920
113238
144052
187533
510055
138560
396594
1998
419905
417007
25676
22909
72526
51410
525628
497778
194903
197885
195563
529272
69031
279047
1999
198149
201815
23465
21614
70025
51035
556126
548432
161255
145068
238168
594686
84821
341815
2000
65099
61389
46929
29474
53658
39058
618725
638538
48967
49222
240885
615591
132341
451807
2001
112982
138273
41216
30245
48852
35321
502549
511741
126431
133074
244055
622930
81233
240923
2002
172315
199268
31745
21579
39590
31727
472996
503559
319921
250631
190179
502836
32345
138591
2003
320596
281477
29131
22893
35550
29418
441383
459270
263253
238298
160721
495720
73332
186477
2004
353231
263911
30065
21882
37535
29316
381617
432459
285510
228330
156244
510620
101664
218709
2005
406374
312412
30065
19934
32845
26884
264728
362580
341945
264704
168952
567109
85087
144989
3082521
3047803
2965568
2984107
3239852
3228373
3498540
3236475
3091684
2869825
2907283
3037519
3051094
3028607
7.88
7.87
7.62
7.46
7.98
7.46
8.18
7.42
7.08
6.87
7.83
7.90
7.94
7.88
-------
Agriculture Data by
County:
Colorado Agricultural
Statistics, Census of
Agriculture
Knowledge of Farming Practices
by County: crop rotations by
year, soil type, and tillage/residue
management
Richard Sparks, USDA, NRCS
Estimate total acreage of crop
rotations by soil type and
tillage practices
Wind Erosion Equation
(WEQ): tons/acre for
different soil types and crop
rotations for SLB
Wind Erosion: total wind
erosion in tons/year for SLB
Figure 4-A. 1 - General approach to estimating wind erosion in the San Luis Basin.
-------
-------
5.0
Emergy
5.1 Introduction
The effects of changes in system states on sustainability
can be quantified using environmental accounting-
methods based on the transformations of available
energy (Odum 1996). Environmental accounting
using emergy and Emergy Analysis (EmA) provide an
alternative to economic methods, which decision makers
can use to assess the effects of their policies on their
system.
When the available energy transformed from one form to
another is traced through a network of interactions, the
amount of energy required to create any item (e.g., an
energy flow or storage) in the system can be determined.
Tracing the energy flows required to support each
network storage or flow and converting those flows to
energy of a common unit (e.g., solar joules), enables the
determination of the value of any item in the network
(Campbell 2001) in terms of the resources required for
that item to exist within the system. This value can then
be expressed as the solar emjoules (sej) required for its
production.
EmA is a method for assessing the available energy
required for the production of a product or service
in a system in units of equivalent quality (i.e., solar
equivalent joules; Brown 2003, Campbell 2001; see
Odum 1996 for a detailed description). The emergy
evaluation of systems with its emphasis on the role
of renewable and nonrenewable resources used in
system support and its characterization of import and
export exchange is a useful framework for the analysis
of sustainable system operations (see Campbell et
al. (2005) and Odum (1996) for detailed methods for
conducting EmA).
Any quantity of energy can be converted into emergy
by multiplying the original units (J or g) by an emergy
per unit factor, the transformity (sej/J) or the specific
emergy (sej/g), to obtain the emergy equivalent of the
quantity in solar emjoules (sej). Other quantities (e.g.,
information or money) are converted into emergy using
the appropriate emergy per unit factors. The emergy
of environmental, economic, and social products and
sendees is then directly comparable so the relative
advantages and disadvantages of any changes can be
readily determined by direct inspection of the values.
Relative advantage is judged by movement toward
greater emergy flow (empower) through a system
network, because maximizing empower in a system
network is the decision variable used to predict the
outcome of evolutionary competition between alternative
systems designs (Lotka 1922a, 1922b, Odum 1996).
Movement toward sustainability, from an energy systems
perspective, is measured by the degree to which a system
depends upon renewable resources for its operation. As
a system decreases its use of non-renewable emergy
and increases the fraction of renewable emergy in its
total emergy use. the system moves toward a more
sustainable state. The concept of sustainability requires
that we specify what system state is to be sustained and
over what time we expect to sustain it (Campbell 1998).
The system structure and function maintained through
greater use of renewable emergy must be equivalent to
that supported by nonrenewable emergy on a per capita
basis, for there to be no net degradation in the quality
of life. Increasing empower, or the amount of emergy
moving through the system may lead toward greater
health and vitality; whereas, decreasing emergy flow
may lead toward a less robust system with lower levels
of overall function. Therefore, the evolution of networks
toward higher empower designs is the overriding process
governing the success of any condition or state of a
system. This process may occur at the expense of other
systems, depending on the emergy sources available
to support its organization (Campbell et al. 1998).
Ultimately, what is sustainable for a system and for how
long depends on the available energy in the suite of
external forcing functions. Often the operation of the
maximum power principle appears to lead to pulsing
designs, in which resources accumulate slowly over
time and are then consumed rapidly once a threshold is
exceeded (Odum 1996, Campbell 2001). Often under
such scenarios, the only pattern that is truly sustainable
is the pulsing cycle of change (Odum et al. 1995).
EmA provides managers with a comprehensive suite
of indices to assess the ecological and economic well-
being of their system (Odum 1996). It gives managers
a comprehensive decision variable (emergy flow in
the system network) against which they can judge the
efficacy of their policies. From this comprehensive suite
of indices and body of theory, we have abstracted a few-
indices (see below) that apply to system sustainability
and used them to characterize conditions in Hie SLB.
EmA has been applied broadly to analyze nations
(Brown 2003, Lefroy and Rydberg 2003, Ulgiati et
al. 1994), states (Campbell 1998, Campbell et al.
-------
2005, Campbell and Ofart 2009, Tilley 1999) and
counties (Lambert 1999). but the application of the
method to regions has been less common (Tiezzi and
Bastianoni 2008). The San Luis Basin Sustainability
Metrics Project is a regional study that presented us
with a unique set of challenges in data gathering and
interpretation. The standard EmA method calculates a
number of indices showing various aspects of system
functions and the relationship of the system to its next
larger system. However, because our focus was on
regional Sustainability, we thought it would be more
straightforward, if we focused initially on a single index
of emergy that relates strongly to Sustainability. We
identified the fraction of renewable emergy to total
emergy used as the best index to assess, the condition of
the regional system and to determine if the region was
moving toward or away from Sustainability. We report
all the indices used in a standard EmA in this Chapter as
well, and discuss aspects of system structure, function,
and Sustainability that are revealed by our analysis of
these measures. Among these additional indices, we
focused on the total emergy use as a measure of overall
well-being of the region and the Emergy Sustainability
Index, ESI, as a measure of the Sustainability of the
relationship between the SLB and its next larger system
(i.e., the State of Colorado and the Nation). Before
calculating the indices, we had to address several
challenges, including setting the regional boundaries
and assembling time-series data relevant to the region.
In this study, we focused on developing emergy indices
of Sustainability over a period of 26 years from 1980 to
2005. However, data availability limited the complete
emergy evaluation to the eleven-year period of 1995 to
2005.
5.2.
The methods used in the emergy portion of this study
needed to address two challenges. The first was a
technical question of how to set boundaries for a regional
study where more than one criterion could be logically
used. The second challenge was to find and assemble
relevant data on the region for a 26-year period. Both of
these challenges have been seldom addressed in emergy
analyses, where the considerable data requirements of
the method often limit the analysis to systems with well-
established boundaries (e.g., a nation, state, or county
examined for one or two years). Detailed methods used
to carry out an EmA are available in Campbell et al.
(2005) and Campbell and Ohrt (2009). We will outline
the methods in the following sections by discussing the
aspects of the SLB regional study that were unique to
this analysis. The seven steps for this emergy evaluation
are described in Appendix 5-A.
5.2.1
The system boundaries of a region could be based with
strong justification on two different sets of criteria (i.e.,
hydrologic or the political boundaries of counties). Both
sets of boundaries were used for the study. Political
boundaries were used to quantify all human activities
(e.g., energy use, economic activity, etc.) and most
natural inputs and storages (e.g., solar radiation and
forest biomass). The hydrologic system boundary was
used to determine the water flows that support human
activities in the Upper Rio Grande River Basin of the
San Luis Basin (SLB; see Chapter 1 for a description
of the SLB). Ultimately, we chose a mixture of both
because they better represent the actual flow of emergy
through the system and they are similar in size and areal
coverage.
5.2.2. Energy Systems
Information about a region is used to draw a detailed
energy systems diagram using the Energy Systems
Language (ESL) to represent the system under study
(for a detailed description of this process, see Campbell
et al. 2005). The resulting diagram should capture the
essence of the system and its operation. The detailed
model is simplified by aggregating components of
similar function, characterizing the inputs and outputs
and reducing the complexity of internal components and
flows to basic measures of economic activity, resource
supply, and total emergy flow. Energy Systems Models
aggregated in this manner provide indices of system
operation related to Sustainability and self-sufficiency, as
well as other system properties.
We reexamined the indices used to evaluate two aspects
of the relationship between a system and the associated
larger system. First, the emergy yield ratio (EYR) for
a local system, often characterized for a nation, state,
or region as U/F (where U is total emergy flow in the
system and F is feedback from the next larger system),
is redefined as Y/F, where Y or yield is equivalent to
exports. This definition more closely matches the
original definition of EYR formulated for production
processes (Odum 1996). Second, the ratio formed by
the tola! emergy flow in the system (U) and the feedback
(F) from the larger system is defined as the local effect
of emergy investment because it shows the effect of
feedback from the larger system on the local system's
empower (see Appendix 5-B).
5.2.3 Emergy Analysis
EmA tables provide a template for the creation of
the accounts needed to construct an emergy income
statement and balance sheet. The tables provide a simple
way to demonstrate and record the calculation of the
-------
emergy values for animal flows of mass, energy, and
information. In the emergy tables, data on the mass of
flows are converted to energy and then to emergy to
aid in comparisons and to provide information for the
analysis of public policies (Table 5.1). Note that for
many items, transformities are available in the literature
or they can be calculated (Campbell et al. 2005, Odum
1996; see Appendix 5-C for values used in this study
and the other appendices 5-D through 5-G for other
results and discussion, e.g.. Table 5-G.l provides sources
used to obtain data for the EmA of the SLB system).
Figure 5.1 shows a detailed Energy Systems Language
diagram of the SLB region and Table 5.2 lists the energy,
material, and information flows, which are linked to the
diagram through the k coefficient values shown in the
table and the diagram.
The energy content of many items has been widely
tabulated, but a similar compendium of transformities
and other emergy per unit factors does not exist at
present. If llie production process is unknown or not
easily documented, transformities may be determined
using one of the ten methods given in Odum (1996) and
Campbell and Ohrt (2009). The emergy per unit factors
used in this study are given in Appendix 5-C.
We estimated the transformity for each commodity class
to determine the emergy per ton of each commodity
imported. These transforinities were approximated by
averaging known transformities of items within the
class (without sendees); however, not all items in a class
were included in the determination of the transformity.
In some cases, when a transformity was not known for
any item in the class, the parent material was used as a
surrogate for the item's transformity. The use of parent
materials resulted in a minimum estimate of the emergy
imported and exported in these commodity classes. The
information needed for EmA is most often reported as
annual flows of energy and/or mass. Usually mass can
be easily converted to energy as the energy content of
many objects has been widely tabulated (e.g.. USDA
2009). The transformrty (sej/J) and specific emergy
(sej/g) have been calculated for many items and can
be used to convert energy and mass flows to emergy
(Appendix 5-C). In this study, emergy accounts for
renewable and nonrenewable local inflows, renewable
production, imports, and exports were calculated for the
seven counties comprising the SLB.
Initially, we assumed all imports were used in the
system. However, we discovered a situation in the SLB
where Antonito, Colorado (Conejos County) is a railhead
and trans-shipment point for heavy materials, such as
unexpanded perlite and volcanic scoria (San Luis Valley
Development Resources Group 2007). To account for
the movement of these materials through the system.
we removed processed nonmetallic minerals from the
import-export balance, to avoid falsely attributing the
emergy of these materials to use by the SLB system.
In addition, the actual amount of local nonrenewable
resources (e.g., crushed stone, sand and gravel) used in
the SLB was unknown. Export vastly exceeded import
of these materials, so we assumed the difference was
"apparent production" and assumed that 10% of the
"apparent production" was used in the system.
5.2.4, and Sources
Data on the human system, and much of the "natural"
system, of the SLB were collected based on political
boundaries, whereas, hydrologrc data were collected
using the basin boundary (Fig. 1.1). Water resource
data conformed to the boundaries of our system. Data
on economic and social activities were collected at the
county level (Table 5-G.l). When only the state or
national level data were available, we used average per
capita data from the state or national level to estimate
variables (Table 5-G. 1). All data were yearly averages.
Data needed to construct an emergy income statement
of a region can be categorized as follows: (1)
renewable energy inputs to the system and information
on renewable production carried out in the system;
(2) nonrenewable energy inputs, production and
consumption (including any renewable energy sources
used in a nonrenewable manner; i.e., rate of use exceeds
rate of replenishment); (3) imports to the system
(including raw and finished materials and services); (4)
exports from the system (including raw and finished
materials and services). Data on the SLB were gathered
via literature surveys, interviews with experts, the
purchase of a dataset, and meetings with knowledgeable
people from the region (Table 5-G.l).
The need for data on SLB imports and exports was the
limiting factor in determining the period investigated
using EmA. Commodity Flow Surveys produced by
the US Census Bureau even- five years provide such
data at the state level, but these data did not provide
annual estimates and they were not available to us at the
count}' level. Thus, we identified a private company,
IMS Global Insight, Inc. (GI; Global Insight 2009) "
that used data from a network of shippers, publicly
available sources, and US Census Bureau to compile
freight movements by US count}'. GI data for the seven
counties in the SLB, from 1995 to 2005, were purchased
for approximately $20,000. We report less detailed EmA
results using variables for the period 1980-2005.
-------
5.2,5 Summary Tables and Indices
A summary- of the variables with their definitions are
given in Table 5.3. Summary variables were used to
define total emergy used and fraction of renewable
emergy (Fig. 5.2; Table 5.3). Fraction of renewable
emergy used to total emergy used was the primary
evaluation criterion used to assess local system condition
and movement toward or away from sustainability. A
number of additional indices were computed (Table 5.4),
but these were not necessary for estimating the fraction
of renewable emergy to total emergy that we used in this
study (Appendix 5-B presents the definitions of other
indices important for a complete EmA. including the
Emergy Yield Ratio, the Environmental Loading Ratio,
and the Emergy Sustainability Index).
5.3
In this section, we present the major results of EmA for
the SLB. Other results and associated discussion are
located in the appendices.
5.3,1 The Energy of the
Luis
We produced a detailed diagram of an Energy Systems
Model of the SLB environmental system (Fig. 5.1; see
Chapter 1 and Appendix 5-D for a detailed description
of SLB). The large rectangular box represents the
boundaries of the SLB. External forcing functions or
energy, material, or information sources entering from
outside the system are shown as circles. They enter the
system through pathway lines crossing the boundaries.
They are arranged in order of increasing transformity
from left to right around the system boundary. Flows
of energy and matter leave the system as shown by
arrows crossing the boundary. All the flows on pathways
crossing the system boundaries as well as flows of
energy and materials moving within the system are
defined (Table 5.2). Surface and groundwater flow out
of the region into New Mexico. Unused solar energy
(albedo) is reflected into space and wind blows over
the region and passes on to the surrounding systems. A
fully specified energy source (circles) always contains
all the necessary parameters needed to define the source
over the simulation time planned for the models created
from the detailed diagram. For example, wind speed and
direction over many annual cycles would be important
variables to model the development and maintenance
of the dunes in Great Sand Dunes National Park and
Preserve. Although this detailed information about the
wind is implied by the wind source symbol, it is not
included in this highly aggregated depiction of the SLB.
Used energy leaves the lower boundary of llie box
delineating the SLB system. No other flows exit from
the lower boundary and the heat sink symbol, which
receives the energy dissipated by all components
and flows in the system, including the background
temperature of the environment. The inclusion of
temperature as part of the specification of energy systems
models allows entropy calculations to be performed.
The major physiographic systems within the SLB are
aggregated into three classes: 1) mountains which are
to a large extent forested, 2) the valley floor, which
is divided into an area covered by the natural shrub
vegetation of the region including phreatophytes and
xerophytes (e.g., sagebrush), and 3) agricultural land,
which is generally irrigated with surface or groundwater
(Emery 1996). The primary economic systems of the
region are centered around its natural resources (e.g.,
food processing and shipment, mining and processing
ores and building materials, and recreation arc important
activities tied to natural resources). In addition, power
companies recently have recognized the high solar power
potential of the region (SLVDRG 2007). For example, a
solar power plant that was under construction during the
course of our study is now operational. However, during
Hie time of this study, all the electricity used in the SLB
was imported.
5.3.2 Aggregate Diagram
Aggregated ESL diagrams of Hie SLB were used to
evaluate the SLB each year under study. They show-
how emergy from renewable resources and local
nonrenewable resources (including renewable resources
like soil and ground water that were being used in a
nonrenewable manner) interacted with imported goods
and sendees to support regional economic activity
and produce exports (Fig. 5.2, Table 5.3). Although
Hie major social, economic, and environmental flows
were evaluated, there were several gaps in the data.
For example, we were not able to find data needed to
document internal mineral resources (i.e., primarily
sand and gravel and crushed stone) used in the system,
thus this value was estimated by assuming 10% of the
total resources remained in the region. In addition, we
were not able to apply the method utilized by Campbell
et al. (2005) to determine the import and export of pure
services to and from the SLB. Thus, (here is a "?" in the
places designated by P,L, P2E,, and E( on the diagram
(Fig. 5.2).
5.3.3 Total Emergy Percent
Data used to compute total emergy used and the
renewable emergy used were available from 1995 to
2005. These data are the main components of our
sustainability metric based on EmA. During this 11-
year period, the overall trend indicates little change in
-------
total emergy use and a slight decrease in the renewable
emergy used (Fig. 5.3). Total emergy use per person
rose 8% from 1995 to 2000 and then fell 14.7% from
its peak in 2000 to 2005 with the largest year to year
decrease (9.8%) occurring from 2002 to 2003 (Fig.
5.4). Total outflows exceeded inflows until 2001,
when the two were almost equal (Fig. 5.5). In 2000.
outflows began to decline more rapidly than inflows
and as a result, inflows slightly exceeded outflows in
2003; however, in 2004 and 2005 outflows increased
rapidly to more than 2.5 times the inflows (Fig 5.5). The
components of total emergy inflow determine the total
emergy used, which is defined as the total emergy inflow
plus the emergy from local sources (Fig. 5.6). The
increase in emergy use by the system seen from 2003 to
2005 followed similarly timed increases in the emergy of
renewable sources and imports.
Both the emergy used from local renewable sources
and the fraction of total emergy used from home
sources (those within the SLB region) were in a slight
declining trend from 1997 to 2001, but the fraction of
the local renewable sources was more variable (Fig. 5.7).
Renewable emergy available per capita declined from
1980 to 2005 such that the renewable emergy per person
in 2005 was only around 75% of that available in 1980
(Fig. 5.8). Further, there was up to a 22% year-to-year
(2001-2002) decrease in the renewable emergy per capita
along its declining trend due to variations in rain and
snowfall.
5.3.4 Summary of Renewable and
Non-renewable Emergy
The geopotcntial energy of snow was the largest emergy
inflow into the system over the 26-years examined (Figs.
5.9 and 5.10). Agricultural production is considered to
be fundamentally renewable in most emergy analyses
(i.e., renewable inputs of rain, wind and soil fertility
can support agriculture without large subsides of
nonrenewable emergy). We recognize that nonrenewable
emergy is a major input to most agriculture in the US,
but this study docs not include a detailed analysis of
the agricultural subsystems of the SLB. Emergy in
livestock, primarily cattle, was the largest agricultural
product of the SLB from 1980 until 1997 (Fig.
5.11). Wheat accounted for the largest emergy of
crop production from 1980 to 1985 (Fig. 5.11). Hay
accounted for the largest emergy from crop production
from 1986 until the conclusion of the study period,
except in 1989, when wheat once again was briefly
dominant (Fig. 5.11).
Data on nonrenewable emergy production in the SLB
(crude oil, metallic ores, broken stone or riprap, sand and
gravel, and non-metallic minerals) were taken from the
GI dataset, which was for the years 1995 to 2005. The
use of nonrenewable emergy in fuels and electricity in
the SLB remained constant over the study period (Fig.
5.12). However, nonrenewable emergy production
fluctuated, with emergy of sand and gravel and broken
stone/riprap in a declining trend from 1995 to 2005, but
showing intermittent spikes in production (Fig. 5.12).
5.3.5 Summary of Import-Export Exchange
From 1995-2005, exported emergy exceeded imported
emergy, with the most dramatic increase in the difference
from 2003 to 2005 (Fig. 5.13). In addition, there was
a dramatic spike in net immigration in 1988, 1989 and
1990, and a smaller surge of people into the SLB in 2004
and 2005 (Fig. 5-E.2).
5.3.6 Emergy
Data on petroleum, coal, natural gas and electricity
imported, and the sendees required to deliver these
energy inputs were collected for 1980 to 2005 (Fig.
5.14). Data were available from 1995 to 2005 for
material imports other than fuels, services in materials
(other than fuels), and tourism (Fig. 5.14, see Appendix
5-E for detailed data on Imports). "Materials other
than Fuels" carried the largest amount of emergy into
the SLB over the period for which we had data. The
emergy in these materials was more than twice as large
as the next largest input, which was the service required
to deliver the material inputs. From 1995 to 2005,
Cinergy of "Agricultural Goods." principally fertilizers
and agricultural chemicals, increased from about 18%
to 38% of material imports, whereas "Construction
Goods" were around 22% of imports from 1995 to
1999, but then declined to 13% of total imports in 2005
(Fig. 5.15). Emergy of "Industrial and Mining Goods"
was about 35% of inputs from 1995 to 1998 and then
declined to 24% in 1999, after which it rapidly rose to a
peak of around 44% of inflows from 2000 to 2001, led
by emergy of industrial gases and electrometallurgical
products, respectively (Fig. 5.15). After 2001, emergy of
"Industrial and Mining Goods" fell to 27.2% of inflows
in 2005. Over the 11-year period, emergy of "Consumer
Goods" remained around 25% of the emergy of material
imports until 2000 when it fell to around 17% of inflows
in 2001 and 2002. After 2002, the import of the emergy
of "Consumer Goods" recovered to its former level
before declining to 22% of the input in 2005.
5.3.7 Exported Emergy
Data on agricultural crops and livestock were available
from 1980-2005 and data on emigration were available
from 1985-2005. Remaining export data were obtained
from the GI dalaset and were only available from 1995
to 2005. If we assume that almost all agricultural
-------
production was exported, livestock was the major
agricultural export of emergy from the region until 1996
when hay (Table 5-E.2) became the dominant export of
emergy (Fig. 5.16). From 1995 until 2003, "Gravel or
Sand" was the largest emergy export from the region
(Table 5-E.5).
The overall pattern of major exports of emergy showed
a declining trend from 1995 to 2003, after which there
was a dramatic resurgence of the annual empower
exported (Fig. 5.16). The resurgence of emergy exports
was lead by total materials and by "All Other Materials"
(i.e., materials without agricultural crops, livestock, and
mineral and forest products; Fig. 5.17). The emergy
in services in material exports that were required to
make them followed the trend of materials, but was of
smaller magnitude, except for the services in "All Oilier
Materials," which exceeded the emergy of the materials
during the export resurgence from 2003-2005 (Fig.
5.17). From 2003 to 2004, there was a steep increase
in the emergy exported in the "All Other Materials"
category, such that it was three times greater than
the emergy exported in sand and gravel in 2004 and
2005 (Fig. 5.17). For the period from 1995 to 2000,
"Miscellaneous (Misc.) Food Preparations, nee (not
elsewhere classified)" was the largest export of emergy
from the SLB, but in 2001 "Misc. Printed Matter" was
the largest component briefly until in 2002 when "Misc.
Agricultural Chemicals" became the dominant export
(Fig. 5.18). After falling to its lowest point in 2003,
emergy of "All Other Materials" exported had a sharp
jump upward in 2004 lead by emergy of "Miscellaneous
Agricultural Chemicals." "Miscellaneous Agricultural
Chemicals" includes such products as insecticides,
herbicides, soda ash and sulfur, soil conditioners,
and mulch. While all "other" materials exported
increased from 2003 to 2005 (Fig. 5.18), the increase
in "Miscellaneous Agricultural Chemicals" exported
was six times greater than that of the next largest export
("Misc. Printed Matter"; Fig. 5.18).
5.3,8 Population and per Indices
The population of the SLB grew steadily from 1980
to 2005, increasing nearly 15% during the period and
there was a corresponding increase in the emergy of
electricity and fuel use per capita during the same period
(Fig. 5.19). The rate of growth in the use of both fuel
and electricity use increased after 1992 (Fig. 5-G.3). Per
capita use of emergy from fuels and electricity declined
from 1980 to 1983 and then began to increase in two
stages (1983-1997 and 1997-2005) with some variation
(±3%) within a stage (Fig. 5.19). The most rapid growth
in per capita energy use occurred between 1999 and
2001, when it increased by approximately 10%.
5.4
Emergy analyses that consider how a system changes
over many years are becoming more common.
Historical studies of nations have been performed (e.g.,
Abel 2007, Rydberg and Jansen 2002, Sundberg et al.
1994, Tilley 2006) and nations have been evaluated over
decades (e.g.. Cialani et al. 2005, Ferreyra and Brown
2007, Hagstrom and Nilsson 2005, Huang et al. 2006).
Also, Pulselli et al. (2008) studied the sustainability of
two regional systems in Italy over 32 years using the
Index of Sustainable Economic Welfare and emergy
analysis; however, studies of the sustainability of
regional environmental systems with nonstandard
boundaries (i.e., boundaries that were not based on
standard political or administrative reporting units)
carried over many years appear to be rare. The SLB, as
defined for this study, is such a nonstandard region.
5.4.1 Spatial Boundaries
This study is unique in that the major organizing
renewable energy flows for the region are the available
geopotential and chemical potential energy in rain
and snow. Although the transformity of snow was
determined by Campbell and Ohrt (2009), this is the first
EmA where snow was shown to account for the largest
renewable emergy input to a system. The variation in the
geopotential energy of snow is the primary driver behind
the fluctuations in total geopotential energy inflow and
the renewable emergy base of the SLB system over
the 26-year period of the study (Fig. 5.8). The present
human system in the SLB is largely organized around
agriculture, which is dependent on the surface and
ground water flows that originate in the high mountain
snowpack (Emery 1996). Thus, as mentioned above,
the boundaries for the analysis of this nonstandard
region might be logically set to conform to those of the
Upper Rio Grande River Basin, or to conform to the
political economic reporting units based on the county
boundaries. We chose to deal with the initial boundary
problem by choosing those boundaries that were
appropriate for the process under consideration (i.e.,
economic and social data were collected for counties,
whereas water flows were determined by the Upper Rio
Grande River Basin).
5.4.2 Availability
Another major obstacle was finding a complete set of
data for the local region that we could use. Technical
problems related to determining the inflow and outflow
of commodities to and from the region that had been
solved at the state level (Campbell et al. 2005. Campbell
and Ohrt 2009) had to be revisited in the context of
the SLB. Of all the data needed to perform an EmA
for the region, information on the import and export
-------
of commodities was the most difficult to obtain over
the entire period of interest. Once we found a source
from which these data could be purchased for a subset
of years, we found that it contained inconsistencies
and uncertainty greater than data from the U.S. Census
Bureau Commodity Flow Survey.
5.4.3 Renewable Inputs to the
Ordinarily, the renewable emergy base of a system is
determined by averaging the independent inputs of
emergy to the system over many years. This rule is
applied because societal structures have many years
to adjust to the emergy signature of a location (i.e.,
economic activities adjust to take advantage of the
amounts of available energy of various kinds in the
region and their expected availability). This study
departs from this custom by looking at the year-to-year
variability in the availability' of renewable resources.
For our study of the SLB, where agriculture and the
region's unique natural systems are highly dependent
on the year-to-year availability of water resources, we
think the inclusion of this level of detail is justifiable.
Evapotranspiration (ET) appears nearly constant because
we had only one composite annual estimate for the ET
of each vegetation class and thus its variability' depends
on the change in land use over the 26-year period
(Appendix 5-E). If we had annual estimates of ET, we
would expect it to vary with the hydrological cycle in a
manner similar to precipitation.
5.4.4 Temporal Patterns of Development
and the Water Cycle
Until 2002, emergy of production from renewable
sources (e.g., crops, livestock, and forests) was more
than twice the emergy of total fossil fuel and electricity
use including services. After 2002, this ratio fell due
to a marked decline in the total emergy of agricultural
production and the steady increase in the emergy
supporting society. The salient characteristic of the
renewable emergy base for Hie SLB is the geopotential
energy expended due to snowmelt and its subsequent
runoff, which is the largest renewable emergy input to
the region. However, the work done by the snow occurs
after a period of storage; thus, there is a lag between the
time the work is delivered from the time the input enters
the system. Snowmelt delivers high quality energy
in a pulse that carries twice the emergy of the next
largest input. ET. This pulse is essential to maintaining
rare geological features like the sand dunes and for
recharging groundwater. If climate change results
in drier winters or in wanner winter temperatures as
indicated by climate change predictions for the region, as
evidenced by State of New Mexico (2005), the formation
and duration of the snow pack may be affected. Changes
in the long-term snowfall or winter temperatures could
result in significant changes in the hydrological and
geological features of this unique system and have
effects on agricultural production.
5.4.5 Nonrenewable Resources
The use of nonrenewable resources, such as, fossil
fuels and electricity' in the SLB showed a steady growth
trend over the period examined that was proportional to
population growth. The construction material (sand and
gravel and broken stone) produced in the SLB declined
by about 40% from 1995 to 2005. Miscellaneous
norurietallic minerals in unprocessed form were mined
in the SLB but apparently were not used there. Exports
in this category precipitously increased from 2003 to
2005, and were part of a general economic resurgence
that occurred in the region during this time. Soil erosion
from wind steadily declined over the period (Fig.
5.20 and Appendix 4-A). likely due to improvement
in cultivation methods. The use of groundwater for
irrigation was episodic and it corresponds with periods of
decreased precipitation in the region. Peak groundwater
withdrawals exceeding the recharge rate occurred with
increasing frequency over the 26-year period with the
greatest decline occurring in 2002. During times of
peak withdrawal, the emergy supplied by groundwater
use was a significant contribution to the annual emergy
used in the region, and this emergy source approached
that supplied by fuels and electricity' in 2002 (Fig. 5.12).
Over-withdrawals of groundwater (i.e., faster than it can
recharge), could threaten the water resource base for
agriculture in the region and the sustainability of this
activity in the SLB.
5.4.6 Patterns of Import and the Emergy Budget
of the SLB
Material goods oilier than energy account for the largest
part of total imports from 1995 to 2005. Processed
nonmetallic minerals accounted for most of the
emergy in imported material goods excluding fuels
and electricity. However, imports in this category fell
45% over the 11-year period examined (Appendix
5-E). Perlite is classified as a processed nonmetallic
mineral used for construction and agricultural purposes.
Because large quantities of perlite entered the SLB
(San Luis Valley Development Resources Group
2007) and there is a defunct perlite mine in Conejos
County, we assumed that materials moving in the
nonmetallic minerals category would be largely perlite.
Perlite is both imported into and exported from die
SLB. however, according to the GI data, much more
is imported than exported. We were puzzled by this
difference and what might have happened to the missing
mass or, alternatively, if the GI data were incorrect.
-------
If the difference was consumed in the SLB, it would
greatly increase the eniergy used there. The apparent
consumption of perlite declined about 40% from 1995
to 2005. Further investigation revealed that perlite
was trucked into the SLB from mines in northern New
Mexico to the railhead at Antonito, where it was mixed
to order and placed on rail cars for shipment out of the
region (San Luis Valley Development Resources Group
2007). Furthermore, the mines in northern New Mexico
supplied 80-85% of the perlite from US sources during
the late 1990s (Barker et al. 2002), so these material
flows pertain to the larger system and not to the SLB. A
similar condition applies to scoria, another processed
nonmetallic mineral, and thus, the eniergy of processed
nonmetallic minerals was removed from the emergy
import-export balance for the SLB.
The emergy of "Agricultural Goods" increased 269%
from 1995 to 2005 suggesting increased use of fertilizer
and agricultural chemicals in the region and by extension
increased cultivation of certain crops. During this
period, BEA (2009) reports the total expenses for
fertilizer and lime increased. "Industrial and Mining
Goods" imported increased two times from 1999 to
2002 and then declined to the 1995 level by 2005, which
may be indicative of increased industrial activity in the
region during this time. The rapid increase in imports
in this category was due to industrial gases (2000-2002)
and electrometallurgical products (2002). The import
of "Consumer Goods" (i.e.. imported commodities
classified as primarily for use by consumers) remained
constant from 1995 to 2000 and then declined to 67%
of tills level in 2001 and 2002 and recovered to 86% of
the 1995-2000 plateaus by 2005. This decrease in the
imported consumer goods occurred in spite of population
growth, which was 10% for this period, indicating that
consumers in the SLB may have decreased their per
capita utilization of imported consumer goods during
this time.
5,4.7 Patterns of Export
Exports of most materials were in a declining trend
from 1995 to 2003, when total materials exported
was at a minimum for the 11 -year period. From 2003
to 2005, the emergy exported in materials increased
markedly due to a general resurgence of the export
economy as indicated by the fact that exports of "All
Other Materials" increased in all categories during this
time. These data show the category "Miscellaneous
Agricultural Chemicals" led this resurgence (Fig. 5.18).
However, we do not know the exact commodity or
industrial activity that accounts for these exports. The
miscellaneous agricultural chemicals commodity class
includes commodities such as pesticides, herbicides,
fungicides, defoliates, soil conditioners, mulch, sulfur
and soda ash. From 1995 to 1998, exports had a slightly
increasing trend in emergy due to the manufacture of
"Misc. Food Preparations," (i.e., processed agricultural
products such as french fries from potatoes) which was
the largest category of all other exports until 2003. We
assumed that almost all the agricultural production in the
SLB was exported so the discussion above pertaining to
the patterns of renewable production applies to exports.
5.4.8 Export-Import Exchange
The fact that exports exceeded imports by a considerable
margin over the entire study period indicates the SLB
is a hinterland supplying raw materials and primary
products to other areas in North America as shown by
the GI data. Exports and imports have different temporal
patterns prior to 2003 (Fig. 5.13), but both reach a low
point at tliis time. From 2000 to 2003, both decline
together, suggesting the import-export sectors of the
economy of the SLB were contracting during that time.
From 2003 to 2005, exports increased precipitously to
the highest level obtained during the 11-year period,
while imports increased very slightly.
5.4.9 Patterns of Population Growth
Consumption
Fuel use in and fuel import to the SLB increased 16%
over the 26-year period and electricity use increased
by 80% compared to a 25% increase in population.
Thus, we must look for factors other than population
growth to explain the increase in energy use. From
1980 to 2005, the consumption of electricity per capita
increased by 44%, which accounts for the additional
growth in electricity consumption. Because of its high
transformity, the use of more electricity by people may
be indicative of an increase in their quality of life over
this time. Per capita fuel consumption declined 7%
over the period but the concomitant population increase
resulted in a 16% rise in fuel consumption. From 1995
to 2005, the population increased 9.8% as the emergy of
the fuels and electricity used increased 20.7% or 28.8%
(services included). Without considering sendees, the
emergy of fuel consumption increased 19.2% as the
emergy of electricity use increased 23.4% from 1995
to 2005 (Fig. 5.19). Increased consumption of fuels
(+8%) and electricity (+12%), along with population
growth, may contribute to the observed increases in
consumption.
The SLB has a very high proportion of renewable
emergy (31-51% of total emergy use) supporting
the system and its people compared to renewable
contributions of around 3% for many developed areas
in the US (Campbell et al. 2005, Campbell and Olirt
-------
2009). Population growth in the SLB had the effect
of decreasing the renewable emergy available per
person (Appendix 5-E). and of putting pressure on the
environment from development and waste production.
In the end, this process is self-limiting, but at the present
rate of increase, it may be some time before the inflow
of people is diminished by degrading natural resources.
Indeed, other factors related to economic development
may continue to draw more people to the region in
the future. The overall quality of life for people in the
SLB as indicated by the Total Emergy Used per person.
which is high (Fig. 5.4) with values equal to those in
Minnesota, a state with high levels of emergy use per
capita compared to West Virginia and the U.S. as a whole
(Campbell and Ohrt 2009).
5.4.10 Analysis of the Emergy Indices
The fraction of emergy use from renewable sources
can be viewed as an indicator of movement toward or
away from sustainability (see Fig. 5.7). If the renewable
emergy is of local origin, it is an indicator of the self-
sufficiency of the system. The fraction of energy use
from locally renewable sources declined from 1997 to
2002 (Figs. 5.6, 5.21). After 2002, it increased abruptly
indicating movement toward sustainability (or perhaps
more accurately away from unsustainability), which was
the result of the excess ground water use that occurred
in 2002. This increase was due both to an increase in
precipitation and to a relatively low level of emergy
imported. On average, renewable emergy accounted
for 44% of total emergy use over the 11-year period
examined, which is five times that of an average location
in the US (Campbell and Ohrt 2009). However, this
index showed a 4% decline over the 11-year period
examined, indicating the region was moving away from
sustainability.
The fraction of emergy derived from home sources has
a pattern similar to, but not the same as the emergy
from renewable sources. As demonstrated in 2000 and
2002, the principal difference between these patterns is
determined by the years of drought when local renewable
resource contributions decline, (along with decreased
precipitation) and the fraction of use from home sources
increases due to increased use of groundwater (compare
Figs. 5.9, 5.20, and 5.21). The use of ground water in
this manner decreases the variability of total emergy use
in the system by compensating for declines in emergy
inflows from the renewable resource base. The fraction
of emergy used from home sources is an indicator of
self-sufficiency and in the SLB home sources exceed
40% in some years making it moderately self-sufficient.
However, because the region has a high potential for
developing solar, wind, and gcothcnnal power, the
potential self-sufficiency of the SLB using renewable
sources alone is large. The principle of matching
environmental emergy with appropriate purchased or
economic inputs in the form of investment implies that
we would expect economic investment from outside
to be drawn into the region. Per capita income in
the region is low, so the necessary investment capital
probably is not available within the region itself. Thus,
we might expect tension to arise between the desire
of outside influence from the next larger system to
obtain ''green" energy and the desire of local people for
self-determination of their lifestyle and energy self-
sufficiency.
One index of how well the local system is doing is the
emergy used annually by the system. The emergy used
in the SLB remained fairly constant over the entire
period (the coefficient of variation was 0.05) despite
variability in the emergy inputs. Because the use of
nonrenewable emergy from inside the system can
be used to compensate for declines in the renewable
emergy inflows in most years, the pattern of total
emergy use in the system appears to follow closely
the third component, imported emergy. Systems at all
levels of organization compete for the available energy
in resources. Emergy matching of a high with a low
transformity resource develops the most emergy flow.
Tims, imports and exports are continually adjusting to
meet the need to increase emergy flow at all levels of
organization under varying circumstances. In theory, the
entire hierarchical, coupled ecological-economic system
continually seeks win-win relationships that increase
system empower. This concept is somewhat parallel
to the general equilibrium model in economics where
supply and demand are brought into relationship through
a set of equilibrium prices. The objective function is
different and the method of calculation is not determined
in terms of emergy matching, but the two ideas arc
theoretically similar. Lu and Campbell (2009), Lu et al.
(2006). and Lu et al. (2007) discussed additional indices
that may be useful for answering particular questions
about the relationship between emergy and economic
flows between a local system and its larger system.
5.4,11 of Missing Unknown
Information, Uncertainty
The commodity flow data taken by the US Census
Bureau met the data quality objectives for measuring
total commodity movements for the studies of West
Virginia and Minnesota (Campbell ct al. 2005, Campbell
and Ohrt 2009). In these studies, the estimated
movements of individual commodities had higher
uncertainties than the average of all commodities. The
GI data were critical in allowing us to apply the revised
-------
method (Campbell et al. 2005) for calculating emergy
import and export to and from a region. The methods
reported by GI appear to be logical and deliberate, even
though there is considerable uncertainty associated with
some of their measurements. Principally, the export of
"nonmetallic minerals, processed" appears to be grossly
underestimated (see Appendix 5-F) and others were
questionable.
The transformities and specific emcrgics by which
the energy or mass flows are multiplied, respectively,
to obtain emergy are critical numbers in the analysis
(Appendix 5-C) and they are important to understand
the general level of uncertainty associated with these
numbers. Campbell (2003) analyzed five global water
budgets, and determined the transformities of global
hydrological flows, such as rain, evapoiranspiration,
and river flow, within an average standard deviation
of 5.9% of the mean value. Odum (1996) determined
the transformiry of coal from its relative efficiency
in producing electricity and its geological production
process. The former method gave an estimate of 4.3E+4
sej/J and the latter 3.4 E+4 sej/J. The two values are
within 12% of the mean value, which may be a rough
estimate of the model uncertainty in determining
transformities.
5.4.12 Strengths and Limitations of the Method
Emergy methods are comprehensive and based on
thcrmodynamic laws and principles governing the order
and organization of all systems; therefore, they have a
broad capability to address many problems. In addition,
Energy Systems Theory is a mela-lheory in the sense
that theory and methods from all other disciplines that
are translated into Energy Systems Language and further
analyzed within the context of the thermodynamic
constraints that govern them. This is true because all
real systems have an underlying energy basis. However,
EmA is not a substitute for other approaches but a
complement to them. Perhaps the greatest strength of
EmA is that it can transform disparate quantities into
similar quantities with the same units that are then
directly comparable for decision-making.
The strengths of EmA are also its weaknesses. Because
it is comprehensive, it often requires more data and
information than other approaches. In addition, the
more aspects of a system that one considers, the more
complicated the emergy work can become, because it is
forced by the imperative to be complete to incorporate
all relevant data and theory into the emergy assessment.
In addition, the uncertainty of the analysis grows with
the complexity of data and models used. One of the
greatest challenges in an emergy assessment is to set
the proper boundaries for the study so that a particular
set of research questions are answerable. Perhaps the
greatest weakness of emergy is that it is an abstract
concept that can be difficult to communicate to scientists
and managers. To use it with facility requires years of
study and a broad background in the natural and social
sciences.
5.5 Conclusions
The process of assembling information on the SLB
and diagramming the system demonstrated that it is
a fully developed agricultural system with crop and
livestock production, food processing, and animal waste
processing components. This system is a high desert
and may be under additional pressure due to climate
change if drier, warmer years become more frequent in
the future, which could threaten unique natural features
like the sand dunes, local wetlands, and wildlife refuges.
Farming in the region may come under pressure and
the area irrigated may have to contract to meet the
diminished capacity of the annual water budget. The
alternative of drawing excess water for agriculture from
groundwater storage may be a short-term solution to deal
with shortages. Long term, however, such a solution
poses a potential recipe for disaster, not only for the
agricultural system, but also for the SLB as a whole.
Examination of Fraction of Renewable Emergy to
Total Emergy Used revealed movement away from
sustainability. Renewable emergy use declined 4% over
the 11-year study period, and only about half of the
present population could be supported at their standard
of living on renewable emergy alone, indicating that the
region as a whole is moving away from sustainability.
Social and economic systems in the SLB have been
growing over the period examined in this study and
we expect that this trend will continue in the future as
the region becomes a center for the development of
renewable energy. The EmA showed there are pulses
within the general growth trend of the economy and the
emergy exported experienced declines followed by rapid
increases that make the overall trend irregular. Even
though the SLB receives 31-51% of its emergy base
from renewable sources, it depends on the larger system
for the purchase of the majority of the emergy used; and
therefore, it is subject to larger national trends. Perhaps,
the national and global trend that is most likely to affect
the SLB in the near future is the current movement
toward developing renewable energy sources. The SLB
has a rich potential for future development of solar,
wind, and geothermal energy and could rely more on
renewable sources of power in the future.
-------
5.6
Abel, T. 2007. World-systems as complex human
ecosystems, in: Hornborg. A., Crumley, C. (Eds.),
The World System and the Earth System: Global
Socioenvironmental Change and Sustainability
Since the Neolithic. Left Coast Press, Walnut Creek,
California, pp. 56-73.
Barker, J.M., Austin, G.S., Santini, K., Alatorre, A.
2002. Perlite deposits and market trends in North
America., in: Scott, P.W., Bristow, C.M. (Eds.),
Industrial Minerals and Extractive Industry Geology.
Geological Society, London, pp. 73-82.
BEA (Bureau of Economic Analyses). 2009. National
Income and Products Accounts Tables. Table CA45
Farm Income and Expenses. US Department of
Commerce, Washington, DC. http://www.bea.gov/
regional/reis/Last Accessed August 14, 2009.
Brown, M.T. 2003. Resource Imperialism: Emergy
perspectives on sustainability, balancing the welfare
of nations and international trade, in: Ulgiati, S.
(Ed.), Advances in Energy Studies. Proceeding of the
conference held in Porto Venere, Italy, October 2002.
University of Siena, Italy.
Campbell, D.E. 1998. Emergy analysis of human
carrying capacity and regional sustainability: an
example using the State of Maine. Environmental
Monitoring and Assessment 51, 531-569.
Campbell, D.E. 2001. Proposal for including what is
valuable to ecosystems in environmental assessments.
Environmental Science and Technology 35, 2867-
2873.
Campbell, D.E. 2003. A note on the uncertainty in
estimates of transformities based on global water
budgets, in: Brown, M.T, Odum, H.T., Tilley, D.R.,
Ulgiati, S. (Eds.), Emergy Synthesis 2. Proceedings
of the Second Biennial Emergy Analysis Conference.
Center for Environmental Policy, University of
Florida, Gainesville, Florida, pp. 349-353.
Campbell, D. E., Brandt-Williams, S., McischM. 2005.
Environmental accounting using emergy: evaluation of
the state of West Virginia. EPA600/R-05/006. United
States Environmental Protection Agency, NarragansetU
Rhode Island.
Campbell, D.E., Galloway, W.B., Pesch, G.G. 1998.
Appreciative Inquiry: A mechanism for maximizing
empower in social systems. In Allen. J.K, Wilby, i.
(Eds.) Proceedings of the 42nd annual conference of the
International Society for the Systems Sciences, ISBN
0-9664183-01.
Campbell, D.E., Ohrt, A. 2009. Environmental
accounting using emergy: evaluation of Minnesota.
USEPA Project Report, EPA/600/R-09/002, 138 pp.
Cialani, C., Russi, D., Ulgiati, S. 2005. Investigating
a 20-year national economic dynamics by means
of emergy-based indicators, in: Brown, M.T,
Bardi, E., Campbell, D., Comar, V., Huang, S.L.,
Rydberg, T., Tilley, D., Ulgiati, S. (Eds.), Emergy
Synthesis. 3-Theory and Applications of the Emergy
Methodology, Center for Environmental Policy,
University of Florida, USA, pp. 401-416.
Emery, P. A. 1996. Hydrology of the San Luis Valley,
Colorado, an overview and a look at the future, in:
Thompson, R.A., Hudson, M. R., Pillmore, C.L.
(Eds.), Geologic Excursions to the Rock}' Mountains
and Beyond. Field Trip Guidebook for the 1996
Annual Meeting, Geologic Society' of America; Oct.
28 1996-Oct. 31 1996; Denver, Colorado. Colorado
Geological Survey, Department of Natural Resources,
Denver, CO. Available on line at http://www nps.
gov/arcnive/grsa/resources/docs/Trip2023.pdf. Last
accessed August 4, 2010.
Ferreyra, C., Brown, M.T. 2007. Emergy perspectives
on the Argentine economy during the twentieth
century: a tale of natural resources, exports and
external debt. International Journal of Environment
and Sustainable Development 6, 17-35.
Global Insight. 2009. TRANSEARCH Insight Freight
Flow Database. Lexington. Massachusetts 02142-
3158, USA.
Hagstrom, P., Nilsson, P.O. 2005. Emergy evaluation of
Swedish economy since the 1950s, in: Brown, M.T.,
Campbell, D., Comar, V, Huang, S.-L., Rydberg, T,
Tilley, D., Ulgiati, S. (Eds.), Emergy Synthesis, vol. 3.
Proceedings of the 3rd Biennial International Emergy
Research Conference. The Center for Environmental
Policy. University of Florida, Gainesville, Florida.
Huang, S.L., Lee, C.L., Chen, C.W. 2006.
Socioeconomic metabolism in Taiwan: emergy
synthesis versus material flow analysis. Resources,
Conservation and Recycling 48, 166-196.
Lambert, D. 1.999. A spatial emergy model for Alachua
County, Florida. PhD Dissertation, University of
Florida, Gainesville.
Lefroy, E.G., Rydberg, T. 2003. Emergy analysis of
three cropping systems in south western Australia.
Ecological Modelling 161. 195-211.
-------
Lotka, A.J. 1922a. Contribution to the energetics of
evolution. Proceedings of the National Academies of
Sciences USA 8, 147-151.
Lotka. A.I 19225. Natural selection as a physical
principle. Proceedings of the National Academies of
Sciences USA 8, 151-4.
Lu, H.F., Campbell, D.E. 2009. Ecological and
economic dynamics of the Shunde agricultural system
under China's small city development strategy.
Journal of Environmental Management 90. 2589-2600.
Lu, H.F., Campbell D.E., Li, Z-A, Ren H. 2006.
Emergy synthesis of an agro-forest restoration system
in lower subtropical China. Ecological Engineering
27, 175-192.
Lu, H.F., Campbell, D.E., Chen, I, Qin, P., Ren, H.
2007. Conservation and economic vitality of nature
reserves: an emergy evaluation of the Yancheng
Biosphere Reserve. Biological Conservation 139.
415-438.
Odum, H.T. 1994. Ecological and General Systems.
University Press of Colorado, Niwot, Colorado.
Odum, H.T. 1996. Environmental Accounting: Emergy
and Environmental Decision Making. John Wiley &
Sons, New York.
Odum, W.P., Odum, E.P., Odum, H.T. 1995. Nature's
pulsing paradigm. Estuaries 18,547-555.
Pulselli, F.M.. Bastianoni, S., Marchettini, N., Tiezzi,
E. 2008. The Road to Sustainability. WIT Press,
Southhampton, UK.
Rydbcrg, T, Janscn, J., 2002. Comparison of horse and
tractor traction using emergy analysis. Ecological
Engineering 19, 13-28.
San Luis Valley Development Resources Group. 2007.
Comprehensive economic development strategy.
Available online at http://www.slvdrg.org/ceds.php .
Last accessed June 17, 2009.
Sundbcrg, U., Lindegren, J., Odum, H.T, Doherty,
S. 1994. Forest energy basis for Swedish power in
the 17th century. Scandinavian Journal of Forest-
Research. Supplement (Sweden), 50 pp.
State of New Mexico. 2005. Potential Effects of
Climate change on New Mexico. Agency Technical
Work Group, State of New Mexico. Available online
at http://www nmenv.state nm.us/aqb/cc/Potential_
Effects__Climate___Change___NM.pdf. Last accessed June
11.2010
Tiezzi, E., Bastianoni, S. 2008. Sustainability of the
Siena Province through ecodynamic indicators.
Journal of Environmental Management 86, 329-331.
Tilley, D.R. 1999. Emergy basis of forest systems.
Ph.D. dissertation, University of Florida, Gainesville,
Florida, 296 pp.
Tilley, D.R. 2006. National metabolism and
communication technology development in the United
States: 1790 to 2000. Environment and History 12,
165-190.
Ulgiati, S., Odum, H.T., Bastianoni, S. 1994. Emergy
use, environmental loading and Sustainability: an
emergy analysis of Italy. Ecological Modelling 73,
215-268.
USD A. 2009. USDA National Nutrient Database for
Standard Reference, Release 21. http://www.nal.usda.
gov/fnic/foodcomp/search. Last accessed August 17,
2009.
-------
Table 5.1- Tabular format for an emergy analysis, providing a template for the calculation of the emergy values.
Column 1
Note
Column 2
Item
Column 3
Data
J,g.S
Column 4
Solar Emergy/Unit
sej/J. sej/g. sej/S
Column 5
Solar Emergy
sej, sej/y
Table 5.2 - Definition of pathway flows for the systems model of the San Luis Basin (seven county area). Colorado
shown in Fig. 5.1.
Pathway
RO
R,
k0
k,
k.
k3
K
\
k6
k7
k8
k«
k10
kn
k,2
k,3
kw
"S5
"S6
k,7
k,s
k,9
ka
k21
k2.
k23
kM
k,,
k26
k,.
k28
kz,
k»
k*
k,
k33
k34
k33
k36
k37
k38
k»
k«
k4,
Definition of Flow
Albedo
Wind passing through the Region
Solar radiation absorbed by the Region
Solar radiation absorbed by agricultural crops
Solar radiation absorbed by natural vegetation (shrubs)
Solar radiation absorbed by forests
Wind energy absorbed by agricultural crops
Wind energy absorbed by natural vegetation (shrubs)
Wind energy absorbed by forests
Wind energy absorbed by sand flats and dunes
Rain and snow falling on agricultural crops
Rain and snow falling on natural vegetation (shrubs)
Rain and snow falling on forests
Snow falling in the mountains
River water flowing out of the SLB
Runoff to rivers from the forests
Infiltration from forests to groundwater
Infiltration from rivers in forestland to groundwater
Groundwater uptake by forest vegetation
Nutrient and rainwater uptake by the forest
Evapotranspiration by the forest
Snow melt feeding mountain rivers
Sublimation of snow
Groundwater loss to deeper levels
Evapotranspiration from rivers in mountains
\Vater use for mining and processing ores
Forest biomass growth
Timber harvest
River flow crucial to dunes formation
Erosion of mountains
Wind erosion of dunes
Groundwater flow from mountains to the valley
River flow from mountains to rivers in valley Shrub land
River flow from mountains to agricultural land
Runoff to rivers from the shrub land
Infiltration from shrub land to groundwater
Infiltration from lakes and rivers in shrub land to groundwater
Groundwater uptake by shrub land vegetation
Nutrient and rainwater uptake by the shrub vegetation
Evapotranspiration by the shrub vegetation
Shrub biomass growth
Evapotranspiration from rivers and lakes in Shrub land
Groundwater loss to deeper levels
Groundwater flow from agricultural areas to the closed basin
-------
Pathway
k«
k43
k44
*«
k«
k4,
k«
k«
kw
k«
k52
k«
k54
k,5
k56
k,7
k,B
k»
k60
k»
k62
K3
k,,
k63
k«
k67
k6!
k6S
k™
k71
k72
k73
k74
k,,
k,«
k,,
k7S
k7S
k!0
K,
k82
k83
k84
k85
k»
^
k8!
k8S
k90
k»,
k,2
k93
kM
Definition of Flow
Surface water use in the region
Water use by production manufacturing
Runoff to rivers from the agricultural land
Infiltration from agricultural land to groundwater
Infiltration from rivers in farmland to groundwater
Groundwater and river water used to irrigate crops
Groundwater loss to deeper levels
Groundwater flow to New Mexico
Water and nutrients taken up by crops
Crop biomass growth
Evapotranspiration by crops
Evapotranspiration of water in rivers
Livestock biomass growth
Crops eaten by livestock
Water used by livestock
Waste produced by livestock
Waste produced by crops
Livestock processed or shipped
Crops processed or shipped
Geologic processes building laudform and mineral deposits
Sand mined and processed
Crushed rock mined and processed
Minerals mined and processed
Electricity and fuels used by the mining industry
Goods and sendees used by the mining industry
Government control of mining
Mining industry inputs to manufacturing and construction
\Vaste produced by mining and processing ores
Knowledge and labor used in the mining industry
\Vater use by food processing
Water use by service and commerce
Water use by manufacturing and construction
Water use by the recreational systems
Fuels and electricity input to the recreational systems
Goods and services input to recreational systems
Government regulation of recreational systems
Human knowledge and labor used by recreational systems
Transport of fuels and electricity into the State
Transport of goods and sendees into the State
Government regulation of transportation
Human knowledge and labor used in the transportation sector
Goods and sendees input to the transportation sector
Fuels and electricity input to the transportation sector
Fuels and electricity used by the government sector
Goods and sendees input to the government sector
Human knowledge and labor used in the government sector
Federal government regulations
Federal taxes
Federal outlays
Money spent on fuels
Money spent on goods and services
Solar electricity generated in SLB joins the regional grid
Fuels and electricity input to the power distribution system
-------
Pathway
k,,
k«
k,7
k»
k9'l
kioo
kioi
kI02
kI03
k104
k105
k™
k,07
k,os
k109
kno
km
k112
k,13
k,14
k,15
ku.
ku,
ku,
kus
k,20
k,2,
k122
k123
i,
K124
kI2?
kIM
km
k128
k129
k130
k,3,
k,32
k,33
k134
k135
k136
k,37
k,3S
k,3,
k,«
kwl
kW2
k,43
k,44
k,4.
Defliiltlon of Flow
Goods and services input to power distribution system
Government regulation of power distribution system
Human knowledge and labor used by power distribution system
Fuels and electricity input to education systems
Goods and services input to education systems
Government regulation of education
Human knowledge and labor used in the schools
Teaching
Learning
Increase in human knowledge and skills
Loss of information (knowledge and skills)
Gain of knowledge and skills with immigrants
Loss of knowledge and skills with emigrants
Government regulation of people
Goods and services used by people and households
Fuels and electricity used by people and households
Water used by people and households
Waste produced by people and households
Immigration
Emigration
Raw and processed ores exported
Manufactured products exported
Raw and processed food exported
Fuels and electricity used by production and manufacturing
Goods and services used by production and manufacturing
Government regulation of industry
Human knowledge and labor used in manufacturing
Waste produced by industry
Fuels and electricity used by food processing
p|S " 1 I p .
Government regulation of food processing industry
Human knowledge and labor used in food processing
Food processing inputs to manufacturing
Waste produced by food processing
Production and manufacturing inputs to service and commerce
Food processing inputs to service and commerce
Fuels and electricity used by service and commerce
Goods and services used by service and commerce
Government regulation of service and commerce
Human knowledge and labor used in service and commerce
Service and commerce used by tourists
Exports from the service and commerce sector
Tourists entering the State
Tourists leaving the State
Money gained from the sale of products and sen-ices
Money spent by tourists
Effects of wastes on forests
Effects of wastes on shrub land
Effects of wastes on agricultural lands
Wastes leaving the Region in water or air
Residents using recreation and cultural resources
-------
Table 5.3 - Summary of annual emergy flows in the San Luis Basin, 1995-2005.
Note
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
86
87
88
89
90
91
92
93
94
95
96
Symbol
RA
RI
N
N'
NO
N,
N2
F
Fj
F2
G
I
Ij
I2
13
I,
I5
P2I
P2Ij
P2I2
PA
PA
PA
B
E
E,
E2
E3
E4
P2E
PA
PA
P2E3
X
P2
Item
Renewable sources used
Renewable Electric Produced
Nonrenewable source flows
Internal fuels and minerals extracted
Dispersed Rural Source
Concentrated use, minerals, hydroelectric)
Exported without Use
Imported Fuels, Minerals+U, Electric
Fuels and minerals used (F+Nj-Renewable&Electric)
In region minerals used (Fl- F)
Imported Goods (materials)
Dollars Paid for all Imports
Dollars Paid for Service in Fuels, Minerals, Electric
Dollars Paid for Service in Goods
Dollars Paid for Services
Dollars Spent by Tourists
Federal Transfer Payments
Imported Services Total
Imported Services in Fuels, Minerals, Electric
Imported Services in Goods without fuels
Imported Services
Emergy Purchased by Tourists
Net Emergy Purchased by Federal $
Exported Products without minerals
Dollars Paid for All Exports
Dollars Paid for Goods without minerals
Dollars Paid for Minerals Exported
Dollars Paid for Exported Services
Federal Taxes Paid
Exported Services Total
Exported Services in Goods other than minerals
Exported Services in Minerals
Exported services
Gross Regional Product
Emergy $ ratio for the US in 2000 sej/$
1995
Emergy
(E+20 sej)
32.4
0.0
12.6
47.6
1.2
11.4
43.2
6.9
11.4
4.4
13.3
8.1
1.9
6.3
0.0
1.4
0.0
27.4
20.7
18.4
2.3
0.0
2.4E+12
Dollars
(E+6 S)
312.5
71.8
240.7
0.0
55.3
23.5
826.8
736.8
90.1
0.0
0.0
1161.8
2000 EmS
(E+6 EmS)
1379.8
0.0
535.6
2025.2
52.6
483.1
1836.7
294.6
483.1
188.5
566.7
346.0
79.5
266.5
0.0
61.2
0.0
1164.0
0.8
0.1
0.0
1996
Emergy
(E+20 sej)
28.8
0.0
16.7
43.9
5.3
11.3
39.8
7.3
11.3
4.1
13.9
7.9
2.0
5.9
0.0
1.5
0.0
31.5
21.8
19.6
2.2
0.0
2.4E+12
Dollars
(E+6 S)
304.9
76.8
228.1
0.0
57.4
25.1
861.3
inn
83.6
0.0
0.0
1229.6
2000 EmS
(E+6 EmS)
1227.1
0.0
710.0
1869.6
227.1
482.9
1695.7
309.0
482.9
173.9
591.1
336.9
84.8
252.1
0.0
63.4
0.0
1342.3
0.8
0.1
0.0
-------
Note
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
86
87
88
89
90
91
92
93
94
95
96
Symbol
RA
RI
N
N
No
N:
N2
F
FI
F2
G
I
I:
I2
I3
I«
I5
P2I
PA
P2I2
PA
PA
PA
B
E
E!
E2
E3
E4
P2E
PA
P2E2
PA
X
P2
Item
Renewable sources used
Renewable Electric Produced
Nonrenewable source flows
Internal fuels and minerals extracted
Dispersed Rural Source
Concentrated use, minerals, hydroelectric)
Exported without Use
Imported Fuels, Minerals+U, Electric
Fuels and minerals used (F+Nj-Renewable&Electric)
In region minerals used (Fj- F)
Imported Goods (materials)
Dollars Paid for all Imports
Dollars Paid for Service in Fuels, Minerals, Electric
Dollars Paid for Service in Goods
Dollars Paid for Services
Dollars Spent by Tourists
Federal Transfer Payments
Imported Services Total
Imported Services in Fuels, Minerals, Electric
Imported Services in Goods without fuels
Imported Services
Emergy Purchased by Tourists
Net Emergy Purchased by Federal $
Exported Products without minerals
Dollars Paid for All Exports
Dollars Paid for Goods without minerals
Dollars Paid for Minerals Exported
Dollars Paid for Exported Services
Federal Taxes Paid
Exported Services Total
Exported Services in Goods other than minerals
Exported Services in Minerals
Exported services
Gross Regional Product
Emergy $ ratio for the US in 2000 sej/$
1997
Emergy
(E+20 sej)
36.5
0.0
12.8
47.0
1.3
11.5
42.7
7.1
11.5
4.4
14.6
8.1
2.0
6.1
0.0
1.5
0.0
31.3
23.2
20.8
2.3
0.0
2.4E+12
Dollars
(E+6 $)
315.8
76.4
239.5
0.0
59.3
24.5
937.9
847.0
90.9
0.0
0.0
1354.1
2000 Em$
(E+6 Em$)
1555.0
0.0
544.2
2001.6
56.9
487.3
1815.7
301.5
487.3
185.8
620.1
343.8
83.1
260.6
0.0
64.5
0.0
1331.0
0.9
0.1
0.0
1998
Emergy
(E+20 sej)
32.1
0.0
14.3
48.6
2.5
11.8
44.1
7.3
11.8
4.5
14.8
8.0
1.8
6.2
0.0
1.5
0.0
30.8
23.4
21.0
2.3
0.0
2.4E+12
Dollars
(E+6 $)
323.0
72.9
250.2
0.0
61.8
24.0
3536.1
3420.2
116.0
0.0
0.0
1385.7
2000 Em$
(E+6 Em$)
1367.0
0.0
607.4
2069.6
104.7
502.6
1877.5
310.6
502.6
192.0
630.1
339.1
76.5
262.6
0.0
64.9
0.0
1309.4
0.9
0.1
0.0
-------
Note
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
86
87
88
89
90
91
92
93
94
95
96
Symbol
RA
RI
N
N'
No
N:
N2
F
FI
F2
G
I
Ii
I,
I3
I,
I,
P2I
PA
PA
PA
PA
PA
B
E
E,
E2
E3
E4
P2E
PA
P2E2
P2E3
X
P2
Item
Renewable sources used
Renewable Electric Produced
Nonrenewable source flows
Internal fuels and minerals extracted
Dispersed Rural Source
Concentrated use, minerals, hydroelectric)
Exported without Use
Imported Fuels, Minerals+U, Electric
Fuels and minerals used (F+Nj-Renewable&Electric)
In region minerals used (Fl- F)
Imported Goods (materials)
Dollars Paid for all Imports
Dollars Paid for Service in Fuels, Minerals, Electric
Dollars Paid for Service in Goods
Dollars Paid for Services
Dollars Spent by Tourists
Federal Transfer Payments
Imported Services Total
Imported Services in Fuels, Minerals, Electric
Imported Services in Goods without fuels
Imported Services
Emergy Purchased by Tourists
Net Emergy Purchased by Federal $
Exported Products without minerals
Dollars Paid for All Exports
Dollars Paid for Goods without minerals
Dollars Paid for Minerals Exported
Dollars Paid for Exported Services
Federal Taxes Paid
Exported Services Total
Exported Services in Goods other than minerals
Exported Services in Minerals
Exported services
Gross Regional Product
Emergy $ ratio for the US in 2000 sej/$
1999
Emergy
(E+20 sej)
32.2
0.0
11.8
34.6
1.3
10.4
31.5
7.3
10.4
3.1
18.2
8.2
1.8
6.5
0.0
1.5
0.0
27.3
20.2
18.7
1.6
0.0
2.4E+12
Dollars
(E+6 $)
357.5
77.6
279.9
0.0
63.8
24.2
962.1
867.1
94.9
0.0
0.0
1486.0
2000 Em$
(E+6 Em$)
1370.5
0.0
501.3
1472.0
57.0
444.3
1339.9
312.2
444.3
132.1
776.3
351.0
76.2
274.8
0.0
62.6
0.0
1160.3
0.8
0.1
0.0
2000
Emergy
(E+20 sej)
29.6
0.0
16.9
35.8
5.9
11.0
32.6
7.7
11.0
3.2
21.3
9.1
2.3
6.9
0.0
1.5
0.0
27.2
19.5
17.7
1.8
0.0
2.4E+12
Dollars
(E+6 $)
387.8
96.0
291.8
0.0
66.0
25.4
876.5
808.7
67.9
0.0
0.0
1590.9
2000 Em$
(E+6 Em$)
1261.6
0.0
718.5
1524.9
251.2
467.2
1386.7
328.9
467.2
138.3
905.3
387.5
95.9
291.6
0.0
66.0
0.0
1155.6
0.8
0.1
0.0
-------
Note
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
86
87
88
89
90
91
92
93
94
95
96
Symbol
RA
RI
N
N'
No
N!
N2
F
FI
F2
G
I
Ii
I2
I3
I4
I,
P2I
PA
PA
PA
PA
PA
B
E
E!
E2
E3
E4
P2E
PA
P2E2
P2E3
X
P2
Item
Renewable sources used
Renewable Electric Produced
Nonrenewable source flows
Internal fuels and minerals extracted
Dispersed Rural Source
Concentrated use, minerals, hydroelectric)
Exported without Use
Imported Fuels, Minerals+U, Electric
Fuels and minerals used (F+Nj-Ren.&Elec.)
In region minerals used (Fl- F)
Imported Goods (materials)
Dollars Paid for all Imports
Dollars Paid for Service in Fuels, Minerals, Electric
Dollars Paid for Service in Goods
Dollars Paid for Services
Dollars Spent by Tourists
Federal Transfer Payments
Imported Services Total
Imported Services in Fuels, Mineral, Electric
Imported Services in Goods without fuels
Imported Services
Emergy Purchased by Tourists
Net Emergy Purchased by Federal $
Exported Products without minerals
Dollars Paid for All Exports
Dollars Paid for Goods without minerals
Dollars Paid for Minerals Exported
Dollars Paid for Exported Services
Federal Taxes Paid
Exported Services Total
Exported Services in Goods other than minerals
Exported Services in Minerals
Exported services
Gross Regional Product
Emergy $ ratio for the US in 2000 sej/$
2001
Emergy
(E+20 sej)
30.3
0.0
13.0
41.3
1.2
11.8
37.7
8.1
11.8
3.7
21.3
8.8
2.3
6.5
0.0
1.5
0.0
19.9
12.1
10.1
1.9
0.0
2E+12
Dollars
(E+6 $)
402.2
106.3
295.9
0.0
69.7
26.6
654.0
566.1
88.0
0.0
0.0
1600.3
2000 Em$
(E+6 Em$)
1289.4
0.0
551.3
1758.6
50.5
500.7
1602.4
344.6
500.7
156.2
905.3
374.4
99.0
275.4
0.0
64.9
0.0
845.8
0.4
0.1
0.0
2002
Emergy
(E+20 sej)
22.5
0.0
19.9
37.4
9.1
10.8
34.7
8.0
10.8
2.8
22.0
8.2
1.9
6.2
0.0
1.4
0.0
17.7
14.9
13.4
1.5
0.0
2E+12
Dollars
(E+6 $)
394.3
93.4
300.9
0.0
67.0
27.1
718.9
648.5
70.4
0.0
0.0
1630.2
2000 Em$
(E+6 Em$)
956.9
0.0
845.5
1593.6
387.0
458.6
1476.5
341.5
458.6
117.1
937.0
346.9
82.2
264.7
0.0
58.9
0.0
753.6
0.6
0.1
0.0
-------
Note
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
86
87
88
89
90
91
92
93
94
95
96
Symbol
RA
RI
N
N'
No
N:
N2
F
FI
F2
G
I
Ii
I2
I3
I4
I,
P2I
PA
PA
PA
PA
PA
B
E
E!
E2
E3
E4
P2E
PA
P2E2
P2E3
X
P2
Item
Renewable sources used
Renewable Electric Produced
Nonrenewable source flows
Internal fuels and minerals extracted
Dispersed Rural Source
Concentrated use, minerals, hydroelectric)
Exported without Use
Imported Fuels, Minerals+U, Electric
Fuels and minerals used (F+Nj-Renewable&Electric)
In region minerals used (Fl- F)
Imported Goods (materials)
Dollars Paid for all Imports
Dollars Paid for Service in Fuels, Minerals, Electric
Dollars Paid for Service in Goods
Dollars Paid for Services
Dollars Spent by Tourists
Federal Transfer Payments
Imported Services Total
Imported Services in Fuels, Minerals, Electric
Imported Services in Goods without fuels
Imported Services
Emergy Purchased by Tourists
Net Emergy Purchased by Federal $
Exported Products without minerals
Dollars Paid for All Exports
Dollars Paid for Goods without minerals
Dollars Paid for Minerals Exported
Dollars Paid for Exported Services
Federal Taxes Paid
Exported Services Total
Exported Services in Goods other than minerals
Exported Services in Minerals
Exported services
Gross Regional Product
Emergy $ ratio for the US in 2000 sej/$
2003
Emergy
(E+20 sej)
28.5
0.0
14.3
23.7
4.0
10.3
21.4
8.0
10.3
2.3
15.3
7.4
2.1
5.3
0.0
1.4
0.0
18.5
18.7
17.8
1.0
0.0
2E+12
Dollars
(E+6 $)
371.5
105.9
265.6
0.0
71.3
29.8
933.3
885.9
47.4
0.0
0.0
1709.5
2000 Em$
(E+6 Em$)
1211.7
0.0
610.0
1007.6
171.4
438.6
911.3
342.2
438.6
96.4
650.3
316.9
90.3
226.5
0.0
60.8
0.0
786.2
0.8
0.0
0.0
2004
Emergy
(E+20 sej)
31.0
0.0
12.3
31.3
1.5
10.8
28.8
8.3
10.8
2.5
16.5
7.7
2.6
5.1
0.0
1.5
0.0
62.3
71.8
69.5
2.4
0.0
2E+12
Dollars
(E+6 $)
379.0
127.4
251.6
0.0
73.6
0.0
3536.1
3420.2
116.0
0.0
0.0
1849.8
2000 Em$
(E+6 Em$)
1318.7
0.0
523.7
1331.3
62.2
461.4
1225.0
355.2
461.4
106.3
704.2
327.7
110.2
217.5
0.0
63.6
0.0
2652.8
3.0
0.1
0.0
-------
Note
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
86
87
88
89
90
91
92
93
94
95
96
Symbol
RA
RI
N
N'
No
N:
N2
F
FI
F2
G
I
Ii
I2
I3
I«
I5
P2I
PA
PA
PA
PA
PA
B
E
E!
E2
E3
E4
P2E
PA
P2E2
P2E3
X
P2
Item
Renewable sources used
Renewable Electric Produced
Nonrenewable source flows
Internal fuels and minerals extracted
Dispersed Rural Source
Concentrated use, minerals, hydroelectric)
Exported without Use
Imported Fuels, Minerals+U, Electric
Fuels and minerals used (F+Nj-Renewable&Electric)
In region minerals used (Fj- F)
Imported Goods (materials)
Dollars Paid for all Imports
Dollars Paid for Service in Fuels, Minerals, Electric
Dollars Paid for Service in Goods
Dollars Paid for Services
Dollars Spent by Tourists
Federal Transfer Payments
Imported Services Total
Imported Services in Fuels, Minerals, Electric
Imported Services in Goods without fuels
Imported Services
Emergy Purchased by Tourists
Net Emergy Purchased by Federal $
Exported Products without minerals
Dollars Paid for All Exports
Dollars Paid for Goods without minerals
Dollars Paid for Minerals Exported
Dollars Paid for Exported Services
Federal Taxes Paid
Exported Services Total
Exported Services in Goods other than minerals
Exported Services in Minerals
Exported services
Gross Regional Product
Emergy $ ratio for the US in 2000 sej/$
2005
Emergy
(E+20 sej)
29.9
0.0
12.0
31.1
1.3
10.8
28.7
8.3
10.8
2.4
17.0
8.0
3.0
5.0
0.0
1.4
0.0
67.2
72.4
70.0
2.3
0.0
2E+12
Dollars
(E+6 $)
416.3
155.2
261.1
0.0
73.1
0.0
3788.4
3667.0
121.4
0.0
0.0
1975.7
2000 Em$
(E+6 Em$)
1272.5
0.0
511.6
1325.0
53.3
458.2
1221.0
354.3
458.2
103.9
721.8
338.4
126.1
212.2
0.0
59.4
0.0
2861.6
3.0
0.1
0.0
-------
Table 5.4 - Emergy indices and indicators for the San Luis Basin, 1995-2005. Our focus for this study was on Item 106.
Item
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
Name of Index
Renewable Use
In Region Non-renewable Use
Imported Emergy
Total Emergy Inflows
Total Emergy Used
Total Exported Emergy
Emergy used from Home Sources
Imports-Exports
Ratio of Exports to Imports
Fraction Used, Locally Renewable
Fraction of Use Purchased Outside
Fraction Used, Imported Services
Fraction of Use that is Free
Ratio of Purchased to Free
Environmental Loading Ratio
Investment Ratio
Use per Unit Area
Use per Person
Renewable Carrying Capacity
Developed Carrying Capacity
SLB Gross Regional Product
Ratio of SLB Emergy Use to GRP
Ratio of U.S. Emergy Use to GDP
Ratio of Electricity Use /Emergy Use
Fuel Use per Person
Population
Area
Renewable empower density
Emergy Yield Ratio
Emergy Index of Sustainability
Local Effect of Investment
Expression
RA
NO+N1
F+G+P2I
R+F+G+P2I
U=(RA+NO+F1+G+P2I)
B+P2E+N2
(NO+F2+R)/U
(F+G+P2I)- (B+P2E+N2)
(B+P1E+N2)/ (F+G+P2I)
R/U
(F+G+P2I)/U
P2I/U
(R+NO)/U
(F1+G+P2I)/(R+NO)
(F1+NO+G+P2I)/R
(F+G+P2I)/(R+NO+F2)
U/Area
U/Population
(R/U)* Population
8*(R/U)*Population
GRP1
U/GRP
U/GDP
El/U
F2/Population
Population
Area
RA/Area
(B+P1E+N2)/(F+G+P2I)
EYR/ELR
U/(F+G+P2I)
Unitsf
sejy-1
sejy-1
sejy-1
sej y- 1
sej y- 1
sej y- 1
sej y-1
sej/m2
sej/pers.
people
people
$/yr
sej/$
sej/$
J/sej
sej/pers.
people
m2
sej m-2
1995
(x E+20)*
32.43
12.59
28.37
60.80
66.46
91.25
0.57
-62.87
3.22
0.49
0.43
0.12
0.51
0.97
1.05
0.74
2.76E+11
1.52E+17
21366
170926
1.16E+09
5.72E+12
2.60E+12
0.037
l.OOE+16
43793
2.41E+10
1.34E+11
3.22
3.06
2.34
1996
(x E+20)*
28.84
16.68
29.07
57.91
67.33
93.21
0.57
-64.14
3.21
0.43
0.43
0.12
0.51
0.97
1.33
0.76
2.79E+11
1.51E+17
19088
152703
1.23E+09
5.48E+12
2.60E+12
0.038
1.02E+16
44566
1.20E+11
3.21
2.40
2.32
1997
(x E+20)*
36.54
12.79
29.73
66.28
71.98
97.12
0.59
-67.38
3.27
0.51
0.41
0.11
0.53
0.90
0.97
0.70
2.99E+11
1.59E+17
22992
183940
1.35E+09
5.32E+12
2.56E+12
0.037
9.73E+15
45289
1.52E+11
3.27
3.37
2.42
'Units in 1023 except as noted. f Where the "Units'
1 See Appendix 5-G for estimation of GRP.
column is blank, the indicators are dimensionless.
-------
Item
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
Name of Index
Renewable Use
In Region Non-renewable Use
Imported Emergy
Total Emergy Inflows
Total Emergy Used
Total Exported Emergy
Emergy used from Home Sources
Imports-Exports
Ratio of Exports to Imports
Fraction Used, Locally Renewable
Fraction of Use Purchased Outside
Fraction Used, Imported Services
Fraction of Use that is Free
Ratio of Purchased to Free
Environmental Loading Ratio
Investment Ratio
Use per Unit Area
Use per Person
Renewable Carrying Capacity
Developed Carrying Capacity
SLB Gross Regional Product
Ratio of SLB Emergy Use to GRP
Ratio of U.S. Emergy Use to GDP
Ratio of Electricity Use /Emergy Use
Fuel Use per Person
Population
Area
Renewable empower density
Emergy Yield Ratio
Emergy Index of Sustainability
Local Effect of Investment
Expression
RA
NO+N1
F+G+P2I
R+F+G+P2I
U=(RA+NO+F1+G+P2I)
B+P2E+N2
(NO+F2+R)/U
(F+G+P2I)- (B+P2E+N2)
(B+P1E+N2)/ (F+G+P2I)
R/U
(F+G+P2I)/U
P2I/U
(R+NO)/U
(F1+G+P2I)/(R+NO)
(F1+NO+G+P2I)/R
(F+G+P2I)/(R+NO+F2)
U/Area
U/Population
(R/U)*Population
8*(R/U)*Population
GRP1
U/GRP
U/GDP
El/U
F2/Population
Population
Area
RA/Area
(B+P1E+N2)/ (F+G+P2I)
EYR/ELR
U/(F+G+P2I)
Unitsf
sejy-1
sejy-1
sej y- 1
sej y- 1
sej y- 1
sejy-1
sejy-1
sej/m2
sej/pers.
people
people
$/yr
sej/$
sej/$
J/sej
sej/pers.
people
m2
sej m-2
1998
(x E+20)*
32.12
14.27
30.07
62.20
69.17
98.25
0.57
-68.18
3.27
0.46
0.43
0.12
0.50
1.00
1.15
0.77
2.87E+11
1.51E+17
21317
170536
1.39E+09
4.99E+12
2.47E+12
0.039
9.94E+15
45902
1.33E+11
3.27
2.83
2.30
1999
(x E+20)*
32.21
11.78
33.83
66.03
70.48
78.98
0.52
-45.16
2.33
0.46
0.48
0.12
0.48
1.10
1.19
0.92
2.92E+11
1.52E+17
21193
169544
1.49E+09
4.74E+12
2.31E+12
0.039
9.90E+15
46377
1.34E+11
2.33
1.96
2.08
2000
(x E+20)*
29.65
16.88
38.11
67.76
76.91
79.24
0.50
-41.12
2.08
0.39
0.50
0.12
0.46
1.16
1.59
0.98
3.19E+11
1.63E+17
18154
145234
1.59E+09
4.83E+12
2.35E+12
0.037
1.03E+16
47097
1.23E+11
2.08
1.30
2.02
'Units in 1023 except as noted. f Where the "Units'
1 See Appendix 5-G for estimation of GRP.
column is blank, the indicators are dimensionless.
-------
Item
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
Name of Index
Renewable Use
In Region Non-renewable Use
Imported Emergy
Total Emergy Inflows
Total Emergy Used
Total Exported Emergy
Emergy used from Home Sources
Imports-Exports
Ratio of Exports to Imports
Fraction Used, Locally Renewable
Fraction of Use Purchased Outside
Fraction Used, Imported Services
Fraction of Use that is Free
Ratio of Purchased to Free
Environmental Loading Ratio
Investment Ratio
Use per Unit Area
Use per Person
Renewable Carrying Capacity
Developed Carrying Capacity
SLB Gross Regional Product
Ratio of SLB Emergy Use to GRP
Ratio of U.S. Emergy Use to GDP
Ratio of Electricity Use /Emergy Use
Fuel Use per Person
Population
Area
Renewable empower density
Emergy Yield Ratio
Emergy Index of Sustainability
Local Effect of Investment
Expression
RA
NO+N1
F+G+P2I
R+F+G+P2I
U=(RA+NO+F 1 +G+P2I)
B+P2E+N2
(NO+F2+R)/U
(F+G+P2I)- (B+P2E+N2)
(B+P1E+N2)/(F+G+P2I)
R/U
(F+G+P2I)/U
P2I/U
(R+NO)/U
(F1+G+P2I)/(R+NO)
(F1+NO+G+P2I)/R
(F+G+P2I)/(R+NO+F2)
U/Area
U/Population
(R/U)* Population
8*(R/U)*Population
GRP1
U/GRP
U/GDP
El/U
F2/Population
Population
Area
RA/Area
(B+P1E+N2)/ (F+G+P2I)
EYR/ELR
U/(F+G+P2I)
Unitsf
sej y-1
sej y-1
sej y-1
sej y-1
sej y-1
sej y-1
sej y- 1
sej/m2
sej/pers.
people
people
$/yr
sej/$
sej/$
J/sej
sej/pers.
people
m2
sej m-2
2001
(x E+20)*
30.30
12.96
38.17
68.47
73.33
69.59
0.48
-31.42
1.82
0.41
0.52
0.12
0.43
1.33
1.42
1.09
3.04E+11
1.56E+17
19383
155064
1.60E+09
4.58E+12
2.19E+12
0.039
1.11E+16
46907
1.26E+11
1.82
1.28
1.92
2002
(x E+20)*
22.49
19.87
38.20
60.68
72.53
67.27
0.47
-29.07
1.76
0.31
0.53
0.11
0.44
1.30
2.23
1.11
3.01E+11
1.53E+17
14698
117581
1.63E+09
4.45E+12
2.07E+12
0.041
1.06E+16
47404
9.33E+10
1.76
0.79
1.90
2003
(x E+20)*
28.47
14.33
30.77
59.24
65.54
58.60
0.53
-27.83
1.90
0.43
0.47
0.11
0.50
1.02
1.30
0.89
2.72E+11
1.38E+17
20681
165445
1.71E+09
3.83E+12
2.00E+12
0.046
1.06E+16
47598
1.18E+11
1.90
1.46
2.13
'Units in 1023 except as noted. f Where the "Units" column is blank, the indicators are dimensionless.
1 See Appendix 5-G for estimation of GRP.
-------
Item
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
Name of Index
Renewable Use
In Region Non-renewable Use
Imported Emergy
Total Emergy Inflows
Total Emergy Used
Total Exported Emergy
Emergy used from Home Sources
Imports-Exports
Ratio of Exports to Imports
Fraction Used, Locally Renewable
Fraction of Use Purchased Outside
Fraction Used, Imported Services
Fraction of Use that is Free
Ratio of Purchased to Free
Environmental Loading Ratio
Investment Ratio
Use per Unit Area
Use per Person
Renewable Carrying Capacity
Developed Carrying Capacity
SLB Gross Regional Product
Ratio of SLB Emergy Use to GRP
Ratio of U.S. Emergy Use to GDP
Ratio of Electricity Use /Emergy Use
Fuel Use per Person
Population
Area
Renewable empower density
Emergy Yield Ratio
Emergy Index of Sustainability
Local Effect of Investment
Expression
RA
NO+N1
F+G+P2I
R+F+G+P2I
U=(RA+NO+F 1 +G+P2I)
B+P2E+N2
(NO+F2+R)/U
(F+G+P2I)- (B+P2E+N2)
(B+P1E+N2)/(F+G+P2I)
R/U
(F+G+P2I)/U
P2I/U
(R+NO)/U
(F1+G+P2I)/(R+NO)
(F1+NO+G+P2I)/R
(F+G+P2I)/(R+NO+F2)
U/Area
U/Population
(R/U)*Population
8*(R/U)*Population
GRP1
U/GRP
U/GDP
El/U
F2/Population
Population
Area
RA/Area
(B+P1E+N2)/ (F+G+P2I)
EYR/ELR
U/(F+G+P2I)
Unitsf
sej y-1
sej y-1
sej y-1
sej y-1
sej y-1
sej y-1
sej y-1
sej/m2
sej/pers.
people
people
$/yr
sej/$
sej/$
J/sej
sej/pers.
people
m2
sej m-2
2004
(x E+20)*
30.99
12.31
32.60
63.58
67.54
162.97
0.52
-130.37
5.00
0.46
0.48
0.11
0.48
1.08
1.18
0.93
2.80E+11
1.40E+17
22117
176938
1.85E+09
3.65E+12
2.03E+12
0.044
1.10E+16
48207
1.29E+11
5.00
4.24
2.07
2005
(x E+20)*
29.90
12.02
33.24
63.14
66.84
168.30
0.50
-135.06
5.06
0.45
0.50
0.12
0.47
1.15
1.24
0.99
2.77E+11
1.39E+17
21520
172157
1.98E+09
3.38E+12
1.91E+12
0.046
1.09E+16
48101
1.24E+11
5.06
4.10
2.01
'Units in 1023 except as noted. f Where the "Units" column is blank, the indicators are dimensionless.
1 See Appendix 5-G for estimation of GRP.
-------
To New Mexico
Figure 5.1 - An Energy Systems Diagram of the San Luis Basin showing the major features of this environmental system.
-------
,' GRP C~"
i 7590.9, "
\ X
Figure 5.2 - An aggregated model used to calculate summary variables and indices for the San Luis Basin system from
1995 to 2005. Year 2000 is presented here as an example.
• Renewable Energy Used
- Total Emergy Used
I
o
s
Q.
LU
90
80
70 -
60 -
50 -
40 -
30
20
10 -
0
1995
1997
1999
2001
2003
2005
Figure 5.3 - Total emergy used and renewable emergy used in the San Luis Basin from 1995 to 2005.
-------
-Energy Used per Person
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Figure 5.4 - The total emergy used per person in the San Luis Basin from 1995 to 2005.
-Total Emergy Inflows
-Total Emergy Outflows
i>,
'5T
I
o
Q.
E
LLJ
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Figure 5.5 - Emergy inflows from renewable sources, imported fuels, electricity, goods and services, emergy outflows
in exported materials, the services in material products, services, and nonrenewable resources exported directly in the
San Luis Basin, 1995-2005.
-------
0)
o
Q.
E
LU
9E+21
8E+21
7E+21
6E+21
5E+21
4E+21
3E+21
2E+21
1E+21
•Renewable Emergy
•Imported Emergy
Emergy Used
Non renewable Use from within the System
1995
1997
1999
2001
2003
2005
Figure 5.6 - A comparison of the total emergy used by the system with the renewable, local nonrenewable and
imported emergy that contribute to it in the San Luis Basin, 1995-2005.
•Fraction Used Locally Renewable
o
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1995
1997
1999
2001
2003
2005
Figure 5.7 - Fraction of renewable emergy to total emergy use in the San Luis Basin from 1995 to 2005.
-------
i
Q.
n
u
a
HI
1E+17
-Renewable Emergy Used per Person
1985
1990
1995
2000
2005
Figure 5.8 - The renewable emergy used per person in the SLB region from 1980 to 2005.
I
o
Q.
E
UJ
—•—Solar Energy Absorbed
—•—Wind Energy Absorbed
—*- Earth Cycle
Rain Chemical Potential
—•—Rain Geopotential as Runoff
—i—Snow Geopotential as Runoff
Total Geopotential Absorbed as Runoff
Evapotranspiration
1980
1985
1990
1995
2000
2005
Figure 5.9 - The renewable emergy inputs to the San Luis Basin, 1980-2005.
-------
•Rain Geopotential as Runoff
Evapotranspiration
-Snow Geopotential as Runoff
I
E 1.0E+21
UJ
5.0E+20
O.OE+00
1980
1985
1990
1995
2000
2005
Figure 5.10- The renewable emergy base for the San Luis Basin, 1980-2005.
o>
(A
Q.
UJ
-Hay
Wheat
-Fish
1.2E+21
1E+21 -
8E+20 -
<
6E+20 -
4E+20
2E+20 1
-Barley
- Potatoes
-Timber Harvest
Oats
•Livestock (Cattle)
t—i—i—i—i—i——i
1980 1985 1990 1995 2000 2005
Figure 5.11- Agricultural production in the San Luis Basin supported primarily by renewable resources, 1980-2005.
-------
-•-Electricity Used
Broken Stone, Rip Rap
-•-Non-metalic Minerals, nee.
—Fuels used
4.0E+21
Metallic Ores
Sand and Gravel
Soil Erosion (wind)
Groundwater use
O.OE+00
1980
1985
1990
1995
2000
2005
Figure 5.12-Major categories of nonrenewable resources used and/or produced in the seven counties of the San Luis
Basin, 1980-2005.
Imported Emergy
Exported Emergy
Q)
O
Q.
E
UJ
1.8E+22
O.OE+00
1995 1997 1999 2001 2003
Figure 5.13- The emergy exported from and imported to the San Luis Basin from 1995 to 2005.
2005
-------
-Total Fuels and Electricity -"-Materials Other Than Fuels
Services in Materials Other Than Fuels -•-Services in Fuels and Electricity
6E+21
5E+21
~ 4E+21
~ 3E+21
2E+21
1E+21 -
1980
1985
1990
1995
2000
2005
Figure 5.14- Major categories of emergy imported into the San Luis Basin, 1980-2005.
-•—Consumer Goods
Agricultural Goods
1E+21
9E+20
Industrial and Mining Goods
Construction Goods
1995
1997
1999
2001
2003
2005
Figure 5.15- The categories of material goods other than fuels imported to the SLB from 1995 to 2005.
-------
-•-Total Materials Exported
-*-Total Exports
-*- Livestock
—•— Forest Products Total
Services in Agricultural Products
-•-Services in Forest Products
Emigration
1.8E+22
-Services in Material Exports
Agricultural Crops
•Minerals Total
-All Other Materials
Services in Minerals
Services in All Other Materials
0)
jn,
<5
o
o.
E
1.6E+22
1.4E+22
1.2E+22
1E+22
8E+21
6E+21
4E+21
2E+21 •
1980
1985
1990
1995
2000
2005
Figure 5.16-Emergy exported from the San Luis Basin, 1980-2005: totals and aggregate categories Total exports
include total materials and total services exported. Total materials exported includes agricultural and forest products,
livestock minerals, and all other materials. Services in materials exported includes services in services in minerals, all
other materials, agricultural and forest products. Export of pure services was not evaluated and emigration is shown,
but was not included in the total.
- Potatoes
Livestock
-Gravel Or Sand
- Forest Products Total
Services in All Other Materials
7.0E+21
6.0E+21
•Grain , Wheat, Barley, Oats
Broken Stone Or Riprap
•Misc Nonmetallic Minerals, Nee
•All Other Materials
a>
o
a.
E
LU
1.0E+21
O.OE+00
1980
1985
1990
1995
2000
2005
Figure 5.17- Selected categories of agricultural, forest product, and mineral exports from the San Luis Basin, 1980-
2005. The "All Other Materials" and the services in that category are shown to facilitate further discussion.
-------
•MiscAgricultural Chemicals
Newspapers
•Meat, Fresh Or Chilled
•Misc Food Preparations, Nee
Misc Printed Matter
•Animal By-prod,in edible
3.5E+21
3.0E+21
_ 2.5E+21 •
>
'57 2.0E+21
•S2.
i_
I 1.5E+21
o
Q.
m 1.0E+21 -I
5.0E+20
O.OE+00
1995
1997
1999
2001
2003
2005
Figure 5.18- The six largest classes of exports contained in "All Other Materials" exported from the San Luis Basin,
1995-2005.
•Fuels and Electricity Used per Person
'a.
o
JO,
c
o
S2
0)
Q.
i_
0)
Q.
0)
1.75E+16
1.7E+16
1.65E+16
1.6E+16
1.55E+16-
1.5E+16
0)
E 1.45E+16
1980 1985 1990 1995 2000 2005
Figure 5.19- The emergy of fuels and electricity used per person in the San Luis Basin from 1980 to 2005.
-------
en
I
Q.
LU
•Soil Erosion (wind)
Groundwater use
9E+20
0 *
1980
1985
1990
1995
2000
2005
Figure 5.20 - Renewable emergy sources of the San Luis Basin (1980-2005) that are being used in a nonrenewable
manner.
O
'•5
•Fraction Locally Renewable
Fraction from Home Sources
1995
1997
1999
2001
2003
2005
Figure 5.21 - The fraction of total emergy use that comes from local renewable emergy and the fraction of use from
home sources in the San Luis Basin, 1995-2005.
-------
Appendix 5-A: for
Emergy
An emergy evaluation of the SLB was performed in
seven steps.
1. The spatial and temporal boundaries of the system
under analysis were specified. System boundaries
were chosen based on the research question to be
answered, considering the issue to be studied.
2. A diagram of a detailed Emergy Systems Model was
created using the Energy Systems Language (ESL;
Odum 1971). The diagram represented the main
features of the system components and flows that
exist within the specified boundaries. Performance of
the second step required prior study of the system and
data gathering from the literature and from meetings
with experts on the SLB.
3. Translated this knowledge of the system into an
aggregated ESL diagram addressing the specific
questions to be answered with the simplest possible
model formulation. Simplification of the detailed
model was done by aggregating components or
flows of similar function, thus information was not
discarded in this process.
4. Emergy analysis tables were set-up and descriptions
of the system's sources, components, and pathways
in the aggregated diagram were transferred to the
tables where the calculations needed to evaluate these
pathways quantitatively were compiled.
5. Gathered and assembled the raw data needed to
complete the emergy analysis tables along with the
conversion factors (energy contents, transformities,
specific emcrgics, etc.) needed to convert the raw data
into emergy.
6. After the raw data were converted into emergy,
summary variables were defined and calculated using
the aggregate diagram as a guide (Lu et al. 2007,
Odum 1996).
7. Calculated emergy indices representing various
aspects of system structure and function and
interpret the meaning of these indices relative to our
knowledge of their values in other systems. The
methods used to carry out an emergy analysis of West
Virginia and Minnesota are presented in detail in two
recent EPA publications, Campbell et al. (2005) and
Campbell and Glut (2009).
References
Campbell, D. E., Brandt-Williams. S., Meisch, M. 2005.
Environmental accounting using emergy: evaluation of
the state of West Virginia. EPA600/R-05/006. United
States Environmental Protection Agency, Narragansett,
Rhode Island.
Campbell, D.E., Ohrt, A. 2009. Environmental
accounting using emergy: evaluation of Minnesota.
USEPA Project Report, EPA/600/R-09/002, 138 pp.
Lu, H.F., Campbell, D.E., Chen, I, Qin, P., Ren, H.
2007. Conservation and economic vitality of nature
reserves: an emergy evaluation of the Yancheng
Biosphere Reserve. Biological Conservation 139,
415-438.
Odum, H.T. 1971. An energy circuit language for
ecological and social systems, in: Patten. B. (Ed.),
Systems Analysis and Simulation in Ecology, Vol. 2,
Academic Press, New York. pp. 139-211.
Odum, H.T. 1996. Environmental Accounting: Emergy
and Environmental Decision Making. John Wiley &
Sons. New York.
-------
Appendix 5-B:
for the of
the
We present revised definitions for two aspects of the
relationship between a socioeconomic system and its
associated larger systems. First, the Ernergy Yield Ratio
(EYR) for a local system, often characterized as U/F,
is redefined as Y/F (see original definition in Odum,
1996), where Y or yield is equivalent to exports of the
socioeconomic system. Second, the ratio formed by the
total emergy flow in the system (U) and the feedback
(F) from the larger system is defined as the Local
Effect of Investment (LEI) or the effect of feedback
from the larger system on the local system's empower.
This ratio is different from existing energy investment
ratios , which look at the matching between purchased
feedbacks and local renewable and nonrenewable
resources. Table 5-B.l and Fig. 5-B.l present some
emergy indicators and indices used in the analysis of the
sustainability of regional environmental systems.
The EYR for the SLB was defined to be the emergy
value of the exports divided by the feedback from the
larger system, which in this case is represented by
imports. This definition is more consistent with the
original conception of the EYR (Odum 1996). which
was used to characterize a production process, rather
than with the commonly used definition of the EYR
for a state nation, or region as the ratio of total system
emergy use (U) divided by the feedback from the larger
system. In this case, exports are the productive yield to
the larger system received from the local system, thus
the ratio of exports to imports is more appropriately
termed the EYR of a regional system. The ratio of U
to the feedback from the larger system is redefined as
the local effect of investment, LEI, which represents the
local changes in empower that result from investments
by the larger system in the local system. It captures the
local empower return on investment and as such it is a
measure of the change in well-being of the local system
as a function of emergy invested by the larger system.
The redefined indices more accurately characterize
both indices based on their function and lead to logical
inferences about the welfare of both local and larger
systems (Campbell 2008).
The Emergy Sustainability Index (ESI; Brown and
Ulgiati, 1997) of the SLB is determined according
to the variation of its two components, the EYR and
the Environmental Loading Ratio (ELR; Fig. 5-B.2).
Trends in these indices show that a slow increase in
environmental loading combined with a slow decline in
the EYR resulted in declining suslainabilily from 1997
until 2002. After 2002, the ELR falls rapidly as the EYR
increased rapidly resulting in a 5-fold increase in the ESI
from 0.79 in 2002 to 4.1 in 2005.
The balance of total emergy inflow and outflow is
a new index, which we calculated for this study. It
characterizes the macroscopic emergy balance by
including all inflows and outflows across Hie system
borders including the renewable emergy. It gives an
overall balance of inflows and outflows, which one
might expect to be approximately balanced for a system
in a dynamic steady state. Although this flow is not large
compared to the others evaluated in this study, the index
in its current form does not include the emergy of water
flowing out of the system as an outflow of renewable
emergy. The pattern of this index shows that outflows
exceed inflows in most years except in 2001 and 2003
when the two arc nearly in balance. This situation is in
contrast to the export-import balance for which outflows
always exceed inflows by a considerable margin.
Internal resources used are not included in the inflows
to the system, but can show up in outflow; thus, when
outflow markedly exceeds inflow, it may indicate that
capital storages within the system are being depleted.
The ESI characterizes sustainability as the ratio of the
EYR to the ELR, thus it has high values when the larger
system receives larger yields from its investment in the
local system, and when the intensity7 of nonrenewable
emergy use to attain those yields is less. The ESI was
above 2 in eight of the eleven years evaluated, which
indicates the SLB would be classified as a sustainable
system within the context established by Brown and
Ulgiati (1997) and subsequent studies. However, the fact
that this index falls below 2 from 2001 to 2003 raises
concern the system may become less sustainable in the
future. The ELR increased to a maximum in 2002 as the
EYR declined to a minimum in the same year resulting
in minimum sustainability in that year. The year 2002
was a drought year in which the system was supported
by a large, unsustainable use of groundwater, and after
this time the ESI increased rapidly as a result of a large
increase in EYR (i.e., during this period, for almost the
same inputs of emergy from the larger system, there
was a marked increase in the emergy of exports to the
larger system). This increase could only come either
from the system's internal resources or from additional
value added to imports by the improved application of
local labor. Over the course of the 11 years examined,
this index showed the system was moving toward
sustainability, but this trend was determined entirely
by the upsurge of exports in the final three years of the
study.
-------
A final index of the relative sustainability of the SLB
is the renewable carrying capacity of the system (Table
5.4). This index shows the system had the potential
to support 21520 people out of the current (2005)
population of 48101 at their present standard of living
using renewable emergy inflows alone. This assumes
that these inflows have not been further upgraded to
support socio-economic activities (e.g., by generating
electricity from solar or wind power). During the 11-
year period examined, there was a 4% decline in the
renewable carrying capacity of the SLB from 48.7% to
44.7% of the current population, but in 2002 only 31%
of the people in the SLB could have been supported by
the renewable emergy that was available in that year.
This index shows the SLB was moving away from
sustainability over the period examined and although
the decline was moderate, it could become worse, if the
natural precipitation regime is altered by regional climate
change in a similar manner to that predicted for the State
of New Mexico (2005). The percent renewable emergy
used is the most important factor determining this index
and thus the two indices give related results.
Another look at system well-being is shown by the
relationship between EYR, ESI, and LEI (Fig. 5-B.2).
EYR is in an overall declining trend from 1998 until
2002, and ESI and LEI are in a declining trend from
1997 until 2002. Between 2002 and 2005, the EYR
increased 2.87 times, the ESI increased 5.18 limes,
and the LEI increased 1.06 times. Although the EYR
and ESI recover strongly after 2002, the total emergy
use in the local system increases only slightly. The
ESI, EYR, and LEI reflect the relative advantages and
disadvantages that may exist in the relationship between
the SLB and its larger system trading partners (imports).
The pattern of the EYR and the ESI shows large benefits
accruing to the larger system in (lie economic recovery
that took place in the region from 2002 to 2005, but the
LEI indicates that this increase resulted in only a slight
increase (6%) in the efficacy of external investments in
the region in increasing well-being in the SLB.
The EYR shows the benefit to the larger system and
ranged from a low of 1.76 times the emergy invested
in 2002 to a high of 5.06 in 2005. From 1995 to 2005,
the LEI shows that about 2 times the invested emergy
flow was stimulated in the SLB by each unit of emergy
invested from outside. This multiplier remained positive
(1.90 to 2.42) over the 11-year time period and thus
investment from the larger system has improved local
well-being over the entire time.
An examination of other emergy indices of sustainability
indicated there was a movement away from
sustainability in the SLB over the period from 1995 to
2005. Renewable emergy use declined 4% over the
11-year study period and approximately half of the
present population could be supported at their current
standard of living on renewable emergy alone. The
ESI, which increased 34% from 1995 to 2005, provided
perspective on the relationship between the SLB and its
larger systems. From our results, it appears the SLB and
its larger systems go through cycles of improving and
declining sustainability and that pulsing apparently plays
a role in this variability.
References
Brown, M.T. Ulgiati, S. 1997. Emergy-based indices
and ratios to evaluate sustainability: monitoring
economics and technology toward environmentally
sound innovation. Ecological Engineering 9. 51-69.
Campbell, E.T. 2008. Environmental accounting of
natural capital and environmental sendees of the
National Forest system. Master's Thesis, Graduate
School of the University of Florida, Gainesville,
Florida, 227 pp.
Odum, H.T. 1996. Environmental Accounting: Emergy
and Environmental Decision Making. John Wiley &
Sons, New York.
State of New Mexico. 2005. Potential Effects of
Climate change on New Mexico. Agency Technical
Work Group, State of New Mexico. Available online
at http://www.imieiw.state.imx iis/aqb/cc/Potential_
EiTecls_Climate_Change_NM.pdf. Last accessed June
11,2010
-------
Table 5-B. 1 - Some emergy indicators and indices used in the analysis of the sustainability of regional environmental
systems. Symbols and expressions refer to Fig. 5-B. 1.
Name of Index
Total Emergy Flow
Renewable Emergy Inflow Absorbed
Percent renewable
Feedback or investment from the larger
system
Local Effect of Investment (LEI)
Yield
Emergy Yield Ratio (EYR)
Nonrenewable Emergy Inflow
Environmental Loading Ratio (ELR)
Emergy Sustainability Index (ESI)
Symbol or Expression
U
RA
RA/U
F = M + S
U/F
Y
Y/F
N,N0
(F+N+N0)/R
EYR/ELR
Definition
Empower (emergy/time) of the regional network. A measure of the
local system's well-being.
Renewable emergy inputs to the system without double counting.
A measure of the potential for sustainability of the present system.
Feedback is the sum of material, M, and service, S, feedbacks.
Effect of external investment on system empower.
For a regional system Y is equivalent to exports, i.e., yield to larger
systems.
Emergy return to the larger system on its investments. Y is equal to
exports in the case of a local system and F is equivalent to imports.
N, nonrenewable, i.e. fossil fuels, N0, renewable sources being used in
a nonrenewable manner, e.g., soil or groundwater used faster than their
replacement rate.
Potential effect of nonrenewable emergy use on the environment
In this index sustainability increases with greater returns to the larger
system and decreases with greater potential for environmental damage
to the local system.
= RA+ N + Nu
Local Renewable and
Nonrenewable In Hows
Figure 5-B. 1 - An Energy Systems Diagram used to define the emergy indices (Table 5-B. 1) to assess regional
sustainability.
-------
- Emergy Yield Ratio
Environmental Loading Ratio
Emergy Index of Sustainability
0)
I
x
0)
T3
C
1995
1997
1999
2003
2005
Figure 5-B.2 - The Emergy Sustainability Index (ESI = EYR/ELR) and its components the emergy yield ratio (EYR)
and the Environmental Loading Ratio (ELR) in the San Luis Basin from 1995 to 2005.
-------
Appendix 5-C: Calculating
per Unit
To calculate emergy units, data are converted first to
energy or mass units and then into emergy units by
multiplying the values by the appropriate coefficients
(the emergy per unit values). When only the
transformity (emergy per unit available energy) of
a quantity was available and the variable was given
in units of mass, it was first converted to energy by
multiplying by the energy per unit mass. Then it
was converted into emergy by multiplying by the
transformity. Quantities were aggregated by similar
function into composite variables considered most
important for defining key interactions. An example
of the calculation process follows: crop production is
reported as bushels of grain harvested. The bushels of
a particular crop are converted to mass by multiplying
by the average dry weight of a bushel of that type of
grain. The mass of grain is converted to energy by
multiplying by the energy content in joules per gram dry
weight (gdwt). Next, the energy content of the grain is
converted to emergy by multiplying by the transformity.
Finally, the emergy of all grain crops are summed to
obtain the total emergy of grain production.
Accounting units of mass and volume are most often
used to record the physical quantity of items used
by society. Other physical units may be used for
environmental data (e.g., velocity for wind, energy flux
for solar radiation, heat flux for geothermal energy,
etc.). In almost all cases, these original units can be
converted to energy or mass using readily available
formulae and/or average conversion factors (Campbell
and Ohrt 2009). The mass of a bushel of wheat and
the energy content of a gram of wheat are examples
of conversion factors that can easily be found in the
literature, the transformity of a joule of wheat, although
available in the ernergy literature, might be harder to
find for the average investigator. The energy content
of many items has been widely tabulated, but a similar
compendium of transformities and other eniergy per unit
factors does not exist at the present time. Therefore, in
any emergy evaluation, the investigator will invariably
have to calculate transformities for certain items. The
first choice is to determine the transformity in the
context of the production system of the item used in the
system under study. If this information is not available,
the emergy per unit values may be determined by
analyzing a fast and efficient production process (i.e.,
one operating at the optimum efficiency for maximum
power; see Odurn and Pinkerton 1955) for the item.
If the production process is unknown or not easily
documented, transfonnities may be determined using one
of the ten methods given in Odum (1996). The emergy
per unit factors used in this study are given in Tables
5-C.l to 5-C.3. This appendix contains the derivation
and assumptions used for new and revised emergy per
unit calculations that were needed to complete this study.
-------
Table 5-C. 1 - Emergy per Unit Factors (solar transformities and specific emergies) used in the San Luis Basin study.
The units for the emergy per unit factors are solar emjoules per joule (sej/j) for transformities unless otherwise noted.
Item
Emergy/ unit#
Adjustment
Ratiof
Adjusted
Transformity
Renewable Emergy Inflows
Solar energy reaching the surface of the earth by definition
Solar energy absorbed
Kinetic Energy of Wind (Odum 1996)
Earth Cycle Energy (Odum 1996)
Rain, Chemical Potential (Odum 1996, Campbell 2003a)
Chem. Potential Energy of Evapotranspiration (Campbell 2003a)
Rain, Geo-Potential on Land (Odum 1996 Campbell 2003a)
Snow, Geo-Potential on Land (Campbell and Ohrt 2009)
Rain, Geo-Potential Runoff (streams) (Odum 1996)
Snow, Geo-Potential Runoff (Campbell and Ohrt 2009)
Wave Energy (Lake Superior) (Campbell and Ohrt 2009)
Rivers, Chemical Potential (Odum 1996, Campbell 2003a)
Rivers, Geo-potential (Odum 1996)
Nitrogen Deposition ammonia (Odum 1996) sej/g
Adjusted from calculations made using the 15.83 baseline
NH3 (Campbell 2003b) sej/g
NOx (Campbell 2003b) sej/g
Ammonia (Campbell 2003b) sej/g
NO, N02, N03 (Campbell 2003b) sej/g
Sulfur Deposition S transformity (Campbell and Ohrt 2009) sej/g
Chloride Ion, Cl , Deposition (Campbell and Ohrt 2009) sej/g
1
1.21
1496
34377
18200
NA
10488
1.03E+05
27764
1.03E+05
30550
48459
27764
3.9E+09
2.4E+09
1.2E+10
NA
NA
NA
NA
1
1.21
0.9809
0.9809
NA
NA
0.9809
0.9809
0.9809
0.9809
0.9809
NA
0.9809
0.9809
NA
NA
NA
NA
1
1.21
1467
33720
18100
28100
10100
1.01E+05
27200
1.01E+05
30000
50100
27200
3.8E+09
1.4E+09
6.8E+09
1.58E+11
1.31E+10
Renewable Products
Agricultural Crops (without services, listed below)
Hay (Campbell et al. 2005); see also Brandt- Williams (2002) sej/g
Wheat (Campbell and Ohrt 2009) sej/g
Barley (average of Oats and Wheat) sej/g
Oats (Campbell et al. 2005; see also Brandt- Williams 2002) sej/g
Potatoes, (Campbell et al. 2005; see also Brandt- Williams 2002) sej/g
Livestock (Odum et al. 1998)
Beef (Brandt- Williams 2002)
Fish Production (Odum et al. 1998)
Hydroelectricity (Odum 1996)
Net Timber Growth (Tilley 1999)
Timber Harvest with Service (Tilley 1999)
Ground Water (Odum et al. 1998)
NA
NA
NA
NA
NA
2.00E+06
5.64E+05
2.00E+05
1.23E+05
2.10E+04
7.00E+04
1.62E+05
NA
NA
NA
NA
NA
0.9809
0.9809
0.9809
0.9809
0.9809
0.9809
0.9809
7.39E+08
2.88E+09
3.46E+09
4.05E+09
2.32E+08
1961800
553228
1961800
120258
20599
68663
159068
# - Emergy per unit values calculated on the 15.83 E24 sej/y baselines can be converted to the 9.26 line by multiplying by 0.585. | - The adjustment
ratio converts values calculated using the 9.44 E24 sej/y planetary baseline to their values on the 9.26 E24 sej/y baseline used in this study and
recommended by Campbell et al. (2005).
-------
Item
Solid Waste Production (Brown & Buranakarn 2000) sej/g
Cement with fly ash
Concrete recycled aggregate
Recycled steel
Recycled aluminum
Recycled lumber
Lumber from recycled plastic
Tile from glass
Average for Solid Waste
Emergy/ unit*
6.40E+09
1.40E+10
4.82E+09
3.09E+09
1.20E+10
3.22E+09
5.58E+09
2.16E+09
6.40E+09
Adjustment
Ratiot
0.9809
Adjusted
Transfonnity
6.28E+09
Nonrenewable Inflows
Energy
Coal (Odum 1996, Campbell and Ohrt 2009)
Geologic processes. 9.44E24 sej/y baseline
Relative efficiency in electricity generation
Average
Rounded value commonly used.
Natural Gas (Odum 1996, Bastianoni et al. 2005)
Geologic processes. 9.26 baseline
Relative efficiency. 9.44 baseline
Average on the 9.26 baseline
Crude Oil (Odum 1996, Bastianoni et al. 2005)
Geologic processes. 9.26 baseline
Relative efficiency. 9. 44 baseline
Average on the 9.26 baseline
Electricity from coal (Campbell et al. 2005, Odum 1996)
4.00E+04
3.40E+04
4.30E+04
3.85E+04
4.00E+04
4.80E+04
4.00E+04
4.80E+04
4.35E+04
5.40E+04
5.54E+04
5.40E+04
5.42E+04
1.74E+05
NA
NA
NA
0.9809
3.78E+04
4.35E+04
5.42E+04
1.70E+05
Minerals
Iron Ore (Odum 1996)
Sand and Gravel (Campbell et al. 2005) sej/g
Limestone (Odum 1996) sej/g
Dolomite (Campbell and Ohrt 2009)
Dolomite (Campbell and Ohrt 2009) sej/g
Peat (Odum 1996) sej/g
Granite (Odum 1996) sej/g
Clay (Odum 1996) sej/g
Sandstone (Odum 1996) sej/g
Alumina/Bauxite (Odum 1996)
Soil Erosion, top soil, (Odum 1996)
Tourism (dollars) (Campbell and Lu 2009)
6.20E+07
NA
l.OOE+09
NA
NA
3.60E+08
5.00E+08
2.00E+09
l.OOE+09
1.50E+07
74000
NA
0.9809
NA
0.9809
NA
NA
0.9809
0.9809
0.9809
0.9809
0.9809
0.9809
NA
6.08E+07
1.31E+09
9.81E+08
1.98E+07
1.08E+10
3.53E+08
4.90E+08
1.96E+09
9.81E+08
14713500
72600
2.60E+12
-------
Item
Tourism (virtual) (calculated this study)
Uranium (Cohen et al. 2007) sej/g
Electricity from Uranium (Campbell and Ohrt 2009)
Imported Goods
Petroleum (refined) (Odum 1996, Bastiauoni et al. 2009)
Fuels (services for 2000) (Campbell and I,u 2009)
Electricity (sen-ices 2000) (Campbell and Lu 2009)
Services (2000) (Campbell and Lu 2009)
Federal Government (spent in US) (Campbell and Lu 2009)
Goods w/o iron ore and fuels (materials)
Iron ore as taconite (Campbell and Ohrt 2009) sej/g
Steel (Brown and Buranakarn 2000) sej/g
Wood and Wood Products
Lumber and Wood Products, (Tilley 1999)
Furniture and Fixtures
Paper Products, (Campbell and Cai 2007)
N fertilizer, (Odum 1996) sej/g
P fertilizer, (Odum 1996) sej/g
Potash, (Odum 1996) sej/g
Storages
Forests (Tilley 1999)
Water, lakes (Campbell and Ohrt 2009).
Soil (Top soil) (Odum 1996)
Iron ore (Campbell and Ohrt 2009)
Iron (specific emergy) sej/g
Peat (Odum 1996)
Sand & Gravel (Campbell et al. 2005) sej/g
Limestone (Odum 1996)
Dolomite (Campbell and Ohrt 2009) sej/g
Nickel (Campbell and Ohrt 2009) sej/g
Copper (Campbell and Ohrt 2009) scj/g
Platinum (Campbell and Ohrt 2009) sej'g
People (Odum 1996) sej /individual
Preschool sej/individual
School sej/individual
College Grad sej/individual
Post-College sej/individual
Elderly (70+) (Campbell et al. 2005) sej/individual
Emergy/ unit"
NA
1.6 El 11
NA
6.60E+04
NA
NA
NA
NA
NA
NA
3.45E I 09
NA
NA
NA
NA
3.80E'09
3.90E-r09
l.lOE-09
NA
NA
74000
6.20E-07
1.20E-10
1.90E-04
NA
1.00E-r09
NA
2.00E+11
9.80E-10
3.70E-11
NA
3.40E-16
9.40E+16
2.80E-17
1.31E+18
NA
Adjustment
Ratio*
NA
0.585
NA
NA
NA
NA
NA
NA
NA
NA
0.9809
NA
NA
NA
NA
0.9809
0.9809
0.9809
NA
NA
0.9809
0.9809
0.9809
0.9809
NA
0.9809
NA
0.9809
0.9809
0.9809
NA
0.9809
0.9809
0.9809
0.9809
NA
Adjusted
Transformity
4.60E-I-12
9.36E I 10
4.81E-I-04
6.58E+04
2.35E-I-12
2.35E I 12
2.35E-H2
2.35E+12
Variable
3.61E+09
3.38E I 09
7.90E+04
2.22E+05
3.73E I 09
3.83E-I-09
1.08E-I-09
28234
1.81E+04
72600
6.08E-I-07
3.51E+09
1.86E-KJ4
1.31E+09
9.81E-r08
1.08E-10
2.55E+10
1.14E-11
1.13E-11
NA
3.34E-16
9.22E+16
2.75E-17
1.28E+18
1.69E-17
-------
Item
Public Status sej/individual
Legacy sej/individual
Emergy/ unit*
3.93E+18
7.85E+18
Adjustment
Ratiot
0.9809
0.9809
Adjusted
Transformity
3.85E+18
7.70E+18
Emergy to Money Ratio for the United States (Campbell and Lu 2009) sej/$
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
NA
5.5421E+12
4.9734E+12
4.3017E+12
3.7936E+12
3.7571E+12
3.5117E+12
3.1773E+12
3.1736E+12
3.1685E+12
2.9942E+12
2.8753E+12
2.6836E+12
2.5672E+12
-------
Table 5-C.2 - Specific emergies for imports (sej/g). All commodity specific emergies are without sen-ices and relative
to the 9.26E24 sej/y baseline. Values of the emergy per unit factors for the commodity classes in this table were
determined from the data given in Table 5-C. 1 and other data given in the calculations shown in Table 5-C.3. The note
numbers connect to Table 5-C.3. Wwt = wet weight; otherwise numbers are dwt = dry weight.
Note
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
New Commodity Classes
Fresh and Leafy Vegetables
Fresh Fish or Whale Products
Manganese Ores
Bituminous Coal
Natural Gasoline
Dimension Stone, quarry
Broken Stone or Riprap
Gravel, Sand
Clay and ceramics
Chemical or Fertilizer Mineral Crude
Meat, Fresh or Chilled
Meat, Fresh Fro/en
Meat Products
Animal By-product, inedible
Poultry, fresh, frozen and processed and eggs
Milk & Milk Products
Canned or Cured Sea Foods
Fruit and Vegetables processed
Frozen Specialties
Flour or Other Grain Mill Products
Prepared or Canned Feed
Flour and meal
Dog, cat or Other Pet Food
Sugar and candy
Malt Liquors
Malt
Wine, brandy" or Brandy Spirit
Distilled or Blended Liquors
Soft Drinks or Mineral Water
Misc Flavoring Extracts
Soybean Oil or By-products
Marine Fats or Oils
Macaroni, spaghetti, etc.
Miscellaneous Food Preparations
Men's or Boy's Clothing
Canvas Products
Specific Emergy w/o sendees
3.58E-08
1.58E+10
3.50E+11
1.11E+09
2.92E-09
8.17E-08
4.90E+08
1.3 IE ' 09
1.96E-09
2.88E-09
1.83E-10
2.1 IF, -1-10
2.28E-10
9.87E-10
9.95E-08
3.19E-09
1.62E+10
3.19E'09
1.22E-MO
3.36E-09
5.07E-09
6.19E+09
5.24E'09
1.33E-09
1.65E-08
3.46E-09
1.18E-09
5.68E-KJ8
1.20E-08
7.10E-08
3.11E+10
1.37E+11
5.22E-09
7.84E-09
1.41E-10
1.41E+10
Units
sej/g wwt
sej/g wwt
scj/g
sej'g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g wwt
sej/g wwt
sej/g wwt
sej 'g wwt
sej/g wwt
sej/g wwt
sej/g
sej/g wwt
sej/g wwt
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
scj/g
sej/g
sej/g
sej/g
sej/g
scj/g
sej/g
sej/g
sej'g
sej/g
sej/g
-------
Note
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
New Commodity Classes
Primary Forest Materials
Lumber I Sawmill
Millwork or Cabinetwork
Plywood or Veneer
Prefabricated & Structural Wood Products.
Treated Wood Products
Miscellaneous Wood Products
Mattresses Furniture
Pulp or Pulp Mill Products
Paper and Paper Products
Paper or Building Board
Newspaper and Printed matter
Potassium or Sodium Compound
Industrial Gases
Misc Industrial Inorganic Chemicals
Misc Indus Organic Chemicals
Plastic Matter or Synthetic Fibers
Fertilizers
Miscellaneous Agricultural Chemicals
Adhesives
Explosives
Chemical Preparations
Petroleum Refining Products
Liquefied Gases, Coal or Petroleum
Asphalt
Tires, Rubber Products
Plastic
Portland Cement
Brick, ceramics
Concrete Products
Ready mix concrete, wet
Lime or Gypsum Products
Non-metallic Minerals Processed
Blast Furnace Or Coke
Primary Iron or Steel Products
Electrometallurgical Products
Iron or Steel Shapes
Primary Aluminum Smelter Products
Copper or Alloy Basic Shapes
Specific Emergy w/o services
3.77E+08
5.18EI08
7.78E-I-08
2.11E-09
1.97E+09
2.02E+09
1.72EI09
2.89E-I-09
7.28E-I-08
4.26E+09
2.15E+09
4.26E+09
7.45E-I-09
9.63E-HO
2.75E+09
5.94E+09
3.09E+09
2.88E-I-09
1.45E-HO
3.73E-I-08
3.73E+09
7.45E+09
2.64E-I-09
3.11E-09
2.64E+09
4.22E+09
4.13E+09
1.93E+09
2.15E+09
1.32E-I-09
1.62EI09
1.61E+09
3.69E+09
2.12E+09
3.31E-I-09
6.64E+10
3.31E+09
1.27E+10
5.03E-HO
Units
scj/g
sej/g
sej/g
sej/g
sej/g
scj/g
sej/g
sej/g
sej/g
sej/g
scj/g
sej'g
sej/g
sej/g
sej/g
scj/g
sej'g
sej/g
sej/g
sej/g
scj/g
sej'g
sej/g
sej/g
sej/g
scj/g
sej'g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g
sej/g
scj/g
sej'g
sej/g
sej/g
-------
Note
76
77
78
79
80
81
82
83
84
85
86
87
New Commodity Classes
Aluminum Or Alloy Basic Shapes
Miscellaneous Nonferrous Basic Shapes, Wire
Metal products
Industrial Machinery and motors
Carbon Products For Electric Uses
Household Machinery & Components
Electronics, precision instruments
Toys, Games, Sports, Household
Matches
Rubber or Plastic Scrap
Warehouse & Distribution Center
Rail International Drayage
Specific Emergy w/o services
1.27E+10
3.15E'10
4.02E-09
7.76E-09
1.45E-H1
4.02E+09
1.91E'10
5.43E-09
8.68E-09
4.17E-09
NA
NA
Units
sej/g
sej/g
sej'g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g
-------
Table 5-C.3 - Determination of the specific emergies for imported commodity classes in Table 5-C.2. All commodity
specific emergies are without sendees and relative to the 9.26E24 sej/y baseline.
Notes
1
2
3
4
5
6
7
8
9
10
Commodity Class, Calculations
Average for Vegetables given below
Average of 6 vegetables.
Fresh Fish Or Whale Products
Average of 3 species
Manganese Ores
Bituminous Coal
Natural Gasoline
Dimension Stone
Limestone
Granite
Sandstone
Average of 3 stones
Clav
Fertilizer
N fertilizer
P fertilizer
Potash
Average
Source, Assumptions
(Campbell and Ohrt 2009, Brandt-
Williams 2002)
Cabbages
Tomatoes
Cucumbers
Green beans
Bell Peppers
Sweet corn
Tilapia, (Brown etal. 1992)
Salmon, (Odum 2000)
Catfish, (Odum etal. 1998)
Avg. heat content fish
Tilapia
Salmon
Catfish
Average specific emergy of 3 fish species
(Cohen et al. 2007)
(Campbell and Ohrt 2009)
Coal heat content
Specific Emergy
(Bastianoni et al. 2009)
(Odum 1996)
(Odum 1996)
(Campbell et al. 2005)
Average
(Odum 1996)
(Odum 1996)
(Odum 1996)
(Odum 1996)
Values
3.58E+08
2.29E-KJ8
3.25E-i-08
2.59E^07
9.68E-i-08
2.85E-H)8
3.13E+08
5310
6410
5650
5790
5.53E-05
5.67E-06
1 .96E-i-06
2.73E 1 06
1.58E+10
3.78E+04
29400
1.11E+09
2.92E+09
8.17E^08
9.81E+08
4.90E-08
9.81E-H38
8.17E+08
1.96Ei9
2.87E+09
3.73E-09
3.83E-H39
1 .08E4-09
2.87Ei09
Units
sej/g
sej/g
sej/g
sej/g
sej/g
sej/'g
sej/g
J/cr
J/g
J''g
J/g
sej/J
sej/'J
sej/J
sej/J
sej'g
sej/J
J/g
sej/g
sej/g
scj/g
sej'g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g
sej/g
sej/g
-------
Notes
11
12
13
14
15
16
17
Commodity Class, Calculations
Meat, Fresh or Chilled
Assume, average of beef and pork plus energy
used in slaughter and cutting.
Slaughter and cutting
Estimate for Meat, Fresh or Chilled
Meat Processing
slaughter, cutting
grinding, freezing
storage
frying or cooking
Total
Estimate Energy required for meat processing
Meat, Fresh Frozen
Add grinding, freezing and storage to 11.
Meat Products
add cooking and frying to above
Animal By-product, inedible
Pig carcass is 1 8% bone and skin
Poultry', fresh, frozen and processed and eggs
Milk & Milk Products
(Carlsson-Kanayama & Faist 2000)
Canned Or Cured Sea Foods
fish plus pickling
Source, Assumptions
Beef, (Brandt- Williams 2002)
Pork
Hogs, (Cavalctt ct al. 2006)
Hogs
Average
(Carlsson-Kanayama and Faist 2000)
http : //www. in fra. kth . se/fms/pdf/
energyuse.pdf
Energy Use for Meat Processing
MJ/90 g (megajoules per 90 g)
0.815
0.14
1 .375
0.895
3.225
Assume electricity is used for processing
energy.
Assume Beef carcass is 19% bone
Specific emergy
(Odumetal. 1998)
Milking, cheese making
Storage
Total
Processing
Cheese
Milk
Average Milk and Cheese
Pickling
Values
1.33E-I-TO
2.01E-I-TO
1.28E+06
15730
1.67E+10
1.54E+09
1.83E-I-TO
9056
1556
15278
9944
35833
2.87E-I-09
1.69E+09
2.28E-HO
0.185
9.87E-HO
9.95E-I-08
0.24
0.04
0.28
3.18E-I-09
4.78E-I-09
1.60E+09
3. 19E-I-09
2297
3.91 E+08
Units
sej/g
sej/g
sej/'J
J''g
sej/g
sej/g
sej/g
J.'g
J/g
J/g
J/g
J''g
sej/g
sej/g
sej/g
sej/g
sej/g wwt
MJ/15 g
MJ/1 5 g
MX/15 g
(sej/g wwt)
(sej/g wwt)
(sej/g wwt)
(sej/g wwt)
J/g
sej/g
-------
Notes
18
19
20
21
22
23
24
25
Commodity Class, Calculations
Fruit and Vegetables processed
Oranges -f freeze drying and pickling
(Carlsson-Kanayama and Faist 2000)
Frozen Specialties
Average of Frozen Meat and Fruit and
Vegetables Processed
Flour or Other Grain Mill Products
Wheat + energy of milling
(Carlsson-Kanayama and Faist 2000)
Prepared or Canned Feed
Flour and meal
Average com and wheat plus milling
Dog. cat or other pet food
Sugar and candy
Malt Liquors
Use fraction water, barley, and alcohol in beer
httpi'/bioeuergy.oml.gov/papers/misc/energy
conv.html
Source, Assumptions
Oranges
Freeze drying
Pickling
Average
Emergy
Wheat
Milling
Corn, (Brandt- Williams 2002)
Soybean, (Brandt- Williams 2002)
Sorghum, (Brandt- Williams 2002)
Oats, (Brandt- Williams 2002)
Average
Wheat
Corn
Milling
Corn, (Brandt- Williams 2002)
Soybean, (Brandt- Williams 2002)
Chicken
(Odum et al. 1986)
Transform! ty
Specific Emergy
Rain water sej/g (Campbell 2003a)
Beer composition (content)
Alcohol content
Barley content
Ethanol heat content
Transformity ethanol (Odum 1996)
Specific emergy ethanol
Emergy
Valyes
1.37E4-08
3.35E4-04
2.30E+03
1.79E-H)4
3.19Ei09
1.22E+10
2.88E+09
2.84E-03
3.36E+09
8.53E-KJ9
6.19E-09
1 .49E4-09
4.05E-H)9
5.07E-H)9
2.88E+09
8.53E+09
2838
6.19E-09
8.53E-KJ9
6.19E-09
9.95E-i-08
5.24E^09
16190
8.24E 1 04
1.33E+09
85740
0.9196
0.06
0.02
2.67E-04
58860
1.57E+09
1.65E^08
Units
sej/g
J/g
J'g
J/g
(sej/g wwt)
sej/g wwt
sej/g
J/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
J/g
sej/g
sej/g
sej/g
sej/g
scj/g
J'g
sej/J
sej'g
sej/g
J/g
sej/J
sej/g
-------
Notes
26
27
28
29
30
31
32
Commodity Class, Calculations
Malt
Wine, brandy or brandy Spirit
Alternate estimate for wine
Distilled or Blended Liquors
Soft Drinks or Mineral Water
Fraction times specific emergy of ingredients
Miscellaneous Flavoring Extracts
Soybean Oil or By-products
http://www.ces.purdue.cdu/extincdia/GQ/GQ-
39.html
Marine Fats or Oils
Source, Assumptions
Use barley
(Bastianouietal. 2001)
Wine
Grapes
Wine
Water content
Alcohol content
Grape essence
factor (weight of grapes per weight of
wine)
Water (fraction x specific emergy water)
Alcohol (fraction x specific emergy
alcohol)
Grapes (fraction x factor* specific emergy
grapes)
Total Wine
86 proof whisky (water content)
Alcohol by weight
Emergy
water
sugar
Emergy
Vanilla extract
Water
Alcohol by weight
Sugar
Emergy
Soybeans (Brandt- Williams 2002)
Fat content
specific emergy soybean oil
Oily fish avg. of herring, mackerel.
Chinook salmon
Fraction oil
Fish oil (energy content)
specific emergy fish oil
Values
3.41E+05
1.18E+09
9.72E+08
0.8658
0.104
0.0302
1.68
7.42E+04
1.64E+08
1.64E-I-09
1.80E+09
0.639
0.361
5.68E+08
0.9031
0.0897
1.20E-I-08
0.5258
0.344
0.1265
7.10E+08
6.19E-K19
0.1994
3.1 IE I 10
0.1155
7710
1.37E+11
Units
sej/J
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/J
sej/g
-------
Notes
33
34
35
36
37
38
39
40
41
42
43
44
45
Commodity Class, Calculations
Macaroni, spaghetti, etc.
Assume noodles are similar to bread.
Wheat -f emergy to process
(Carlsson-Kanayama and Faist 2000)
Miscellaneous Food Preparations
Corn, soy. wheat, oats + emergy to process
Men's or Boy's Clothing
Canvas Products
Primary Forest Materials
Lumber + Sawmill
Millwork or Cabinetwork
Assume 50% increase over lumber due to
energy and materials used in processing
Plywood or Veneer
Prefabricated & Structural Wood Products
Assume 50% increase due to energy and
materials used in processing plywood and
lumber
Treated Wood Products
Lumber plus 0. 1 paint
Miscellaneous Wood Products
Mattresses Furniture
Pulp or Pulp Mill Products
Source, Assumptions
Bread making
Milling
Baking
Storage
Milling
Baking
Storage
Total
Specific Emergy for processing
Average
Processing emergy as fraction of total
Specific Emergy
Cotton (Odum 1996)
Cotton (Odum 1996)
Logs avg. soft and hardwood (Tilley
1999)
Energy per gram dry weight.
Specific Emergy
Lumber (Tilley 1999)
Specific emergy
Specific emergy
Plywood (Tilley 1999)
Specific emergy
Average lumber and plywood
Specific emergy of Prefabricated wood
products
Paint (Biiranakam 1998)
Specific Emergy
Average value of 4 preceding wood
products
(Buranakarn 1998)
Wood pulp (Campbell and Cai 2007,
Tilley 1999)
Specific Emergy
Valyes
0.21
0.725
0.08
2838
9797
1081
13716
2.34E+09
5.22E-KJ9
5.41E-09
144.81%
7.84E^09
1.41EUO
1.41E+10
19620
19200
3.77E+08
27000
5.18E-H38
7.78E-08
110000
2.11E+09
1.32Ei09
1.97Ei09
1.50E+10
2.02E+09
1.72E+09
2.89E-09
37900
7.28E-08
Units
mJ/74 g
in.T/74 g
mJ/74 g
J/g
J'g
J'g
J/g
sej/g
sej/g
sej/g
scj/g
sej/g
sej'g
sej/J
J/g
sej/g
sej/J
sej/g
sej/g
sej/J
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g
sej/g
sej/J
sej/g
-------
Notes
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
Commodity Class, Calculations
Paper and Paper Products
Paper or Building Board
Newspaper and Printed matter
Potassium Or Sodium Compound
Industrial Gases
Miscellaneous Industrial Inorganic Chemicals
Miscellaneous Industrial Organic Chemicals
Average pesticides, glue, petroleum products.
Plastic Matter or Synthetic Fibers
Fertilizers
Miscellaneous Agricultural Chemicals
Adhesives
Explosives
Chemical Preparations
Petroleum Refining Products
Liquefied Gases, Coal or Petroleum
Asphalt
Tires, Rubber Products
Plastic
Portland Cement
Brick ceramics
Concrete Products
Ready mix concrete, Wet
Lime or Gypsum Products
Source, Assumptions
Paper (Campbell and Cai 2007, Tilley
1999)
Specific Emergy
Paper board (Campbell and Cai 2007,
Tilley 1999)
Assume paper
Caustic soda (Pritchard 2000)
Nitrogen gas (Campbell 2003b)
Average of hydrated lime, caustic soda,
diatomite, sulfuric acid (Oduni et al.
2000)
pesticides
glue
Specific emergy
plastics (Buranakarn 1998)
Average N, P, K (Oduni 1996)
Assume pesticides (Brown and Arding
1991)
Assume glue (Buranakarn 1998)
Assume similar to ammonium fertilizer
Assume average of glue and pesticides
(Bastianoni et al. 2009)
LPG (Bastianoni et al. 2009)
(Bastianoni et al. 2009)
Rubber (Buranakarn 1998)
(Buranakarn 1998)
Polyethylene
Plastics
Average
(Buranakarn 1998)
(Buranakarn 1998)
(Ilaukoos 1995)
Average sand and gravel and cement
(Pritchard 2000)
Values
222000
4.26E+09
112000
2.15F-I-09
4.26E+09
7.45E I 09
9.63E+10
2.75E+09
1.45E-I-10
3.73E+08
5.94E+09
3.09E--09
2.88E-KJ9
1.45E4-10
3.73F,-i-08
3.73E-K19
7.45E 1 09
2.64E+09
3.11E-I-09
2.64E+09
4.22E-09
5.17E-09
3.09E+09
4.13E+09
1.93E-H)9
2.15Ei09
1.32E+09
1.62E+09
1.61E+09
Units
sej/J
sej/g
sej/J
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g
-------
Notes
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
Commodity Class, Calculations
Non-metallic Minerals Processed
Average of glass and ceramics
Blast Furnace or Coke
Primary Iron or Steel Products
Average iron and steel
Electrometallurgical Products
Iron or Steel Shapes
Primary Aluminum Smelter Products
Copper or Alloy Basic Shapes
Aluminum or Alloy Basic Shapes
Miscellaneous Nonferrous Basic Shapes, Wire
Metal products
Industrial Machinery and motors
Carbon Products For Electric Uses
use graphite @ 10% of world coal
Household Machinery & Components
Electronics, precision instruments
Toys, Games, Sports, Household
Matches
Rubber or Plastic Scrap
Source, Assumptions
(Burauakarn 1998, Haukoos 1995)
Glass
Flat glass
Float glass
Average
Ceramics
Average Specific Emergy
(Bastianoni et al. 2009)
(Buranakarn 1998, Haukoos 1995)
Pig Iron
Steel Conventional
Average
(Odumetal. 1987)
Use iron and steel
(Odum 2002) Al sheet and recycle
(Buranakarn 1998)
(Odum 1996)
Average copper and aluminum
Assume steel
(Odum 1996)
Specific Emergy of Graphite
World Coal
World Graphite
Graphite/coal ratio
Heat Content Graphite
Transformity of Graphite
Specific Emergy Graphite
Assume steel
Use average Steel, plastics, copper
Use average plastic, wood, paper, Al
Use average of lumber and phosphorus
Phosphorus (Odum 1996)
Use average of rubber and plastic
Valyes
4.61E-09
1.57E4-09
7.53E^09
4.57Ei09
2.81E+09
3.69E+09
2.12E+09
2.60E-KJ9
4.02E4-09
3.31E4-09
6.64E-HO
3.31E-H)9
1.27EUO
5.03E+10
1.27E+10
3.15E+10
4.02E+09
7.76E-KJ9
7.55E-M7
2.9F/-H4
0.000384106
32805
9.83E+07
2.89E+12
1.45E-H1
4.02E-09
1.91 E-f- 10
5.43E-09
8.68E-09
1.74E-HO
4.17E-H)9
Units
sej/g
sej/g
sej/g
sej/g
sej'g
sej'g
sej/g
sej/g
sej/g
sej/g
sej/g
scj/g
sej/g
sej'g
sej'g
sej/g
sej/g
sej/g
grams
grams
J/g
sej/J
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
scj/g
-------
Table 5-C.4 - Specific emergies for exports. All commodity specific emergies are without services and relative to the
9.26E24 sej/y baseline. Values of the ernergy per unit factors for the commodity classes in this table were determined
from the data given in Tables 5-C. 1 and 5-C.3, as well as data provided in Table 5-C.5. The note numbers connect
to Table 5-C.5. Wwt = wet weight, otherwise numbers are dwt = dry weight. Any commodity class that is the same
as that found in the calculations for imports was determined in the same manner as shown in Table 5-C3 and the
calculation is not repeated here.
Note
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
New Commodity Classes
Field Crops
Grain
Oil Kernels. Nuts or Seeds
Miscellaneous Field Crops
Citrus Fruits
Fresh Vegetables
Horticultural Specialties
Metallic Ores
Iron Ores
Manganese Ores
Crude Petroleum
Broken Stone or Riprap
Gravel or Sand
Miscellaneous Nonmetallic Minerals
Meat, Fresh or Chilled
Meat. Fresh Frozen and Meat Products
Animal By-products, inedible
Flour or Other Grain Mill Products
Prepared or Canned Feed
Rice, Com flour and Cereal preparations
Dog, cat or Other Pet Food
Soybean Oil or By-products
Miscellaneous Food Preparations
Primary Forest Materials
Lumber or Dimension Stock
Treated Wood Products
Miscellaneous Wood Products
Newspapers
Miscellaneous Printed Matter
Industrial Chemicals
Potassium or Sodium Compound
Industrial Gases
Miscellaneous Industrial Organic Chemicals
Specific Emergv w/o
services
2.32E+08
2.94E-I-09
6.23E+09
2.32E-I-08
1.37E+08
3.58E+08
1.23E-I-09
1.77EH1
3.51E+09
3.50E^11
2.32E+09
4.90E+08
1.31E-I-09
1.96E+09
1.83EI 10
2.11E+10
9.87E+10
3.36E+09
5.07E+09
5.69E I 09
5.24E+09
3.11E+10
7.84E+09
3.77E-I-08
5.18EI08
2.02E+09
1.72E+09
4.26E+09
4.26E+09
O.OOE I 00
7.45E-I-09
9.63E+10
5.94E+09
Units
sej'g wwt
sej/g
sej/g
sej/g
scj/g wwt
sej/g
sej/ g wwt
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/ g wwt
sej'g wwt
sej/g wwt
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
-------
Note
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
New Commodity Classes
Miscellaneous Industrial Inorganic Chemicals
Miscellaneous Agricultural Chemicals
Miscellaneous Glassware, Blown or Pressed
Concrete Products
Ready-mix Concrete, Wet
Nonmelal Minerals, Processed etc.
Steel and Wire Products
Total Machinery - Motor V
Motor Vehicles - Transportation Equipment
Metal Scrap or Tailings
Textile and paper waste
Chemical or Petroleum Waste
Rubber or Plastic Scrap
Warehouse & Distribution Center
Rail and Air Drayage
Specific Emergy w/o
sendees
2.75E-H)9
1.45EUO
4.57E+09
1.32E+09
1.62E+09
3.69E-KJ9
4.02E-HJ9
7.76E-09
7.76E4-09
4.02E-H)9
8.78E 1 09
5.67E+09
4.17E+09
NA
NA
Units
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
-------
Table 5-C.5 - Specific emergies for imports. All commodity specific emergies are without sendees and relative to the
9.26E24 sej/y baseline.
Notes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Commodity Class, Calculations
Field Crops
Grain
Oil Kernels, Nuts or Seeds
Assume pecans
Miscellaneous Field Crops
Citrus Fruits
Fresh Vegetables
Horticultural Specialties
Average 3 arbitrary plants
Assume average plant sold is 5 years old
Metallic Ores
Iron Ores
Manganese Ores (MN)
Crude Petroleum
Broken Stone or Riprap
Gravel or Sand
Miscellaneous Nonmetallic Minerals
Meat, Fresh or Chilled
Meat, Fresh Frozen and Meat Products
Animal By-products, inedible
Flour or Other Grain Mill Products
Prepared or Canned Feed
Rice, Corn flour and Cereal preparations
Average (Rice ; Corn) I Milling energy
Dog. cat or Other Pet Food
Soybean Oil or By-products
Miscellaneous Food Preparations
Primary Forest Materials
Lumber or Dimension Stock
Treated Wood Products
Miscellaneous Wood Products
Newspapers
Source, Assumptions
Assume potatoes
Wheat
Rice
Oats
Average
(Brandt- Williams 2002)
Assume potatoes
Assume oranges
Same as imports
Cabbages
Lettuce
Sorghum
Average
Average Iron and MN
(Cohen et al. 2007)
(Cohen et al. 2007)
(Bastianoni ct al. 2005)
Same as Imports
Same as Imports
Assume clay, (Odum 1996)
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Values
232E-08
2.88E-r09
1.89E-09
4.05E-09
2.94E+09
6.23E-09
2.32E-08
1.37E'08
3.58E+08
2.29E-08
3.90E~r-08
1.19E-08
2.46E-08
1.23E-09
1.77E-H1
3.51F/-KJ9
3.50E+11
2.32E-09
4.90E+08
L31E-09
1.96E-09
1.83E-10
2. HE = 10
9.87E+10
3.36E-09
5.07E-09
5.69E-09
5.24E-09
3.11E-10
7.84E-09
3.77E~r-08
5.18E-08
2.02E-09
1.72E+09
4.26E-09
Units
sej/g wwt
sej/g
sej'g
scj/g
sej/g
sej/g
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/g
sej/g
sej'g
scj/g
sej/g
sej/g
sej/g
sej/ g wwt
sej/ g wwt
sej/ g wwt
sej/g
sej/g
sej/g
scj/g
sej/g
sej/g
sej/g
sej'g
scj/g
sej/g
sej/g
-------
Notes
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Commodity Class, Calculations
Miscellaneous Printed Matter
Industrial Chemicals
Potassium or Sodium Compound
Industrial Gases
Miscellaneous Industrial Organic Chemicals
Miscellaneous Industrial Inorganic Chemicals
Miscellaneous Agricultural Chemicals
Miscellaneous Glassware, Blown or Pressed
Concrete Products
Ready-mix Concrete, Wet
Nonmetal Minerals, Processed etc.
Steel and Wire Products
Total Machinery - Motor V
Motor Vehicles -I- Transportation Equipment
Metal Scrap or Tailings
Textile and Paper Waste
Chemical or Petroleum Waste
Rubber or Plastic Scrap
Source, Assumptions
(Campbell and Glut 2009)
Caustic soda (Pritchard 2000)
Hydrated lime (Pritchard 2000)
Diatomite (Pritchard 2000)
Sulfuric acid (Pritchard 2000)
Average
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Same as Imports
Assume Steel
Same as Imports
Assume machinery
Assume Steel
Average
Cotton (Odum 1996)
Paper, same as imports
Average coke, pesticides, glue
Same as Imports
Values
4.26E+09
2.68E+09
7.31E-I-09
1.31EI09
1.99E+09
8.96E+07
2.68E-I-09
7.45E-I-09
9.63E+10
5.94E+09
2.75E-I-09
1.45E-HO
4.57E+09
1.32E+09
1.62EI09
3.69E-I-09
4.02E+09
7.76E+09
7.76E-09
4.02E'09
8.78E-09
1.33E-10
4.26E-09
5.67E-09
4.17E-09
Units
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej'g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej/g
sej'g
-------
References
Bastianoiii, S., Marchettini. N., Panzieri, M., Tiezzi,
E. 2001. Sustainabiiity assessment of a farm in the
Chianti area (Italy). Journal of Cleaner Production 9,
365-373.
Bastianoni, S., Campbell. D., Susani, L., Tiezzi. E.
2005. The solar transformity of oil and petroleum
natural gas. Ecological Modelling 186,212-220.
Bastianoni, S., Campbell. D.E., Ridolfi, R.. Pulselli,
P.M. 2009. The solar transformity of petroleum fuels.
Ecological Modelling 220, 40-50.
Brandt-Williams, S.L. 2002. Handbook of Emcrgy
Evaluation. Folio #4. Emergy of Florida Agriculture.
Center for Environmental Policy, Environmental
Engineering Sciences, University of Florida,
Gainesville, Florida.
Brown, M.T., Arding, J. 1991. Transformative Working
Paper. Center for Wetlands, University of Florida,
Gainesville, Florida.
Brown, M.T., Green, P., Gonzalez, A., Venegas, J. 1992.
Emergy Analysis Perspectives, Public Policy Options,
and Development Guidelines for the Coastal Zone
of Nay aril, Mexico Volume 2: Emcrgy Analysis and
Public Policy Options. Center for Wetlands and Water
Resources, University of Florida.
Brown, M.T., Buranakarn, V. 2000. Emergy evaluation
of material cycles and recycle options, in: Brown,
M.T., Brandt-Williams, S.L,, Tilley, D.R., Ulgiati, S.
(Eds.), Emergy Synthesis, Proceedings of the First
Biennial Emergy Analysis Research Conference.
The Center for Environmental Policy, Department of
Environmental Engineering Sciences, University of
Florida, Gainesville, Florida, pp. 141-154.
Buranakarn, V. 1998. Evaluation of recycling and
reuse of building materials using the emergy analysis
method. Ph.D. Dissertation, University of Florida,
UMI Dissertation Services, Ann Arbor, Michigan.
Campbell, D.E. 2003a. A note on the uncertainty in
estimates of transformities based on global water
budgets, in: Brown, M.T., Odum, H.T., Tilley, D.R.,
Ulgiati, S. (Eds.), Emergy Synthesis 2. Proceedings
of the Second Biennial Emergy Analysis Conference.
Center for Environmental Policy, University of
Florida, Gainesville, Florida, pp. 349-353.
Campbell, D.E. 2003b. Emergy Analysis of the
prehistoric global nitrogen cycle, in: Brown,
M.T., Odum, H.T.. Tilley. D.R., Ulgiati. S. (Eds.),
Emergy Synthesis 2, Proceedings of the Second
Biennial Emergy Analysis Conference. Center
for Environmental Policy, University of Florida,
Gainesville, Florida, pp. 221-239.
Campbell, D.E., Brandt-Williams, S.L., Meisch, M.E.A.
2005. Environmental accounting using emergy:
evaluation of the State of West Virginia. USEPA
Research Report, EPA/600/R-05/006.
Campbell, D.E., Cai, T.T. 2007. Emergy and economic
value, in: Brown. M.T., Bardi, E., Campbell, D.,
Comar, V.. Huang. S.L., Rydberg. T, Tilley. D.,
Ulgiati, S. (Eds.), Emergy Synthesis 4: Theory and
Application of the Emergy Methodology, Proceedings
of the 4th Biennial Emergy Conference, Gainesville,
Florida, pp. 24.1-24.16.
Campbell, D.E., Lu. H.F. 2009. The emergy to money
ratio of the United States from 1900 to 2007. in:
Brown, M.T., Sweeney, S., Campbell, D.E., Huang,
S.L., Ortega, E, Rydberg, T, Tilley, D., Ulgiati, S.
(Eds.), Emergy Synthesis 5, Theory and Applications
of (lie Emergy Methodology, Proceedings of the
5lh Emcrgy Research Conference. The Center for
Environmental Policy, Department of Environmental
Engineering Science, University of Florida,
Gainesville, Florida, pp. 413-448.
Campbell, D.E., Ohrt. A. 2009. Environmental
accounting using emergy: evaluation of Minnesota.
USEPA Project Report, EPA/600/R-09/002, 138 pp.
Carlsson-Kanyama, A., Faist, M. 2000. Energy use
in the food sector: a data survey. Foundation for
Strategic Environmental Research. Stockholm,
Sweden, Swiss Federal Institute of technology, Zurich,
Switzerland. Available online at http://www.infra.kth.
se/fms/pdf/energyuse.pdf. Last accessed on August 6,
2010.
Cavalett, O., Ferraz de Queiroz, J., Ortega, E. 2006.
Emergy assessment of integrated production systems
of grains, pig and fish in small farms in the South
Brazil. Ecological Modelling 193, 205-224.
Cohen, M.J., Sweeney, S., Brown, M.T. 2007.
Computing the unit emergy value for crustal
elements, in: Brown. M.T, Bardi, E. Campbell, D.E..
Comar, V, Huang, S.L. Rydberg, T, Tilley, D.R.,
Ulgiati, S. (Eds.), Emcrgy Synthesis 4, Theory and
Applications of the Emergy Methodology. Center
for Environmental Policy. University of Florida.
Gainesville, Florida, pp. 16.1-16.11.
-------
Haukoos, D.S. 1995. Sustainable architecture and its
relationship to industrialized building. M.S. Thesis in
Architectural Studies, Department of Environmental
Engineering Sciences, University of Florida,
Gainesville, Florida.
Odum, H.T. 1996. Environmental Accounting: Emergy
and Environmental Decision Making. John Wiley &
Sons, New York.
Odum, H.T. 2000. Emergy evaluation of salmon pen
culture. Paper presented at the International Institute
of fisheries Economics and Trade. Macrobehavior,
Macroresults, and Externalities: Conceptual Issues.
July 10-15, Oregon State University, Corvallis,
Oregon. Available online at http://oregonstate.edu/
dept/iifet/2000/papers/odum.pdf. Last accessed on
September 3, 2009.
Odum, H.T. 2002. Material circulation, energy
hierarchy, and building construction, in: Kibert, C.J.,
Sendzimir, J. Guy, G.B. (Eds.), Construction Ecology,
Nature as the Basis for Green Buildings. SPON Press,
Taylor and Francis, London, New York, pp. 37-71.
Odum, H.T, Pinkerton, R.C. 1955. Time's speed
regulator: the optimum efficiency for maximum power
output in physical and biological systems. American
Scientist 43, 331-343.
Odum, H.T, Brown, M.T., Christiansen, R.A. 1986.
Energy Systems Overview of the Amazon Basin.
Report to the Cousteau Society, Center for Wetlands,
CFW Publication #86-1, University of Florida,
Gainesville, Florida.
Odum, H.T, Odum, B.C., Blissett, M. 1987. The
Texas System, Emergy Analysis and Public Policy. A
Special Project Report, L.B. Johnson School of Public
Affairs. University of Texas at Austin and Office of
Natural Resources, Texas Department of Agriculture,
Austin.
Odum, H.T, Romitelli, S., Tigne, R. 1998. Evaluation
Overview of the Cache River and Black Swamp
in Arkansas. Center for Environmental Policy,
Environmental Engineering Sciences, University of
Florida, Gainesville, Florida.
Odum, H.T, Brown, M.T., Williams, S.B. 2000.
Handbook of Emergy Evaluation: A Compendium of
Data for Emergy Computation Issued in a Series of
Folios. Folio #1 - Introduction and Global Budget.
Center for Environmental Policy, Environmental
Engineering Sciences, Univ. of Florida, Gainesville,
16pp.
Pritchard, L. 2000. Transformities used in calculations,
in: Odum, H.T. (Ed.), Heavy Metals in the
Environment: Using Wetlands for their Removal.
Lewis Publishers, Boca Raton, Florida, pp.261-268.
Tilley, D.R. 1999. Emergy Basis of Forest Systems,
PhD. Dissertation, University of Florida, UMI
Dissertation Services, Ann Arbor, Michigan.
-------
Appendix 5-D: Further of
the Luis
The major physiographic systems within the SLB are
aggregated into three classes: 1) mountains which are
to a large extent forested, 2) the valley floor, which
is divided into an area covered by the natural shrub
vegetation of the region including phreatophytes and
xerophytes (e.g., sagebrush), and 3) agricultural land,
which is generally irrigated with surface or groundwater.
Cottonwood and other trees grow along riparian habitat
and are included in the symbol for shrub lands vegetation
(Fig. 5.1). Each vegetation subsystem includes surface
and groundwater with flows appropriate for the location
(i.e., considering drainage). Groundwater storages
include a specified temperature, because geotliermal
energy is a potentially important energy source in
the region. Much of the valley floor is occupied by
agriculture, growing mainly grains and potatoes.
Specialized agricultural crops (e.g.. quinoa) that require
high altitude can be grown in the SLB, which has an
average elevation of approximately 2300 m. Some
livestock, mostly cattle, are raised in the valley and are
included in the agricultural subsystem.
The high elevation valley is surrounded by still higher
mountains with peaks that rise more than 1800 m above
the valley floor (San Luis Valley Development Resources
Group 2007). The work of geological processes deep
in the earth has given rise to many of the present salient
features of the SLB. The SLB itself was originally
formed as a rift valley (Chapin and Cather 1994). Also
in the past, various periods of orogeny gave rise to the
mountains with stores of rock and concentrated mineral
deposits (e.g., perlite) suitable for mining. Erosion of
the mountains over millions of years filled the deep rift
valley with sediments, which are deeply bedded with
layers of ground water present throughout the sediment
deposit (CWCB 2000, Mayo et al. 2006). In recent
times, sandy plains and the unique wind regime of the
basin along with the dogleg shape of the Sangre de
Cristo range on the eastern border of the valley have
given rise to a dune field containing the highest sand
dunes in North America (San Luis Valley Development
Resources Group 2007). Seasonal river flow from the
Sangre de Cristo Mountains plays an important part in
the maintenance of these dunes (Madole et al. 2008) as
indicated by the interaction symbol in the model (Fig.
5.1) describing dune production as an interaction of the
sand supply on the valley floor with wind and water
flows. Finally, the high mountains surrounding the
valley ensure that a large part of the water that falls on
the basin comes down as snow (Emery 1979).
The primary economic systems of the region are
centered on its natural resources (e.g., food processing
and shipment, mining and processing ores and building
materials, and recreation). Service and commerce,
education, transportation, and government are connected
within the system network (Fig. 5.1) as noted by the
pathways defined in Table 5.2. In addition, people are
an important storage of capital within the system and we
documented the effects of immigration and emigration
on the annual ernergy inflows to and outflows from the
SLB. For example, in 1989 the emergy brought into the
SLB by immigrants exceeded the emergy of petroleum
used in that year. Net movement of people appeared
to follow economic opportunity, so managers might
indirectly work to increase the movement of people to
the SLB by zoning land for industrial and commercial
use. Also, elected officials might alter the tax structure
to attract more or less business to the region. Tourists
enter and leave the system and are apparently attracted
by the extraordinary features in the natural environment
of the region (e.g., Great Sand Dunes National Park and
Preserve, various wildlife preserves with rare species,
hot springs, etc.).
References
Chapin, C.E., Cather, S.M. 1994. Tectonic setting of
the axial basins of the northern and central Rio Grande
rift, in: Keller, G.R., Cather, S.M. (Eds.), Basins of the
Rio Grande Rift: Structure, Stratigraphy, and Tectonic
Setting, Geological Society of America, Special Paper
291, Boulder, CO, pp. 5-25.
Colorado Water Conservation Board (CWCB). 2000.
Rio Grande Decision Support System: An Overview.
Section 2.0 Rio Grande Basin Conditions. CWCB,
State of Colorado. Available online at http://cdss.state.
co.us/DNN/RioGrande/tabid/57/Default.aspx. Last-
Accessed July 21, 2009.
Emery. P. 1979. Geohydrology of the San Luis Valley,
Colorado, USA. Proceedings of the Canberra
Symposium 1979. IAHS-AISH Publication no. 128,
297-305.
Emery, P. A. 1996. Hydrology of the San Luis Valley,
Colorado, an overview and a look at the future, in:
Thompson, R.A., Hudson, M.R., Pillmore, C.L. (Eds.),
Geologic Excursions to the Rocky Mountains and
Beyond. Field Trip Guidebook for the 1996 Annual
Meeting, Geologic Society of America, Oct. 28 1996-
Oct. 31 1996, Denver, Colorado. Colorado Geological
Survey, Department of Natural Resources. Denver,
CO. Available online at http://www.nps.gov/archive/
grsa/resources/docs/Trip2023.pdf. Last accessed
August 4, 2010.
-------
Madoie, R.F., Roniig, J.H., Aleinikoff, J.N., VanSistine,
D.P., Yacob, E.Y. 2008. On the origin and age of the
Great Sand Dunes, Colorado. Geomorphoiogy 99,
99-119.
Mayo, A.L., Davey, A., Christiansen, D. 2007.
Ground-water flow patterns in the San Luis Valley,
Colorado US A revisited: an evaluation of solute and
isotopic data. Hydrogeology Journal 15, 383-408.
San Luis Valley Development Resources Group. 2007.
Comprehensive economic development strategy.
Available online at http://www.slvdrg.org/ceds.php.
Last accessed June 17, 2009.
-------
Appendix 5-E: Emergy Analysis For Inflows, Production,
Productions And Imports And And Additional Figures Not Included In Of
Table 5-E. 1 - Renewable emergy inflows to the San Luis Basin.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Solar Energy
Absorbed
(sej/y)
1.26E+20
1.31E+20
1.23E+20
1.21E+20
1.19E-I-20
1.25E420
1.23E+20
1.22E-I-20
1.26E+20
1.27E+20
1.30E+20
1.21E+20
1.16E+20
1.09E-I-20
1.04E+20
1.10E+20
1.19E-I-20
1.31E+20
1.3 IE -1-20
1.42E+20
1.41E+20
1.41E+20
1.39E+20
1.42E+20
1.39E+20
1.42E+20
Wind Energy
Absorbed
(sej/y)
6.98E-20
1.05E-21
1.05E-21
1.32E-21
LOSE 1-21
1.05E+21
9.75E+20
9.75EI-20
6.98E-20
8.29E-20
8.29E-20
6.68E-20
6.38E-20
3.12E-20
4.33E+20
2.78E+20
5.29EI-20
7.62E+20
4.80EI-20
6.38E-20
6.98E-20
6.38E-20
4.80E-20
6.98E-20
6.38E+20
5.82E+20
Earth Cycle
(sej/y)
2.30E-I-21
2.30E+21
2.30E+21
2.30E+21
2.30E-I-21
2.30FJ-21
2.30E+21
2.30E-I21
2.30E+21
2.30E+21
2.30E-I-21
2.30E+21
2.30E+21
2.30E-I-21
2.30E+21
2.30E+21
2.30E-I21
2.30E+21
2.30E-I21
2.30E-I-21
2.30E+21
2.30E+21
2.30E+21
2.30E-I-21
2.30E+21
2.30E+21
Rain
Chemical
Potential
(sej/y)
1.02E-i21
1.05E+21
1.09E+21
1.11E+21
1.13E-i21
1.21E+21
1.17E+21
9.50E-I-20
9.50E+20
7.50E+20
1.29E-i21
1.02E+21
9.86E+20
1.08E-i21
1.08E+21
1.05E+21
9.16E-I-20
1.21E+21
1.02E-I-21
1.04E-t21
9.38E+20
9.45E+20
6.12E+20
8.74E-i20
9.86E+20
9.42E+20
Rain
Geopoteiitial
on the Land
(sej/y)
1.73E+21
1.77E+21
1.84E+21
1.86E+21
1.92E+21
2.05E+21
1.97E+21
1.60E-I-21
1.60E+21
1.27E+21
2.16E+21
1.72E+21
1.65E+21
1.81E+21
1.81E+21
1.77E+21
1.55E-I-21
2.04E+21
1.72E-I-21
1.76E+21
1.59E+21
1.59E+21
1.03E+21
1.47E+21
1.66E+21
1.59E+21
Snow
Geopotential
on Land
(sej/y)
1.73E-22
1.77E-22
1.84E-22
1.86E-22
1.92E-22
2.05E-22
1.97E-22
1.60EI-22
1.61E+22
1.27E-22
2.17E-22
1.72E-22
1.66E-22
1.81E-22
1.81E-22
1.78E-22
1.55E-22
2.04E+22
1.72EI-22
1.76E-I-22
1.59E-22
1.59E-22
1.03E-22
1.47E-22
1.67E-22
1.59E-22
Rain
Geopotential
as Runoff
(sej/y)
2.35EI-20
2.44E+20
2.56E+20
2.54E+20
2.59E-I-20
2.87E+20
2.75E+20
2.25E-I-20
2.23E+20
1.75E+20
3.08E-I-20
2.46E+20
2.36E+20
2.58E-I-20
2.57E+20
2.46E+20
2.08E-I-20
2.90E+20
2.43E-I20
2.44EI-20
2. 17E+20
2.25E+20
1.42E+20
2.06E-I-20
2.33E+20
2.22E+20
Snow
Geopotential
as Runoff
(sej/y)
1.97E-I21
2.05E+21
2.14E+21
2.13E+21
2.17E-t21
2.41 E+21
2.31E+21
1.89E-t21
1.87E+21
1.47E+21
2.58E-I21
2.06E+21
1.98E+21
2.16E-t21
2.16E+21
2.06E+21
1.74E-t21
2.43E+21
2.04E-I-21
2.05E-I21
1.82E+21
1.88E+21
1.19E+21
L72E-i21
1.95E+21
1.86E+21
Total
Geopotential
as Runoff
(sej/y)
2.20E-I-21
2.29E+21
2.40E+21
2.38E+21
2.43E+21
2.69E-f-21
2.58E+21
2. HE -1-21
2.09E+21
1.65E+21
2.89E-I-21
2.30E+21
2. 21 E+21
2.42E+21
2.41E+21
2.30E+21
1.95E+21
2.72E+21
2.28E+21
2.29E-I-21
2.04E+21
2.11E+21
1.33E+21
1.93E+21
2.18E+21
2.08E+21
Evapo-
transpiration
(sej/y)
l.OOE-l-21
9.78E+20
9.50E420
9.46E-20
9.43E-20
9.40E-20
9.36E-20
9.33E-20
9.35E+20
9.38E+20
9.40E-I-20
9.43E+20
9.45E-20
9.43E-20
9.41E-20
9.38E-20
9.36E-20
9.34E-20
9.3 IE -20
9.29E-I-20
9.26E+20
9.23E+20
9.21E420
9.18E-20
9.15E-20
9.13E-20
Renewable*
Emergy
Base
(sej/y)
3.20EI-21
3.27E+21
3.35E+21
3.33E+21
3.37E-I-21
3.63E+21
3.52E+21
3.04EI-21
3.03E+21
2.58E+21
3.83EI-21
3.24E+21
3.16E+21
3.36E-I-21
3.35E+21
3.24E+21
2.88E-I-21
3.65E+21
3.21E-I-21
3.22EI-21
2.96E+21
3.03E+21
2.25E+21
2.85EI-21
3.10E+21
2.99E+21
* The Renewable emergy base for the systems consists of the sum of the geopotential of runoff'from rain and snow and the evapotranspiration.
-------
Table 5-E.2 - Production in the San Luis Basin that is based primarily on renewable resources.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Agricultural
Crops (Total)
(sej/y)
1.32E-21
1.36E-21
1.41E-21
1.42E-21
1.42E-21
1.56E-21
1.70E-21
1.55E-21
1.54E-21
1.66E-21
1.55E+21
1.42EI-21
1.50E-21
1.43E-21
1.54E-21
1.48E-21
1.73E-21
1.90E-21
1.79E-21
1.87E-21
2.00E-21
1.86E-21
1.42E+21
1.35E+21
1.29EI-21
1.49E-21
Hay
(sej/y)
4.76E+20
3.52E+20
3.57E+20
1.79E+20
2.03E+20
2.65E+20
5.28E+20
6.04E+20
6.06E+20
4.79E+20
5.48E+20
5.25E-i20
5.74E+20
6.46E+20
6.87E+20
6.23E+20
5.62E+20
8.40E+20
8.00E+20
8.07E+20
8.15E+20
9.63E+20
6.73E+20
7.10E+20
7.32E+20
8.98E+20
Barley
(sej/y)
2.30E+20
2.41E+20
3.03E+20
2.79E-I-20
3.33E+20
3.11E-20
4.30E+20
2.14E+20
1.87E+20
2.20E+20
2.75E+20
2.50E-I-20
2.07E+20
1.44E+20
1.52E+20
2.08E+20
1.60E+20
2.34E+20
1.96E-I-20
1.95E-I-20
2.72E-I-20
2.01E+20
1.66E+20
1.87E+20
1.89E-I-20
1.96E-I-20
Oats
(sej/y)
1.13E+20
1.37E+20
1.73E+20
1.81E-I20
1.80E+20
2. 21 E+20
1.35EH-20
9.71E+19
1.43EH-20
1.69EH-20
1.62E+20
1.19E-I-20
1.04E+20
9.41E+19
8.28EH-19
1.38EH-20
1.48E+20
1.32E+20
1.50E-I20
1.32E-I20
2.18E-I20
1.36E+20
7.31E+19
8.83EH-19
4.36E-1 19
3.54E-M9
Wheat
(sej/y)
3.85E-t-20
5.05E+20
4.43E+20
6.30E+20
5.27E+20
5.70E+20
4.07E+20
4.34E+20
4.05E+20
5.71E+20
3.20E+20
2.76E+20
3.78E+20
2.77E+20
3.49E+20
2.56E+20
5.56E+20
4.33E+20
3.72E+20
4.60E+20
4.04E+20
3.32E+20
2.18E+20
1.19E+20
9.39E + 19
1.36E+20
Potatoes
(seJ/57)
1.16E+20
1.22E+20
1.35E+20
1.47EI-20
1.82E+20
1.89E+20
1.99E+20
2.06E+20
2.01 E+20
2.17E+20
2.40E+20
2.5 IE 1-20
2.33E+20
2.67E+20
2.72E+20
2.51E+20
3.08E+20
2.64E+20
2.68E-I-20
2.72E-I-20
2.95E-I-20
2.25E+20
2.94E+20
2.50E+20
2.27E-I-20
2.22EI-20
Livestock
(Cattle)
(sej/y)
6.52E+20
7.03E+20
6.89E+20
7.45E-I-20
7.80E+20
7.56E+20
7.47E+20
7.43E+20
7.39E+20
7.34E+20
7.30E+20
7.25E-I-20
7. 21 E+20
7.16E+20
7. 11 E+20
7.06E+20
7.02E+20
6.96E+20
6.91 E+20
6.86E-I-20
6.81E-I-20
6.81E+20
5.74E+20
5.59E+20
4.66E-I-20
4.98E-I-20
Fish
(sej/y)
O.OOE-00
O.OOE-00
O.OOE-00
O.OOE-00
O.OOE-00
4.73E-16
3.07E-17
3.07E-17
3.31E-17
3.31E+17
3.31E+17
7.09E-17
8.99E-17
8.99E-17
8.99E-17
8.99E-17
8.99E-17
1.37E-18
1.49E-18
1.23E-18
1.51E-18
1.75E+18
1.80E+18
1.84E+18
2.13E-18
2.06E-18
Timber Harvest
(sej/y)
7.43E+19
4.38E+19
7.44E+19
7.78E-! 19
9.54E+19
9.54E+19
1.08E+20
1.02E+20
1.03E+20
1.09E+20
7.40E+19
7.22E-! 19
7.76E+19
8.16E+19
1.13E+20
6.69E+19
3.43E+19
1.20E+19
1.35E-U9
2.09E-! 19
1.71E-U9
9.99E+18
8.30E+18
1.21E+19
1.95E-U9
1.38E-U9
-------
Table 5-E.3 - Nonrenewable emergy inflows in the San Luis Basin.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Coal Used
(sej/y)
2.28E+19
2.09E+19
L48E+19
9.62EM8
1.16EM9
1.10EM9
1.03E+19
9.68E+18
9.07E+18
8.18E+18
8.97E+18
9.06E+18
8.65E+18
8.70EM8
9.31EM8
7.94E+18
4.01E+18
8.27E+18
4.23E-18
5.40E-18
5.11E+18
5.91E-18
4.21E-18
5.40E-18
5.00E-18
4.13E-18
Natural Gas
Used
(sej/y)
1.36E+20
1.19E+20
1.30E+20
1.23E-I20
1.32E-I20
1.24E-I20
1.12E+20
1.17E+20
1.28E+20
1.34E+20
1.32E+20
1.43E+20
1.34E+20
1.47E-I20
1.36E-I20
1.40E+20
1.51E+20
1.49E+20
1.52E+20
1.47E+20
1.52E+20
1.84E-I20
1.84E-I20
1 .72E-I20
1.72E+20
1.78E+20
Petroleum
Fuels Used
(sej/y)
2.91E+20
2.66E+20
2.70E+20
2.76E+20
2.76E+20
2.73E+20
2.78E+20
2.78E+20
2.82E+20
2.67E+20
2.66E+20
2.69E+20
2.66E+20
2.80E+20
2.84E+20
2.91E+20
3.01E+20
2.84E+20
3.00E+20
3.06E+20
3.26E+20
3.30E+20
3.15E+20
3.26E+20
3.53E+20
3.42E+20
Electricity
Used
(sej/y)
1.69E+20
1.91E+20
1.87E+20
1.90E-I-20
2.01E-I-20
2.06E-I-20
2.10E+20
2.17E+20
2.28E+20
2.31E+20
2.32E+20
2.35E+20
2.30E+20
2.32E-I-20
2.42E+20
2.48E+20
2.59E+20
2.63E+20
2.71E+20
2.73E+20
2.87E+20
2.87E-I-20
2.97E-I-20
2.99E-I-20
3.00E+20
3.06E+20
Crude Oil
Produced*
(sej/y)
7.28E+15
1.34E+18
2.44E+18
5.62E+18
2.33E+18
6.62E+17
7.88E+ 17
533E-! 17
8.70E-t 17
1.10E+18
2.27E+18
Metallic
Ores*
(sej/y)
1.49E+20
1.38E+20
1.54E-t-20
1.46E+20
1.78E+20
1.06E+20
2.3 IE +20
7.64E+20
3.94E + 19
3.66E+19
4.80E+19
Broken
Stone, Rip
Rap*#
(sej/y)
1.24E+21
1.14E+21
1.22E+21
1.29E+21
6.77E+20
8.27E+20
9.50EI-20
5.06EI-20
5.43EI-20
7.93E+20
8.26E+20
Sand and
Gravel*#
(sej/y)
3.63E+21
3.35E+21
3.58E+21
3.67E+21
2.74E+21
2.75E+21
3.09E+21
2.52E+21
1.95E+21
1.95E+21
1.86E+21
Non-
metallic
Minerals*
(sej/y)
1.82E+20
1.68E+20
1.80E+20
1.99E+20
1.75E+20
2.27E+20
2.31E+20
2.29E-I-20
6.31E+19
5.94E+20
6.21E+20
Soil Erosion
(sej/y)
1.87E+20
2.01E+20
1.95E+20
1.96EI-20
2.00EI-20
1.85EI-20
1.70E+20
1.43E+20
1.41E+20
1.43E+20
1.35E+20
1.33E+20
1.28E+20
1.26EI-20
1.23E+20
1.24E+20
1.34E+20
1.34E+20
1.45E+20
1.34E+20
1.28E+20
1.19EI-20
1.20EI-20
1.26EI-20
1.26E+20
1.25E+20
Ground -
water
(sej/y)
O.OOE+00
2.63E+20
O.OOE+00
O.OOE-lOO
O.OOE-lOO
O.OOE-lOO
O.OOE+00
O.OOE+00
2.75E+20
2.56E+20
3.59E+19
O.OOE+00
O.OOE+00
O.OOE-lOO
2.62E+19
O.OOE+00
4.00E+20
O.OOE+00
1.0 IE +20
O.OOE+00
4.62E+20
O.OOE+00
7.89E+20
2.77E-I-20
2.00E+19
O.OOE+00
* Assumed from Global Insight Inc. data to be produced in the Basin. " Ten percent was added for local consumption.
-------
Table 5-E.4 - Imports into the San Luis Basin.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Coal
(sej/y)
2.28E+19
2.09E+19
1.48E+19
9.62E+18
1.16E+19
1.10E+19
1.03E+19
9.68E+18
9.07E+18
8.18E+18
8.97E+18
9.06E+18
8.65E+18
8.70E+18
9.31E+18
7.94E+18
4.01E+18
8.27E+18
4.23E+18
5.40E+18
5.11E+18
5.91E+18
4.21E+18
5.40E+18
5.00E+18
4.13E+18
Petroleum
(sej/y)
1.36E+20
1.19E+20
1.30E+20
1.23E+20
1.32E+20
1.24E+20
1.12E+20
1.17E+20
1.28E+20
1.34E+20
1.32E+20
1.43E+20
1.34E+20
1.47E+20
1.36E+20
1.40E+20
1.51E+20
1.49E+20
1.52E+20
1.47E+20
1.52E+20
1.84E+20
1.84E+20
1.72E+20
1.72E+20
1.78E+20
Natural
Gas
(sej/y)
2.91E+20
2.66E+20
2.70E+20
2.76E+20
2.76E+20
2.73E+20
2.78E+20
2.78E+20
2.82E+20
2.67E+20
2.66E+20
2.69E+20
2.66E+20
2.80E+20
2.84E+20
2.91E+20
3.01E+20
2.84E+20
3.00E+20
3.06E+20
3.26E+20
3.30E+20
3.15E+20
3.26E+20
3.53E+20
3.42E+20
Electricity
(sej/y)
1.69E+20
1.91E+20
1.87E+20
1.90E+20
2.01E+20
2.06E+20
2.10E+20
2.17E+20
2.28E+20
2.31E+20
2.32E+20
2.35E+20
2.30E+20
2.32E+20
2.42E+20
2.48E+20
2.59E+20
2.63E+20
2.71E+20
2.73E+20
2.87E+20
2.87E+20
2.97E+20
2.99E+20
3.00E+20
3.06E+20
Total
Fuels and
Electricity
(sej/y)
6.19E+20
5.97E+20
6.01E+20
5.98E+20
6.21E+20
6.13E+20
6.10E+20
6.22E+20
6.46E+20
6.40E+20
6.40E+20
6.56E+20
6.38E+20
6.68E+20
6.71E+20
6.87E+20
7.15E+20
7.04E+20
7.27E+20
7.32E+20
7.71E+20
8.08E+20
8.00E+20
8.02E+20
8.30E+20
8.30E+20
Materials
other than
fuels
(sej/y)
1.33E+21
1.39E+21
1.46E+21
1.48E+21
1.82E+21
1.83E+21
2.13E+21
2.20E+21
1.53E+21
1.65E+21
1.70E+21
Services in
materials
other than
fuels
(sej/y)
6.26E+20
5.92E+20
6.13E+20
6. 17E+20
6.46E+20
6.85E+20
6.47E+20
6.22E+20
5.32E+20
5.11E+20
4.99E+20
Services in
Fuels
(sej/y)
2.37E+20
2.27E+20
2.07E+20
1.74E+20
1.72E+20
1.57E+20
1.14E+20
1.19E+20
1.17E+20
1.17E+20
1.23E+20
1.15E+20
1.08E+20
1.14E+20
1.18E+20
1.22E+20
1.33E+20
1.30E+20
1.15E+20
1.18E+20
1.61E+20
1.71E+20
1.33E+20
1.46E+20
1.89E+20
2.23E+20
Services in
Electricity
(sej/y)
6.76E+19
7.99E+19
7.93E+19
7.11E+19
7.38E+19
7.19E+19
6.57E+19
6.56E+19
7.02E+19
6.72E+19
6.43E+19
6.14E+19
5.80E+19
5.86E+19
6.23E+19
6.46E+19
6.65E+19
6.55E+19
6.51E+19
6.13E+19
6.48E+19
6.19E+19
6.03E+19
6.62E+19
6.94E+19
7.30E+19
Total
Goods and
Services
(sej/y)
2.83E+21
2.90E+21
2.97E+21
3.00E+21
3.38E+21
3.51E+21
3.81E+21
3.82E+21
3.07E+21
3.25E+21
3.32E+21
Immigration
(sej/y)
1.23E+20
O.OOE+00
3.21E+19
1.80E+20
3.80E+20
2.93E+20
1.63E+19
7.39E+19
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
O.OOE+00
9.16E+19
1.45E+20
Tourism
(sej/y)
1.44E+20
1.49E+20
1.52E+20
1.52E+20
1.47E+20
1.55E+20
1.52E+20
1.39E+20
1.43E+20
1.50E+20
1.40E+20
Federal
and State
Outlays
(sej/y)
4.89E+19
5.09E+19
5.31E+19
5.48E+19
5.12E+19
4.98E+19
4.77E+19
4.85E+19
5.24E+19
5.86E+19
6.12E+19
6.52E+19
6.26E+19
5.91E+19
5.59E+19
5.96E+19
5.82E+19
5.60E+19
5.97E+19
-------
Table 5-E.5 - Exports from the San Luis Basin.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Total
Materials
Exported
(sej/y)
7.49E+21
7.55E+21
7.83E+21
7.94E+21
6.19E+21
6.30E+21
6.12E+21
5.52E+21
4.22E+21
9.36E+21
9.84E+21
Services in
material
exports
(sej/y)
2.07E+21
2.18E+21
2.32E+21
2.34E+21
2.02E+21
1.95E+21
1.21E-I-21
1.49E-I-21
1.87E+21
7.18E+21
7.24E+21
Total
Exports
(sej/y)
9.57E+21
9.73E+21
1.01E+22
1.03E+22
8.21E+21
8.25E+21
7.33E+21
7.00E+21
6.09E+21
1.65E+22
1.71E+22
Agricultural
Crops
(sej/y)
7.72E+20
9.25EH-20
9.53EH-20
1.14EH-21
1.11E+21
1.17E+21
1.06E+21
8.87EH-20
8.64EH-20
1.09E+21
9.04E+20
8.19E+20
8.58E+20
7.30E+20
8.07EH-20
7.88EH-20
1.11E+21
9.95E+20
9.22EH-20
l.OOEH-21
1.10E+21
8.35E-I20
7.24E-I20
6.10E+20
5.31EH-20
5.72E+20
Potatoes
(sej/y)
1.16E-20
1.22E-20
1.35E-20
1.47E-20
1.82E-20
1.89E-20
1.99E-20
2.06E-20
2.01 E-20
2.17E-20
2.40E-20
2. 51 E-20
2.33E-20
2.67E-20
2.72E-20
2.51E-20
3.08E-20
2.64E-20
2.68E-20
2.72E-20
2.95E-20
2.25E-20
2.94E-20
2.50E-20
2.27E-20
2.22E-20
Grain,
Wheat,
Barley, Oats
(sej/y)
6.57E+20
8.03E+20
8.18E+20
9.91E+20
9.32E+20
9.85E+20
8.60E+20
6.81E+20
6.63E+20
8.74E+20
6.64E+20
5.68E+20
6.24E+20
4.63E+20
534E+20
5.27E+20
7.95E+20
7.21E+20
6.42E+20
7.16E+20
7.85E+20
5.95E-t20
4.08E-t20
3.38E+20
2.82E+20
3.24E+20
Vegetables,
Fruit, Nuts,
Seeds
(sej/y)
1.08E-18
1.13E-18
1.53E-18
1.34E-18
1.52E-18
1.46E-18
1.78E-18
1.24E-19
1.39E-19
1.32E-19
1.36E-19
Horticulture
Specialties
(sej/y)
8.63E+18
8.18E+18
9.19E+18
1.08E+19
1.20E+19
1.44E+19
1.29E-H9
9.47E-H8
8.56E+18
8.41EH-18
1.20E+19
Livestock
(sej/y)
6.52E+20
7.03E+20
6.89E+20
7.45E+20
7.80E+20
7.56E+20
7.47E+20
7.43E+20
7.39E+20
7.34E+20
7.30E+20
7.25E+20
7.21E+20
7.16E+20
7.11E420
7.06E+20
7.02E+20
6.96E+20
6.91E+20
6.86E+20
6.81E+20
6.81 E+20
5.74E+20
5.59E+20
4.66E+20
4.98E+20
Minerals
Total
(sej/y)
4.76E+21
439E+21
4.70E+21
4.86E+21
3.46E+21
3.58E+21
4.13E-t21
3.74E-t21
2.37E+21
3.13E+21
3.11E+21
Metallic
Ores
(sej/y)
1.49E+20
1.38E+20
1.54E+20
1.46E+20
1.78E+20
1.06E+20
2.31EI-20
7.64E-20
3.94E-19
3.66E-19
4.80E-19
-------
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Crude
Petroleum
(scj/y)
7.19E-H5
1.33E-I-18
2.41E-I-18
5.55E+18
2.30E+18
6.54E+17
7.78E + 17
5.26E+17
8.59E+17
1.08E+18
2.25E + 18
Broken Stone
Or Riprap
(scj/y)
1.13E+21
1.04E+21
U1E-I-21
1.18E+21
6.16E+20
7.51E+20
8.64E+20
4.60E+20
4.94E+20
7.21E+20
7.51E-I-20
Gravel Or Sand
(scj/y)
3.30E+21
3.05E-I-21
3.26E-I-21
3.34E+21
2.49E+21
2.50E+21
2.81E-I-21
2.29E+21
1.77E+21
1.78E+21
1.69E-I-21
Misc
Nonmetallic
Minerals, Nee
(scj/y)
1.82E+20
1.68E-t20
1.80E-t20
1.99E+20
1.75E+20
2.27E+20
2.31E-t20
2.29E+20
6.31E+19
5.94E+20
6.21E-t20
Forest Products
Total
(sej/y)
1.222EH-20
L324E-I20
1.416E-I20
1.377EH-20
1.818E+20
1.737E+20
4.672E-1 19
2.7E+19
2.749EH-19
5397E+19
5.682E-1 19
Primary Forest
Products, Logs
(sej/y)
1.11439E+19
1.20767EM9
1.29163EM9
1.31497E+19
2.31487E+19
2.12302E+19
8.4775EM8
4.98659E+18
5.28069E-18
1.21393E-19
1.29705E-19
Lumber and
dimension stock
(scj/y)
5.434E+19
5.89E-M9
6.299EM9
5.826E+19
6.752E+19
7.06E+19
3.478EM8
3.397E+18
2.06E+18
9.845E+18
1.045EM9
Wood Products
(scj/y)
5.6669E-19
6.1418E-19
6.5687E-19
6.6308E-19
9.1163E+19
8.1919E-19
3.4765E-19
1.8621E-19
2.0149E-19
3.1991E-19
3.3402E-19
-------
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
All other materials
(sej/y)
1.119E+21
1.209E+21
1.294E+21
1.326E+21
8.571 E+20
7.651E+20
4.247E+20
4.464EI-20
6.514E+20
5.184E+21
5.598EI-21
Services in
Agricultural
Products
(s^j/y)
2.296E+20
2.253E+20
2.509E+20
2.757E+20
3.034E+20
2. 661 E+20
2.85E+20
6.722E-i20
6.598E+20
5.802E+20
5.682E-i20
Services in
Minerals
(sej/y)
2.344E-20
2.17E-20
2.325E-20
2.342E420
1.566E420
1.828E+20
1.925E+20
1.456E-I-20
9.502E-H9
2.356E420
2.3 19E -1-20
Services in Forest
Products
(sej/y)
6.7331 Ef 20
7.2813E+20
7.6716E+20
6.9573EH-20
7.9454EH-20
8.192E+20
9.5246E+19
6.1747E-M9
5.1814E+19
1.4459E+20
1.4389E-I-20
Services in All
Other materials
(sej/y)
9.35722E+20
1.01145E+21
1.06632E+21
1.13029E+21
7.68159E+20
6.81173E+20
6.33211E+20
6.06636E+20
1.06415E+21
6.22359E-I-21
6.29179E-I-21
Emigration by age
(sej/y)
0
7.032E+19
0
0
0
0
0
0
4.9303EM9
6.7808E+19
1.282 IE +20
6.4372E+19
4.1769E+19
3.7275E+19
6.0803E+19
6.2125E+19
2.8154E+19
6.6619EM9
4.9303E+19
0
0
-------
•Total Renewable Production
Agricultural Crop Production
Total Fuels and Electricity
3.0E+21
O.OE+00
1980 1985 1990 1995 2000 2005
Figure 5-E.l - Renewable production compared to the emergy of fuels and electricity, and the renewable emergy base
oftheSLB.
-Coal
Petroleum
-Services in Fuels
-Immigration
Federal and State Outlays
-•-Natural Gas
Electricity
-•-Services in Electricity
^~ Tourism
O.OE+00 -i
1980
1985
1990
1995
2000
2005
Figure 5-E.2 - Minor components of the emergy input to the San Luis Basin 1980-2005.
-------
Appendix 5-F: Quality of the
Global Insight. Inc. (GI) data were important in that
they allowed us to apply the revised method (Campbell
et al. 2005) for calculating the import and export of
emergy to and from a region. The commodity flow
data collected by the Bureau of Transportation Statistics
(1993, 1997, 2002) met the data quality objectives for
measuring commodity movements for the studies of
West Virginia and Minnesota (Campbell et al. 2005.
Campbell and Ohrt 2009). GI uses their proprietary
program, TRANSEARCH, to estimate domestic freight
flows by count}', commodity, and mode of transportation.
They used a variety of public data sources to derive these
flows (Global Insight 2006). To verify the accuracy
of TRANSEARCH data GI benchmarks the flows"
against reported freight volume data using two primary
sources: 1) private carrier information that they acquire
as part of a data exchange program with railroads and
truck carriers; and 2) truck count information released
by the state departments of transportation. However,
GI does not perform any statistical quantification of
this process to estimate the uncertainty inherent in
the TRANSEARCH data (Rich Fullenbaum, personal
communication). Even though no statistical estimates
were made to ascertain this uncertainty, we believe
that it is reasonable to assume their methods can)'
uncertainty comparable to that in the US Census Bureau
estimates, because both are based on similar sources.
Furthermore. GI crosschecks their estimates with
independent information (Rich Fullenbaum, personal
communication). Despite the fact the methods reported
by GI appear to be logical and deliberate, there is
considerable uncertainty associated with some of their
measurements. Principally, the export of "nonmetallic
minerals, processed" appears to be underestimated.
based on conversations with officials of two perlite
companies. Both Dicaperl and Harborlite are doing
business in Antonito. CO, and stated that almost all of
the material that arrives at Antonito is shipped out by
rail. At this time we are unable to determine the reason
for this apparent error; however, this is a moot point
because this commodity class passed through the SLB
without local use and therefore we did not include it in
our system's import-export balance.
References
Bureau of Transportation Statistics. 1993. Commodity
Flow Survey. U.S. Department of Transportation,
Washington, DC.
Bureau of Transportation Statistics. 1997. Commodity
Flow Survey. U.S. Department of Transportation,
Washington, DC.
Bureau of Transportation Statistics. 2002. Commodity
Flow Survey. U.S. Department of Transportation,
Washington, DC.
Campbell, D. E., Brandt-Williams, S., Meisch M. 2005.
Environmental accounting using emergy: evaluation of
the state of West Virginia. EPA600/R-05/006. United
States Environmental Protection Agency, Narragansett,
Rhode Island.
Campbell, D.E., Ohrt. A. 2009. Environmental
accounting using emergy: evaluation of Minnesota.
USEPA Project Report, EPA/600/R-09/002, 138 pp.
Global Insight, Inc. 2009. Development of the
TRANSEARCH Database. Global Insight, Lexington,
Massachusetts.
-------
Appendix 5-G:
A thorough emergy analysis often consists of several
analyses and corresponding results. In an effort to be
concise, many of the aspects related to the emergy of
this study were not included in the body of the report.
Below are the reference notes and tables that contain
and the methods and assumptions used for calculating
the energy and mass variables used in the emergy
evaluation of the SLB. Various figures represent the
reference notes and additional aspects of the San Luis
Basin. This information makes available the core data
needed to calculate the emergy flows of the region for
1995. Information for other years sufficient to reproduce
the calculations for that year using the online spreadsheet
for 1995 is available at http://www.epa.gov/aed/research/
desupp5.html.
-------
Table 5-G. 1 - Example calculation of emergy and economic value, assumptions, and data sources used in the emergy
evaluation of the San Luis Basin and Upper Rio Grande River Basin for the year 1995.
Note
1
Note
Variables
Areas
Alamosa
Conejos
Costilla
Hinsdale
Mineral
Rio Grande
Saguache
Area of 7 counties
Area counties not in watershed
Area counties in region
Area counties in mountains
Area of Upper Rio Grande watershed
Area of valley Floor
Area of mountains
Closed Basin watershed (see Figure 5-G.5 ')
Rio Grande watershed excluding the closed
basin (see Figure ')
Values
(mi2)
722.74
1287.22
1227.1
1117.68
875.72
911.6
3168.44
9310.5
7900
3200
4700
Values
(in2)
1.87E+09
3.33E+09
3.18E+09
2.89E+09
2.27E+09
2.36E+09
8.21E+09
2.41E+10
3.65E+09
8.29E+09
1.58E+10
2.05E+10
8.29E+09
1.22E+10
8.05E+09
1.15E+10
Sources
http://quickfacts.census.gov/qfd/states/08/08003.html
http://quickfacts.census.gov/qfd/states/08/08021.html
http://quickfacts.census.gov/qfd/states/08/08023.html
http://quickfacts.census.gov/qfd/states/08/08053.html
http://quickfacts.census.gov/qfd/states/08/08079.html
http://quickfacts.census.gov/qfd/states/08/08105.html
http : //quickfacts .census . go v/qfd/states/0 8/081 09 .html
Sum of 7 counties.
By difference between counties
The region consists of 7 counties.
All area outside the watershed is mountainous.
http://www.nps.gov/archive/grsa/resources/docs/Trip2023.pdf
http://www.nps.gov/archive/grsa/resources/docs/Trip2023.pdf
http://www.nps.gov/archive/grsa/resources/docs/Trip2023.pdf
This study, GIS estimate
This study, GIS estimate
Variable and formulae
Subsets of the
Variable
Value
Units
Sources and additional
information
Renewable Energy Sources
2
Solar Energy
Solar radiation absorbed was
determined from weather data for two
stations from the National Renewable
Energy Laboratory (NREL)
Dataset, Alamosa in the valley and
Wolf creek in the mountains.
Average precipitation and albedo were
prorated by area.
Received, watershed
Absorbed, watershed
Received, 7
counties
Absorbed, 7
counties
Solar insolation for
watershed
Avg. Insolation
(Alamosa)
Solar for watershed
Solar Energy
Received
Albedo
Solar Energy
Absorbed
Solar Insolation for
Counties
Solar Energy
Received
Albedo
Solar Energy
Absorbed
1.2378E+20
9.3062E+19
1.46E+20
1.10E+20
1.68E+07
1.2378E+20
3.07181E+19
9.3062E+19
1.45697E+20
3.60751E+19
1.09622E+20
Jy4
Jy4
Jy-1
Jy4
J m-2 dj
Jy4
Jy4
Jy4
Jy4
Jy4
Jy4
NREL Data (1 station, n=46)
http://rredc.nrel.gov/solar/old data/
nsrdb/1 99 1 -2005/siteonthefly.
cgi?id=724620
Area weighted adjustment by solar
fraction for mountains.
-------
Note
3
4
5
Variable and formulae
Solar Insolation (Wolf Creek/ Alamosa)
Kinetic Energy of Wind
Used at the Surface
(density )(drag coefficient)(geostrophic
velocity )3(area)(seconds/year)
Calculated in Odum (1999)
"Evaluating Landscape Use of Wind
Kinetic Energy."
Earth Cycle Energy
(land area)(heat flow/area)
Rain, Chemical Potential
(area)(rainfall)(density water)(Gibbs
Free Energy water relative to seawater)
The chemical potential of rain
was determined by a GIS analysis
performed for this study.
Rainfall was averaged for each county
and the counties were summed to get
the total for the SLB.
Detailed data on the chemical potential
of rainfall is found in the Geo and
Chemical Potential spread sheet '
Subsets of the
Variable
Albedo
Alamosa
Wolf Creek
Wind Speed, @ 10 m
Air Density
Geostrophic Velocity
Drag Coefficient
(area weighted)
Area, 7 counties
sec / year
Drag coefficient flat
land
Drag coefficient
mountains
Area
Heat flow/area
Heat flow per mVyear
Area
Rainfall
Gibbs Energy
Density
Value
0.979775281
Fraction
reflected
0.25417
0.24417
1.89E+17
2.5
1.3
4.166666667
0.002656302
24114195000
31400000
0.002
0.003
6.81E+16
24114195000
89.53571429
2823598.286
5.83E+16
24114195000
variable by
county
4.74
1000000
Units
dimen-sionless
dimen-sionless
dimen-sionless
Jy1
m s4
kgm-3
m s4
m2
Jy1
m2
mWm"2
J m-2 y-1
Jy1
m2
my4
Jg1
gm"3
Sources and additional
information
These are dimensionless numbers
http://eosweb.larc.nasa.gov/sse/
http://www4.ncdc.noaa.gov/cgi-
win/wwcgi.dll?WWDI~StnSrch
http://smu.edu/geothermal/georesou/
resource.htm
Heatflow is the same for all years.
-------
Note
6
7
Variable and formulae
Rain, Geopotential
Received on the Land
Received: (area)(mean elevation)
(precipitation)(density)(gravity)
We estimated that 1/2 the precipitation
falls as rain over a year.
Rain, Geopotential Energy
Absorbed from Runoff
Rain Geopotential used, Rain Runoff
Absorbed: (area)(mean elevation-
minimum elevation)(runoff)(density)
(gravity)
The geopotential energy absorbed from
the rain runoff was determined from
GIS calculations made for this study.
The closed basin and the Rio Grande
drainage were handled separately,
because the low points, average
elevations, and rainfall of the two
watersheds were not the same (see
Geochemical Potential Table online).
http ://www. epa.gov/aed/research/
desupp5.html
Subsets of the
Variable
1.76E+17
Total Area
elevation
(State location of
station)
rain/yr
Density
Gravity
9.04E+15
Area
Closed Basin
Rio Grande Drainage
Average Elevation
Closed Basin
Rio Grande Drainage
Minimum Elevation
Closed Basin
Rio Grande Drainage
Rainfall
Closed Basin
Rio Grande Drainage
Density
Gravity
Value
J/yr
24114195000
variable by
county
variable by
county
1000
9.81
J/yr
8047770000
11502440000
2679
2836
2289
2252
0.377
0.553
1000
9.81
Units
Jy1
m2
m
my-1
kgm-3
m s"2
Jy1
m2
m2
m
m
m
m
my4
my-1
kgm"3
m s"2
Sources and additional
information
-------
Note
8
9
Variable and formulae
Snow, Geopotential
Received on the Land
Received: (area)(mean elevation)
(precipitation)(density)(gravity)
We estimated that 1/2 the precipitation
falls as rain over a year.
Snow, Geopotential Energy
Absorbed from Runoff
Snow Geopotential used, Snow Runoff
Absorbed= (area)*(mean elevation-
minimum elevation)*(runoff)*
(density)*(gravity)
The geopotential energy absorbed from
the snow runoff was determined from
GIS calculations made for this study.
The closed basin and the Rio Grande
drainage were handled separately,
because the low points, averages
elevation, and rainfall of the two
watersheds were not the same (see Geo
chemical Potential Table online).
Subsets of the
Variable
1.76E+17
Total Area
elevation
(State location of
station)
rain/yr
Density
Gravity
2.04E+16
Area
Closed Basin
Rio Grande Drainage
Average Elevation
Closed Basin
Rio Grande Drainage
Minimum Elevation
Closed Basin
Rio Grande Drainage
Rainfall
Closed Basin
Rio Grande Drainage
Density
Gravity
Value
J/yr
24114195000
variable by
county
variable by
county
1000
9.81
J/yr
8047770000
11502440000
2679
2836
2289
2252
0.377
0.553
1000
9.81
Units
Jy1
m2
m
my4
kgm"3
m s"2
Jy1
m2
m2
m
m
m
m
my4
my4
kgm-3
m s'2
Sources and additional
information
-------
Note
10
11
Variable and formulae
Subsets of the
Variable
Value
Units
Sources and additional
information
Assumptions for snow
Assume all snow runs off into closed
basin except for 8% sublimation
(Campbell and Ohrt 2009)
Assume all snow runs off into Rio
Grande Drainage except for 8%
sublimation and 10% infiltration
(calculated from Emery 1996).
Chemical and Geopotential of Rivers
entering form Outside
Colorado is at the topographic high
point for the Nation; therefore, all
rivers flow out of Colorado to the east
or to the West.
Chemical Potential Energy of
Evapotranspiration
(Area)(Evapotranspiration)(density)
(Gibbs Free Energy per gram)
3.34E+16
Forest land Area
Avg. Forest
Transpiration
Density water
Gibb's Free Energy,
Rainwater
ET Forests
Arable land Area
Avg. Crop
Transpiration
Density water
Gibb's Free Energy,
Rainwater
ET Crops
Pasture land Area
Avg. Pasture
Transpiration
Density water
Gibb's Free Energy,
Rainwater
ET Pasture
J/yr
8820852711
0.30
1000000
4.74
1.23698E+16
2379884072
0.69
1000000
4.74
7.82877E+15
3818083936
0.73
1000000
4.74
1.31932E+16
Jy1
m2
my-1
gm"3
Jg4
Jy4
m2
my4
gm'3
Jg1
Jy-1
m2
my-1
gm'3
Jg1
Jy4
An average of 4 values for the ET of
three species found in the area was
used to
Estimated evapotranspiration.
Lodgepole pine (Knight et al. 1981),
large and small, Ponderosa pine
(Ryan et al. 2000) and cottonwood
(Gazal et al. 2006) were used.
See the online table for Forests, ET,
and GW 1
Average transpiration for crops was
taken from (Campbell et al. 2005)
Average transpiration for pasture
was taken from (Campbell et al.
2005)
-------
Note
Variable and formulae
Subsets of the
Variable
Value
Units
Sources and additional
information
Renewable Production
12
13
14
15
16
Nonrenewable Use
Coal Used
Assume that all coal imported is
consumed.
Corrected for coal used to generate
electricity.
Total Agricultural
Crops
(amount sold)
(energy/unit)
Hay
Barley
Oats
Potatoes
Wheat
Livestock
(annual production
mass)(energy/mass)
Cows
#sold
wt
Energy /unit
Fish Production
(mass)(energy/mass)
Mass
Energy/mass
Timber Harvest
(vol. forest harvested)
(drywt)(J/g)
Volume harvested
dry wt
Forest mass
J/g
2.10E+14
2.11E+12
Mass
8.44E+11
6.01E+10
3.41E+10
1.08097E+12
8.89E+10
1.2769E+15
190249
551065.2
12180
3.7158E+11
190,000
453.59
4311.58
9.7388E+14
3582538.96
1.01446E+11
0.5
50723103070
19200
J/yr
8572
907200
27000
gy1
gy4
gy4
gy4
gy4
gy4
Jy1
g/animal
Jg4
Jy1
Ibs y '
gib1
Jg4
Jy1
ft3
cm3
gem'3
g/yr
Jy1
Short tons/yr
g/short ton
Jg4
Data on crop production is from
Colorado Agricultural Statistics
(2006).
(Animal production data from
Colorado Agricultural Statistics
(2006).
Energy content from US Dept of
Agriculture Handbook
Fish production was estimated by a
survey of 3 local producers
Carried out by Fred Bunch of the
National Parks Service.
Estimate made in the Ecological
Footprint section of this study,
Using harvest records of the Rio
Grande National Forest and other
federal lands.
-------
Note
17
18
19
Variable and formulae
Natural Gas Used
Assume that all natural gas imported is
consumed.
Corrected for natural gas used to
generate electricity.
Petroleum Used
Assume that all petroleum imported is
consumed.
Electricity Used
Subsets of the
Variable
3.22E+15
Amount
4.43E+15
1.46E+15
Amount
Value
J/yr
3055602.64
1055000000
J/yr
811314.00
5.46E+09
J/yr
404174975.10
Units
Jy1
Thousand ft3
J/thousand ft3
Jy1
barrels
J/barrel
Jy1
kW-hr
Sources and additional
information
http ://bioenergy. ornl. gov/papers/
misc/energy conv.html
Renewables Used in a Nonrenewable Manner
20
21
Groundwater use in some years
(vol.)(density)(Gibbs free energy)
Soil Erosion
(area)(erosion rate)(organic fraction)
(energy)
2.57E+15
Volume used
Density
Gibbs
1.70E+15
Soil mass eroded
Organic fraction
Energy per gram
J/yr
5.49E+08
1000000
4.9
J/yr
2.71E+12
0.03
20930
•Jy1
mV
gm3
Jg4
Jy1
gy4
Jg1
Based on the volume of ground
water withdrawn as estimated in the
GNRP section.
Appendix 4-A
Nonrenewable Production
Nonrenewable production is based on
the excess of imports over
Exports for those nonrenewables
known to be produced in the SLB.
Crude Oil Produced
Metallic Ores
Produced
Broken Stone, Rip
Rap Produced
Sand and Gravel
Produced
Non-metallic
Minerals Produced
1.35E+11
8.40E+08
2.30E+12
2.52E+12
9.26E+10
Jy1
gy4
gy4
gy4
gy4
See Imports-exports worksheet1 for
calculation of the nonrenewable
Materials produced based on the
import-export balance.
Assume 10% of crushed stone
remains within the SLB.
Assume 10% of sand and gravel
remains within the SLB.
Imports
22
Fuels
All fuels and electricity used in the
SLB were imported during the time of
this study
Coal
Petroleum
Natural Gas
Electricity
2.10E+14
4.43E+15
3.22E+15
1.46E+15
Jy1
Jy1
Jy1
Jy1
-------
Note
23
24
25
Variable and formulae
Other goods (without nonmetallic
minerals processed) and services
People
Government
Subsets of the
Variable
Materials other than
fuels
Services in
materials other than
fuels
Services in Fuels
Services in
Electricity
Net Immigration
Tourism
Federal and State
Outlays
Value
2.87E+11
2.41E+08
4.56E+07
2.48E+07
0
5.53E+07
2.35E+07
Units
gy4
$y4
$y4
$y4
individuals
y4
$y4
$y4
Sources and additional
information
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Energy Consumption worksheet
i
See the Population worksheet '
Dean Runyon Associates, Colorado
Travel Impacts (1996-2003),
Colorado Tourism Office
See Government Worksheet '
Exports
26
Exports by Category
(without nonmetallic minerals
processed)
Agricultural Crops
Potatoes
Grain, Wheat,
Barley, Oats
Vegetables, Fruit,
Nuts, Seeds
Horticultural
Specialties
Livestock
Minerals
Metallic Ores
Crude Petroleum
Broken Stone Or
Riprap
Gravel Or Sand
Misc Nonmetallic
Minerals, Nee
Forest Products
Primary Forest
Products, Logs
Lumber and
dimension stock
Wood Products
All other materials
1.08E+12
1.83E+11
2.18E+09
7.02E+09
1.28E+15
8.40E+08
3.10E+06
2.30E+12
2.52E+12
9.26E+10
2.96E+10
1.05E+11
2.82E+10
8.21E+10
gy4
gy4
gy1
gy4
Jy4
gy4
gy1
gy4
gy4
gy4
gy4
gy4
gy4
gy4
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
see Livestock worksheet
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
See Imports-Exports worksheet '
-------
Note
27
28
29
Variable and formulae
Services
Services in Agricultural Products
Services in Minerals
Services in Forest Products
Services in All Other materials
(without nonmetallic minerals
processed)
Services
Gross Regional Product
People
Emigration by age
Government
Federal Taxes
State Taxes
Subsets of the
Variable
Value
8.83E+07
9.01E+07
2.59E+08
3.60E+08
N.A.
9.70E+02
Units
$y4
$y4
$y4
$y4
Sy-1
$y4
individuals y1
Sy-1
$y4
Sources and additional
information
See Imports-Exports worksheet 1
See Imports-Exports worksheet 1
See Imports-Exports worksheet l
See Imports-Exports worksheet l
We have not yet calculated exported
services.
Used data from Chapter 4 and
estimated for this Chapter10
See Population worksheet1
Data on federal and state taxes are
only available for 2004.
http ://www.taxfoundation. org/
files/fedtaxburden
bycounty-2007032 1 .pdf
1 http://www.epa.gov/aed/research/desupp5.html
References
Gazal, R.M., Scott, R.L., Goodrich, D.C., Williams,
D.G. 2006. Controls on transpiration in a semiarid
riparian cottonwood forest. Agricultural and Forest
Meteorology 137, 56-67.
Knight, D.H., Fahey, T.J., Running, S.W., Harrison,
A.T., Wallace, L.L. 1981. Transpiration from 100-yr-
old lodgepole pine forests estimated with whole-tree
potometers. Ecology 62, 717-726.
National Agricultural Statistics Service. 2006. Colorado
Agricultural Statistics, http://www.nass.usda.gov/
Statistics_by_State/Colorado/Publications/Annual_
Statistical_Bulletin/2006bulletin.pdf.
Ryan, M.G., Bond, B.J., Law, B.E., Hubbard, R.M.,
Woodruff, D., Cienciala, E., Kucera, J. 2000.
Transpiration and whole-tree conductance in
Ponderosa pine of different heights. Oecologia 124,
553-560
10 Chapter 4 did not specifically estimate Gross Regional Product (GRP). To estimate GRP for this chapter, we used Gross Domestic Product (GDP)
for Colorado and New Mexico and population for the two states. We multiplied population for the SLB by the average per capita GDP from both
states to estimate GRP. This approach requires numerous assumptions. However, the main assumption depends on whether the average per capita
GDP for both Colorado and New Mexico represents consumption in the SLB. If this assumption is not sound, we may have overestimated or under-
estimated GRP.
-------
•Imported Nonmetallic Minerals Processed -B-Exported Nonmetallic Minerals, Processed
3.5E+21
3.0E+21 •
>
'5T
o>
o
o.
E
UJ
O.OE+00
1995
1997
1999
2001
2003
2005
Figure 5-G.l - The emergy of processed non-metallic minerals (i.e., perlite and scoria) imported into the San Luis
Basin and the emergy exported in that category.
Immigration
Emigration
4.0E+20
5.0E+19
O.OE+00
1985
1990
1995
2000
2005
Figure 5-G.2 - The change in annual empower from net immigration or emigration from 1985 to 2005.
-------
•Electricity Used
•Fuels used
o>
42.
I
o
Q.
E
LJJ
6.0E+20
5.0E+20
4.0E+20
3.0E+20
2.0E+20 •
1.0E+20
O.OE+00
1980 1985 1990 1995 2000 2005
Figure 5-G.3 - The emergy of fuels and electricity used per year in the San Luis Basin from 1980 to 2005.
•Outflows - Inflows
o>
(A
0)
O
Q.
E
LU
1.2E+22
O.OE+00
-2.0E+21
1994
1996 1998 2000 2002
2004 2006
Figure 5-G.4 - The difference between the total emergy outflows (B+P E+N ) and total emergy inflows (R+F+G+P I).
-------
Figure 5-G.5 - Digital elevation model of the Upper Rio Grande River Basin showing the boundaries of the closed and
open basins.
-------
6.0
Fisher Information and Order
6.1 Introduction
This chapter presents the theory, data, and methodology
for using Fisher information (FI) to assess Hie dynamic
order and overall stability of the San Luis Basin (SLB)
regional system. In this chapter we provide background
on the method, details of the numerical approach, and
include a simple example demonstrating the computation
method. Further, we delineate the data used to compute
FI for the SLB. present results from the FI analysis and
discuss the strengths and weaknesses of the approach.
Fisher information is a key method used in information
theory and provides a means of monitoring a broad range
of system variables to characterize the dynamic order
of a system, to include its regimes, and identify regime
shifts (Cabezas et al. 2003). Fisher information has not
traditionally been used as a metric of sustainability, and
its application to sustainability is, in fact, a relatively
new concept. However, its usefulness as a measure of
sustainability has been documented in the peer-reviewed
scientific literature (Cabezas and Path 2002, Cabezas et
al. 2003, Path et al. 2003, Karunanithi et al. 2008, Mayer
et al. 2007) over a wide array of applications ranging
from real and modeled ecosystems to climate.
We have used FI as a metric in the SLB project because
it captures a critical aspect of the system that the other
metrics do not. Well-functioning economic, ecological,
and social systems may be necessary to preserve the
self-organization and resilience of an environmental
system and FI captures these aspects. Conceptually, the
repeatability of observations denotes order in dynamic
systems (Karunanithi et al. 2008). Therefore, a dynamic
system is said to be well functioning and orderly when
its behavior over time follows characteristic, regular, and
nearly predictable patterns. FI captures organization and
resilience by assessing the dynamic behavior of systems.
To illustrate this point, we offer three examples.
First consider that biological systems have elaborate
mechanisms that allow them to maintain a specific
degree of organization (Kairffnian, 1993). The fact
that an ecosystem can be identified as such is itself
evidence that it has observable properties that, although
variable, are still sufficiently stable to define the
system. Indeed, the properties do vary, but they do so
in a characteristically orderly and almost predictable
manner. Further, as a system experiences a regime shift
(i.e., loss of characteristic properties), the variation of
key properties generally increase (Carpenter and Brock
2006) which causes a loss of order. Second, similar
arguments can be made for markets that are left to
operate freely. The observable properties of the markets
follow regular patterns, which indicate order. Movement
toward market equilibrium helps to predict how prices
will change (Maurice and Phillips 1992). The existence
of patterns in the markets (i.e., prices as rationing and
signaling devices) allows consumers and producers to
make decisions with the expectation of particular results
(Mankiw 2009; Maurice and Phillips 1992). Finally,
consider that the four seasons of the year constitute an
orderly dynamic system because the seasons succeed
each other in regular, although not exactly precise,
timing year after year (Houghton 2002). Indeed, autumn
may sometimes come late and climatic fluctuations can
occur; however, the seasons follow a pattern.
In summary, ecosystems and social systems follow
very complex but real patterns of change over time.
Many processes in ecosystems cycle and change along
relatively orderly patterns often driven by the daily
and yearly cycles of light, tides and seasons along with
longer term decadal, centurial, and other temporal and
spatial variations. These cycles, although complex,
are sufficiently regular to be studied and understood
(Odum 1971). Markets likewise follow complex but
orderly patterns and they are sufficiently regular to be
studied and understood (Maurice and Phillips 1992).
In brief, many systems, when functioning well, show
regular characteristic behavior, which is an indication of
dynamic order. Accordingly, order, or more specifically
dynamic order, is an important and very fundamental
indicator of the state or condition of the system (Mayer
et al. 2007). It is for this reason that FI has been
incorporated into the suite of metrics used to assess
sustainability in the SLB project.
6.1.1 Capturing Dynamic Shifts with
Fisher Information
When any system, including the aforementioned ones,
undergoes a change from one characteristic pattern
or set of behaviors to another, the change is generally
termed a regime shift. Because no two regimes have
the same observable patterns, a regime shift is typically
accompanied by corresponding changes in dynamic
order, which can be tracked by FI. To illustrate this
point, consider the very simple case where a system is
represented by one variable (e.g., temperature) which
is being measured over time. Keeping in mind that the
-------
system here is represented by its temperature only, if
the temperature is the same (i.e., within measurement
error) every time it is measured, then the system can be
said to exist in one state only. This system has as much
order as it can possibly have, and the FI associated with
the temperature is at a maximum. On the other hand, if
the temperature is different every time a measurement
is made, even after accounting for measurement error,
say the temperature is rising linearly with time, then
the system can be said to exist in many different states
and the FI is at a minimum. Hence, a system with a
well-defined pattern (i.e., constant properties) has high
dynamic order and high FI, and a system that has no
defined patterns (i.e., constantly changing properties)
has low order and low FI. Although these two examples
represent extreme behavior, real systems typically
function between these extremes.
Further exploring this idea, consider a situation where
the aforementioned system is cooled, allowed to remain
at a constant low temperature, within measurement
error, for a period of time, and then heated to a constant
temperature with larger variability, perhaps very small
amounts of heat are added or removed randomly. In this
case, when the temperature is decreasing, the FI will
be very small or zero. Once it reaches a constant low
temperature, the FI will be very high. As the system is
heated and the temperature increases, the FI will again
be very small or zero (during the change), and then
increases to a relatively steady FI. However, the final
"steady" FI value will be less than the original due to
the higher variability in the temperature. The system
described here experienced a dynamic regime shift,
as there was a noted drop in FI between two stable
regimes, one with higher order than the other. From
these explorations, it is evident that FI tracks the order
and stability of the system. Although we used simple
examples to illustrate the very basic relationship between
system dynamics and FI, as discussed later, the FI
method has been adapted to assess the dynamic changes
in complex systems described by multiple and disparate
variables.
6.2 Methods
6.2.7 Theory
Fisher information was formally developed by
statistician Ronald Fisher as the information obtainable
from data when estimating the value of a parameter (Path
et al. 2003, Fisher 1922). It has since been adapted as
a measure of dynamic order. Details on the derivation
of FI from its original form are provided in Path et al.
(2003), Mayer et al. (2007), Karunanithi et al. (2008),
and Appendix 6-A. 1. From the derivation, we obtain a
representation of FI (denoted here as / to preserve the
historical context; Fisher 1922, Frieden, 1998, 2004,
Karunanithi et al. 2008, Mayer et al. 2007) based upon
the probability of a system p(s) being in a particular state
(s):
e ds fdp(s)~\
JXs)L ds \ (6.1)
From this expression, an approach was developed that
affords the ability to assess the dynamic order of real
systems and is derived as follows. In order to minimize
calculation errors from very small probability values, we
replace p(s) in Equation 6.1 with its amplitude, which is
defined, by q2(s) = p(s), such that:
.
ds \ ds
,
as
(6.2a)
Substituting Equation 6.2a into Equation 6.1, the
expression becomes:
/ = .
ds
ds
(6.2b)
Equation 6.2b is adapted for use with discrete data by
using a summation to approximate the integral and
replacing ds and dq(s) with AS=SI-SI+I and Aq=qi-qi+1 such
that: r- -,2
(6.2c)
In Equation 6.2c, m is the number of states and s; is
merely an index denoting a particular state of the system
(e.g., Sj is state 1, s2 is state 2, etc.). Accordingly, s;- si+1
= 1 and Fisher information is then:
Equation 6.3 is our working expression for computing
the FI metric.
6.2.2 Calculation Methodology
The basis of computing FI is assessing changes in the
probability of observing different states of the system
through time. Therefore, we must obtain information
on its condition (state) over time. Borrowing from
standard statistical mechanics approaches, we define a
system by n measurable variables (x) that characterize
the system and its state at any point in time (Mayer
et al. 2007). Note that the correlation structure of the
variable time series is not critical, because our goal is
to assess changes in dynamic order and not to develop
a predictive model of system behavior. By describing a
dynamic system in this way, it is said to have a trajectory
-------
in a phase space representing all possible states of the
system in n-dimensions and time. In the absence of
measurement error, each point in phase space is defined
by specifying a value for each of the variables at a
point in time, pt.: (x^t), x2(t.) x3(t.)... xn (t.)). However,
because measurements of real system parameters
contain error (i.e., uncertainty), the state s of the system
is delineated by a region rather than a point. In other
words, two points that differ from each other by less
than the measurement error are indistinguishable and
can be thought of as two measurements within the same
state. Conversely, two points that differ from each other
by more than the measurement error are, in fact, two
legitimately different points and consequently, exist in
two distinct states of the system. From this conceptual
description, we have a foundation for understanding
the process of characterizing a dynamic system and
assessing changes in its state over time.
In order to capture the dynamic behavior of a system,
once the variables characterizing its state have been
gathered, the time period is divided into time windows.
We achieve this by first defining a parameter (hwin)
that denotes the size (in time steps) of each window and
then determining the amount of overlap (in time steps)
desired for each window (winspace). The parameters
hwin and winspace denote the integration window
parameters used to move through the time series data by
creating a sequence of overlapping windows such that
hwin > winspace as shown in Fig. 6.1. This convention
affords the ability to compensate for changes in
dynamic behavior that may extend beyond the boundary
of each window. The probability densities p(s) and
corresponding FI are computed within each of these
windows. Whereas hwin is selected based upon the
amount of data available, we have empirically found
that it should be at least eight time steps. The next step
in the procedure is "binning" each point within the time
windows into discrete states of the system. Estimating
the uncertainty of the variables is key to defining the
states of the system.
As previously mentioned, inherent in any measurement
is a level of uncertainty. Accordingly, we define a
parameter Ax. as the measurement uncertainty for each
variable (x) such that, if the condition:
**' (6.4)
is true for all variables at time t. and t then the two points
are indistinguishable and subsequently, grouped in the
same state (i.e., binned together). Because multiple
variables are used to characterize a system and each
variable has a distinct measure of uncertainty, a state is
represented as an n-dimensional hyper-rectangle where
each side is defined by an uncertainty (Ax.) for each
variable. Therefore, this Axis the size of the states for
the system.
Typically, knowledge of the measurement uncertainty
of underlying state variables is unknown; therefore,
we have developed approaches for estimating Ax.
One tactic is to find a relatively stable time period
within the system trajectory, calculate the variation in
each variable in this period, and assume this to be the
measurement uncertainty. This approach is implemented
by calculating the standard deviation (SD) for each
variable in the "stable" period and applying Chebyshev's
inequality, which is defined by:
P(X-fi\95% at k=2) if, for instance, the data are
normally distributed. Given that each variable has a
unique measure of uncertainty, the size of states then is
noted as a row vector Ax = [&SDp kSD2 . . . .£SDJ, where
k is a scalar constant. Another method of defining the
size of states is by locating a similar system that exhibits
stability and using the variation within this system as a
measure of uncertainty for the system under study.
Once the integration window parameters (hwin and
winspace) and size of states (Ax) are determined, the data
may be distributed (binned) into different states of the
system. The binning process begins with the first point
within the time window taken as the center of the first
state. A hyper-rectangle with the side lengths defined
by Ax for each variable is established around the point,
such that all of the points falling within its boundaries
are considered to be in the same state. Then, the next
unbinned point in the window is assumed to be the center
of a new state, a new hyper-rectangle is constructed
around it, and all the points within its boundaries are
binned into that state. The process continues until all
of the points within the first window are binned and is
repeated for the remaining time windows until all of
the points in each window are binned into states of the
system and the data are exhausted.
-------
When measurement uncertainty of the underlying state
variables is known and independent, the binning process
alone would work well in defining states of the system.
Unfortunately, data collected from public sources,
as in this project, often do not report the uncertainty
associated with their data. Accordingly, an additional
step is implemented to mitigate the effect of data error
by applying a tightening level parameter (Karunanithi et
al. 2008). The tightening level (TL) adjusts the binning
criteria such that a point can be declared to be within
a given hyper-rectangle (particular state of the system)
when at least a certain percentage of the variables meet
the size of states criteria (Equation 6.4). The percentage
itself is the tightening level. For example, if a system is
characterized by 100 variables and 95 of the variables
indicate that a particular point fits within the state being
evaluated, then the two points would be binned together
at a 95% tightening level. Therefore, points may be
binned into a particular state of the system if the size of
states criteria is true for all variables in particular time
steps or the number of variables that satisfies the size of
states criteria is greater than the product of the tightening
level (TL) and the total number of variables (Fig. 6.2).
When all of the points in a window are binned into states
at a given tightening level, a probability distribution (pt)
is generated for each window using:
fluctuations. As such, Equation 6.7 is used to compute a
three-point mean, for window y:
P. = -
# points in state
total # points in time window (6.6)
Next the amplitude, q (^ = ^/p~) is calculated for each
state and FI for the time window is computed using
Equation 6.3. The computed FI value itself is assigned
to the middle of each time window (see example in
Fig. 6.1). Because there are no rigorous criteria for
setting the tightening level, FI is computed for multiple
tightening levels between strict and relaxed tightening.
Rather than establishing an arbitrary lower bound for
the tightening level, relaxed tightening is set as the
lowest tightening level at which more than one state is
observed in the window and FI is calculated by taking
the arithmetic average of FI between strict tightening
(TL=100%) and lower bound (relaxed tightening level).
Note that this results in multiple pt and FI values for each
window (Karunanithi et al. 2008). The final FI reported
in each window is again an average over the tightening
levels from strict to relaxed tightening. Finally, in
accordance with the Sustainable Regimes Hypothesis
(see section 6.2.3), a regime is denoted as sustainable
when the dynamic order does not change with time
(i.e., (d(Ff)ldt» 0) where indicates a mean FI
value). The is calculated by computing the mean
of neighboring FI values. This convention affords the
ability to focus on trends in dynamic order and not
j=w-l, w-3, w-5, w-7, ... (6.7)
where w is the year corresponding to the last window of
the original FI computation. Based on our experience,
the three-point mean tends to preserve trends and
deemphasize short fluctuations. The is evaluated. However, given
that a true measurement of uncertainty is unknown for
the metrics or underlying variables, we can visually
inspect the trends, but we cannot test for statistical
significance.
6.2.3 Interpreting Fisher Information
The Sustainable Regimes Hypothesis encompasses
conceptual ideas governing the use and interpretation
of FI as a metric for assessing sustainability (Cabezas
and Path 2002, Karunanithi et al. 2008). In summary,
the hypothesis states that: 1) well functioning systems
exist within an orderly dynamic regime with non-zero
FI that does not change with time (i.e., (d(Fl}ldt » 0);
2) steadily decreasing FI signifies a progressive loss of
dynamic order and denotes a system that is becoming
disorganized and losing functionality; 3) steadily
increasing FI indicates that the system that is becoming
more ordered (although, not necessarily more desirable
by humans); and 4) a steep decrease in FI between two
dynamic regimes denotes a regime shift (Karunanithi
et al. 2008). As a note, both conditions of statement 1
must be true for a system to be considered sustainable.
In other words, a completely disorganized system has no
order over time; therefore, a system with ~ 0 is not
sustainable even ifd(Fl}/dt « 0.
Important elements in regime shift detection are
determining the occurrence, pervasiveness, and intensity
of a shift (Karunanithi et al. 2008). The fourth statement
of the Sustainable Regimes Hypothesis indicates that a
regime shift has occurred when there is a significant drop
in FI between two dynamic regimes. Once a regime
shift has been detected, the intensity simply relates to
the level of the drop in FI (e.g., a more severe shift has
a steeper drop). The pervasiveness of the shift relates
to the number of system variables affected by the shift
and is characterized by varying the tightening level and
noting the lowest tightening level at which a particular
shift can be detected. Recall, that the tightening level
relates to the number of variables that must meet the
size of states criteria for binning. Accordingly, the more
relaxed (lower) the tightening level at which the regime
shift is recognized, the more pervasive the shift.
-------
6.2.4 Computing Fisher Information:
A Simple Example
Below a simple example is used to demonstrate the
procedure for computing FI. From Section 6.2.2, the
basic algorithm is as follows: (1) establish the size of the
time windows (hwin), (2) determine the time increment
that the window will be moved forward (winspace) to
create overlapping windows, (3) set a tightening level
(TL), (4) bin all of the points into states within each
window, (5) compute the probability density for each
state in each time window - the result at this point
will be a sequence of probability densities/>; for each
time window, (6) calculate FI from the q. in each time
window, (7) set a new tightening level, (8) repeat steps
4 through 7 until all the computations have been done
from strict tightening (TL = 100%) to relaxed tightening
(the lowest TL at which more than one state is present
in the window), (9) compute an average FI for each
window over the tightening levels from strict to relaxed
tightening and (10) calculate (mean FI) values for
the system.
For the sake of simplicity, we used two demographic
variables (population and personal income) from the
SLB data (Fig. 6.3). Because the goal of this exercise
is simply to step through the computation algorithm,
we used the guidelines for the integration window
parameters (see section 6.2.2) and selected values for
hwin and winspace that met the criteria. Recall that
the parameters should be set such that hwin > 8 and
winspace < hwin. Accordingly, we set hwin to eight
time steps (years) and winspace to three time steps
to ensure we ended up with a manageable number
of windows for this computation demonstration. To
provide insight into how the integration window
parameter settings may affect the computation result, we
performed a sensitivity analysis to examine the impact
changes in these parameters have on FI (Appendix 6-B).
However, the sensitivity analysis is not a requirement for
setting hwin and winspace. The only specific parameter
guidelines are provided in section 6.2.2.
Following the basic algorithm already described, we
used the hwin and winspace values to partition the data
into over lapping windows (Table 6.1). Next, we set the
tightening level (TL) at 100% to include all variables in
the computation and then binned the points into states
of system within each window (Fig. 6.2). The purpose
of binning is to determine the state of the system over
time and the basis of this process is grouping points
(e.g., pt1=(x1(t1), x2(tj)) within an established boundary
of uncertainty. As explained in section 6.2.2, if the level
of uncertainty for the variables under study is unknown,
it may be estimated. In this case, it was estimated by
assuming a measurement uncertainty by computing the
standard deviation (SD) of each variable over the first
five time steps. Each SD was then multiplied by two
in accordance with Chebyshev's theorem (Lapin 1975:
58), such that the level of uncertainty (Ax.) defined
as ikSD^ is a 1x2 vector: Ax. = [2xSDi; 2xSD2] =
[86285.39, 1639.83]. From Fig. 6.2 and Equation 6.4,
if the tightening level multiplied by the total number of
variables in the time series meet the size of states criteria
(i.e., \xt(t) - xt(t)\ < Ax) then the two points at t. and t. are
binned in the same state. The differences were computed
by subtracting the value of the variables at each time step
starting with the first point in the window (Table 6.2).
For example, the difference between variable Xj at time
= 1 and time = 2, i.e.,x1(t2)-x1(tj)\ = |314914-310352| =
4562.
A "bulls-eye" (radar) plot of the differences provides
a visual depiction of point binning (Fig. 6.4). The first
binning pass starts with assessing the distance from
ptF accordingly, ptj is the center of this figure with an
absolute difference value of (0, 0) and the remaining
absolute differences are plotted on the corresponding
axis (e.g., 2-1 = (\X](t2) - Xj(t)\, \x2(t2) - x^)|) = (4562,
511)). The red and blue lines represent the boundary
of uncertainty (i.e., size of states = AXj and Ax2) around
the center for each variable. Points 1, 2, and 3 are
binned into state 1, because pt2 and pt3 are within the
boundary of uncertainty (i.e., less than Ax. from ptj).
The process then moves to the next "un-binned" point
(pt4), establishes it as the center of state 2 and then
bounds state 2 as Ax. from pt4, such that points 4-8 are
binned into state 2. The binning procedure continues
to the next window until all points have been binned
into a state of the system. Once all the points have
been binned, then probability densities are computed
for each window using Equation 6.6. In window 1,
there were three points binned into state 1 and five in
state 2 resulting in a 37.5% chance (p(l)=3/8) that the
system is in state 1 and a 62.5% chance that it is in
state 2. This process was repeated for each window,
resulting in probability densities for each window (Fig.
6.5). Next, the amplitude (q(s) = ^p(s)) was computed
for each state (Table 6.3). To compute FI for each time
window, the gradients of the amplitude (q, -qM) were
used as in Equation 6.3 (see Fig. 6.6). At this point,
the tightening level may be decremented (i.e., TL =
TL-1) and the computation steps repeated as articulated
in section 6.2.2. This produces FI values for the time
series for TL ranging from strict to relaxed tightening.
An average FI result is calculated by taking the average
of all FI values computed within TL range. For this
exercise, we only computed FI at TL = 100% and then
-------
used Equation 6.7 to calculate a three-point to
smooth any fluctuations in the result and focus on the
general trend of the metric (Fig. 6.7). Further details
for using the code and graphical user interface (GUI) to
compute FI are provided in Appendices 6-C and 6-D.
The procedure for computing FI has been automated
inMATLAB (Release 2009a; Mathworks, Inc.) so that
only time series data that characterize the system, size
of states, hwin, and winspace need to be provided by the
user to perform the analysis. The code is provided in
Appendix 6-E.
6.3 Computing FI for the
To assess dynamic order in the SLB, we needed data that
characterized the state of the system. Accordingly, it was
necessary to gather data that represent environmental,
social, and economic aspects of the region. Because
the variables used to compute the other metrics satisfied
this criterion, data for this project were selected from
the datasets used to calculate EFA, GNRP and EmA
(Chapters 3, 4, and 5, respectively). However, any
data that adequately characterize a system may be
used to calculate FI. Further, we selected variables
that contained data for the entire 26 years of the study
(1980 - 2005). Each of the variables was assigned to
one of six categories for computing FI and encompassed
information on consumption (food and forest),
environment, and demographic characteristics, as well
as, the energy- consumption, land use, and agricultural
production aspects of the SLB (Table 6.4). Details on
each variable are provided in the corresponding chapters
(Chapters 3-5).
6.4 Calculation Methodology
Following the approach described in sections 6.2.2 and
6.2.4, as well as, the guidelines for computing FI using
MATLAB (Appendices 6-C and 6-D), the GUI (Main_
Fisher_data_proc.m) was used to compute both FI and
. The file containing data for the 53 variables
over 26 years, the time file which indicates the years
examined (i.e., 1980, 1981, 1982,... 2005), and the file
containing the size of slates were selected. Following
the guidelines in Section 6.2.2, the size of states for the
SLB was determined using a small number of years
where data were relatively stable. Thus, Hie size of states
was defined by calculating the SD of the first five data
points (years) for each variable and then (in accordance
with Chebyshev's inequality) multiplying the result by 2
(k = 2).
Given the amount of data and the results of the
sensitivity analysis (Appendix 6-B), the time window
was set to be eight time steps (hwin = 8) and the window
increment was set to one time step (winspace =1).
Thus, FI was integrated over an eight-year window
that was moved in one-year increments. The FI value
reported represents the FI computed over a specific
period and is reported in the center of that window (i.e.,
FI for 1984 = 1980-1987, FI for 1985 = 1981-1988,
FI for 1986 = 1982-1989, etc). The (for each
window,/) was calculated using Equation 6.7, such that
the . reported is a mean placed in the center of the
three points used to calculate it (e.g., = (FL00 +
FI,001 + FI2002)/3). Further, in order to explore drivers
potentially responsible for changes in dynamic order
of the system overall, we compared FI of the overall
system (included all 53 variables) to that of the variables
grouped by categories (e.g.. energy consumption).
Spearman's rank correlation coefficient analysis was
used to assess the relationship between the dynamic
order of the system and the variables by category. The
statistical significance level established was P < 0.05.
6.5
FI was computed for the SLB over the 26 years of
the study and ranged in value from 2.92 in the period
centering 2001 to 4.37 in the period centering 1996 (Fig.
6.8). ranged from 3.22 in the period centering
1986 to 3.89 in the period centering 1995 (Fig. 6.9; Table
6.5). The system increased initially, peaked in
the period centering 1995, and then decreased slightly
thereafter.
The minimum in the system in 1986 corresponded
(visually) to a minimum in the consumption (food
and forest) and agricultural production categories (Fig.
6.10). The peak in the system in 1995 corresponds
(visually) to a peak in the of the environmental and
energy categories, as well as, agricultural production
which increases up to 1992 and essentially remains
steady until 1998 (Table 6.5; Figs. 6.9 and 6.10).
Although die of consumption (food and forest)
category remains relatively steady, the of the other
categories exhibit larger changes over time.
Similar to the overall system, dynamic order of the
agricultural production category rose from 1989 to 1992,
was relatively steady until 1998, and decreased thereafter
(Fig. 6.10). The opposite was true of the demographic
variables as the peaked in 1989. decreased from
1989 to 1995 and increased thereafter. By the end of
the study period energy, environment, and demographic
categories exhibited an increasing trend (Table 6.5; Fig.
6.10). Further, Spearman's rank analysis indicated a
significant, negative correlation between the overall
system and die demographic category (r = -0.88, df= 4,
P = 0.033).
-------
6.6
According to criteria in the Sustainable Regimes
Hypothesis (Section 6.2.3). the results of this
assessment indicated that although there were changes
in over time, was relatively steady during the
period and there was no indication of an overall system
regime shift (i.e., no sharp drop between two regimes).
The dynamic order of the system increased up to 1995
and exhibited a small decrease thereafter. Although
the demographic, energy, and environmental categories
showed an increase in dynamic order at the end of the 26
years, the overall system, consumption (food and forest),
land use, and agricultural production categories indicated
a decreasing trend in which may denote some
movement away from sustainability (i.e., decreasing
dynamic order). There was, however, no sharp drop in
that could indicate a regime shift.
Spearman's rank analysis indicated that there is an
inverse relationship between the demographic category
and that of the overall system. Toward the goal of policy
setting and decision making, variables within a category
may be explored to uncover possible drivers of dynamic
behavior. As an illustration, a closer look at the FI of the
demographic category revealed that changes in dynamic
order seem to correspond to changes in the underlying
variables (e.g., population and personal income).
Because FI is a measure of system order, increases in
variability of the system variables generally result in
decreasing FI. As a simple exploration to see if we could
identify such variables, we computed the annual percent
change in the demographic variables as a measure
of variability. While population generally increased
during the period examined, the percent change varied
during the 26 years (e.g., +1.6% from 1982 to 1983 and
+0.9% from 1983 to 1984 graphically is a decrease).
Further, unlike the percent change in personal income
over time, the percent change in population decreased
initially, peaked in 1995, and decreased thereafter. A
visual comparison of these patterns and demographic
FI suggests, the variation in population (as described
by percent change) appears to be inversely proportional
to demographic FI (i.e., low variation in demographic
data corresponds to high FI values). For that reason,
changes in annual percent change of population seem
to correspond to changes in dynamic order of the
demographic category (Fig. 6.11). From this simplified
approach, there appears to be a relationship; however,
oilier techniques (e.g., sensitivity analysis) to explore the
effect of changes in underlying variables on FI may help
determine which variables drive changes in dynamic
order.
In summary, there was no indication of a regime
shift in the overall system. Although, the system
exhibited small changes in dynamic order during the
26 years, we conclude the SLB is relatively stable with
a slight indication of possible movement away from
sustainability near the end of the period.
6.7
One of the key strengths of FI is the ability' to collapse
data from complex, multivariate systems into a
fundamental metric that can be computed over time and
used to evaluate the dynamic behavior of systems. This
is important because the characterization of complex,
integrated systems for sustainability (social, ecological,
and economic systems) often requires a large number of
disparate variables. Traditional approaches to assessing
regime shifts and system dynamics (e.g., variance)
have typically only been demonstrated on simple model
systems and work must be done to determine whether
these methods can be used to evaluate real, complex
systems (Scheffer et al. 2009). However, the calculation
of dynamic order using FI is insightful, theoretically
sound, and provides a means of evaluating both model
and real systems that are characterized by multiple,
disparate variables.
As is the case for many methods used to analyze
complex systems, the computation of FI has its
challenges. Some of the challenges include establishing
the integration window parameters (i.e., hwin and
winspace), the need for determining the measurement
error for each variable (used to set the size of states), and
the availability and quality of data to characterize the
system. The integration window parameters are used
to traverse through the time series data and establish
overlapping windows for the FI computations. These
parameters must be carefully selected based on amount
of data available and knowledge of the system. The
size of states is a key parameter used for binning points
into states of the system and is based on estimating
the amount of error present in the data. Strategic
recommendations for determining the size of stales and
the integration window parameters were provided in
sections 6.2.2 and Appendix 6-B, respectively. Data
availability and quality issues are discussed in Chapter 7.
Another challenge is that because PDFs are used to
calculate FI, one FI value is provided in each time
window and reported in the center of that period.
Accordingly, there is a unique FI for each time window
and not each time step. Hence, FI values tend to come
several years behind the latest data point (e.g., the
last data were for 2005, but the last FI value was for
2001). Moreover, according to the Sustainable Regimes
-------
Hypothesis, in order to capture the trends (and not
fluctuations) in the dynamic order, a is calculated.
We therefore use a three-point to evaluate trends
in dynamic order over time. Based on our experience
with these calculations, three points seem to smooth the
result, while maintaining the characteristic changes in
dynamic order associated with trends.
Further, the methodology can identify if a system is
either maintaining its current order (i.e., (d(FI)ldt« 0)or
losing dynamic order (d(Ff) I dt <0) and heading toward
or already experiencing a regime shift (i.e., a sharp
drop in between two stable regimes). The loss of
dynamic order, and certainly a regime shift, indicates the
system is moving away from sustainability. However,
an increase in indicates the system is gaining order,
yet does not necessarily mean the system is moving
toward a more preferable state (i.e., sustainability) in
terms of human wants and needs (Karunanithi et al.
2008). Hence, FI is a one-sided test of sustainability and
care must be taken on interpreting changes in dynamic
order with reference to human preferences.
Lastly, we acknowledge that the details of the conceptual
underpinning of FI analysis can be quite abstract and,
perhaps, difficult for many potential users to understand.
However, we also believe that anyone with a good
grounding in science or engineering can understand
the fundamental concept sufficiently to apply the
method and interpret the results appropriately. We note
that practical and effective end users of methods and
technologies in many professions are not necessarily
experts in the theory behind the tool being used. For
example, most physicians are not experts in Quantum
Mechanics, yet they can effectively use Magnetic
Resonance Imaging (MRI) and interpret the results. Use
of MRI may involve a medical technologist who serves
as a bridge between those who designed the MRI and the
physicians that interpret and use the results. The point
here is that it should be possible to make the benefits of
otherwise abstract concepts (such as FI) available to non-
expert end users through software and/or individuals that
act as intermediaries. To help make the concept more
accessible, we provided an example of the computation
and interpretation of FI using a simple example of
dynamic order (Section 6.2.4). Further, we provided the
MATLAB code necessary to compute FI and stand-alone
software with a GUI to simplify calculating FI. The GUI
improves the usability of the metric by automating the
computation process.
6.8 References
Anderson, MJ. 2005. Trimming the FAT out of
Experimental Methods. OE (Optical Engineering)
Magazine, September 2005, p. 29.
Cabezas, H., Path, B.D. 2002. Towards a theory of
sustainable systems. Fluid Phase Equilibria 3-14,
194-197.
Cabezas, H., Pawlowski, C.W, Mayer, A.L., Hoagland,
N.T 2003. Sustainability: ecological, economic,
technological and systems perspectives. Clean
Technologies and Environmental Policy 5, 176-180.
Cabezas, H., Pawlowski, C.W, Mayer, A.L., Hoagland,
N.T. 2005. Simulated experiments with complex
sustainable systems: ecology and technology.
Resources, Conservation and Recycling 44, 279-291.
Carpenter, S.R., Brock, WA. 2006. Rising Variance: A
Leading Indicator of Ecological Transition. Ecology
Letters 9-3, 311-318.
Path, B., Cabezas, H., Pawlowski, C.W. 2003. Regime
changes in ecological systems: an information theory
approach. Journal of Theoretical Biology 222- 4, 517-
530.
Fisher, R.A. 1922. On the mathematical foundations of
theoretical statistics. Philosophical Transactions of the
Royal Society of London 222, 309-368.
Frieden, B.R. 1998. Physics from Fisher Information: A
Unification. Cambridge University Press, Cambridge,
UK.
Frieden, B. R. 2004. Science from Fisher Information:
A Unification Cambridge University Press,
Cambridge, UK.
Gatenby, R.A. (Eds), Exploratory Data Analysis Using
Fisher Information. London: Springer-Verlag, pp.
217-244.
Grove, J.M. 1988. The Little Ice Age. Routlege, New
York, New York, USA.
Houghton, J. 2002. The Physics of Atmospheres.
Cambridge University Press, Cambridge, UK.
Karunanithi A.T., Cabezas, H., Frieden, B.R.,
Pawlowski, C.W. 2008. Detection and assessment
of ecosystem regime shifts from Fisher information,
Ecology & Society 13(1), 22.
Kauffman, S.A. 1993. The Origins of Order: Serf-
Organization and Selection in Evolution. Oxford
University Press, Oxford, UK.
-------
Lapin, L. 1975. Statistics: Meaning and Method.
Harcourt Brace Jovanovich, New York, New York,
USA.
Mankiw, N.G. 2009. Principles of Economics (5lh Ed).
South-Western, Mason, Ohio, USA.
Mathworks, Inc. 2009. MATLAB Software: The
Language of Technical Computing version 7.8.0.347:
Release R2009a, Natick, MA.
Maurice, S.C., Phillips, O.R. 1992. Economic Analysis:
Theory and Application, 6lh Ed. Irwin, Homewood,
Illinois, USA.
Mayer, A.L., Pawlowski C.W., Cabezas, H. 2006.
Fisher Information and dynamic regime changes in
ecological systems. Ecological Modeling 195. 72-82.
Mayer, A.L., Pawlowski, C.W., Path, B.D., Cabezas,
H. 2007. Applications of Fisher information to the
management of sustainable environmental systems, in:
Frieden, B.R., and
Montgomery, D. C. 1997. Design and Analysis of
Experiments. John Wiley and Sons, Inc., pp. 4.
Odum, E.P 1971. Fundamentals of Ecology, 3rd Ed.
W.B. Saunders Company, Philadelphia, Pennsylvania,
USA.
Scheffer, M., Bascompte, I, Brock, W. A.. Brovkrn.
V., Carpenter, S. R., Dakos, V., Held, H., vanNes, E.
H., Rietkerk, M., Sugrhara. G. 2009. Early-warning
signals for critical transitions. Nature, 461, 53-59.
-------
Table 6.1- Data were divided into a sequence of overlapping time windows. The eighth window was removed because there were not enough data in the
time series to place in it. Therefore, all data were accounted for within seven windows, t = time (i.e., 1980-2005), xl = Population and x2 = Personal income
(in thousands). Point 1 (ptj) was defined by (x^), x2(tj)) and is highlighted as 310352, 38469.
ptj= (x^tj), x2(tl))= (310352, 38469)
#
1
2
3
4
5
6
7
8
Window 1 /
t
1980
1981
1982
1983
1984
1985
1986
1987
X' 1
310352 "
314914
318946
384256
400888
424601
431318
436775
X' /
38469 ^
38980
39467
40112
40490
40369
40804
41104
f
#
4
5
6
7
8
9
10
11
Window 2
t
1983
1984
1985
1986
1987
1988
1989
1990
x|
384256
400888
424601
431318
436775
448738
504014
551552
xz
40112
40490
40369
40804
41104
41289
40903
40682
#
7
8
9
10
11
12
13
14
Window 3
t
1986
1987
1988
1989
1990
1991
1992
1993
x|
431318
436775
448738
504014
551552
549641
562981
625715
xz
40804
41104
41289
40903
40682
41283
41120
41466
10
11
12
13
14
15
16
17
Window 4
t
1989
1990
1991
1992
1993
1994
1995
1996
x,
504014
551552
549641
562981
625715
661317
715547
754872
xz
40903
40682
41283
41120
41466
42564
43793
44566
13
14
15
16
17
18
19
20
Window 5
t
1992
1993
1994
1995
1996
1997
1998
1999
x|
562981
625715
661317
715547
754872
781542
839246
892574
xz
41120
41466
42564
43793
44566
45289
45902
46377
16
17
18
19
20
21
22
23
Window 6
t
1995
1996
1997
1998
1999
2000
2001
2002
x,
715547
754872
781542
839246
892574
910329
967522
1042786
xz
43793
44566
45289
45902
46377
47097
46907
47404
19
20
21
22
23
24
25
26
Window 7
t
1998
1999
2000
2001
2002
2003
2004
2005
x,
839246
892574
910329
967522
1042786
1042145
1048231
1097044
xz
45902
46377
47097
46907
47404
47598
48207
48101
V
\.
22 "%
23
24
25
26
0
0
0
Window 8
t
. 2001
^IWG
200^^
2004
2005
0
0
0
x|
967522
1042786
1042145
^£48231
1(OTI^4
0 ^N
0
0
xz
46907
47404
47598
48207
48101
W °
^L
O^W
-------
Table 6.2 - Absolute difference of variables Xj (population) and x2 (personal income) in each window using the first point as the center. For example, the
absolute difference of x^tj) andXj(t2) (i.e., |x;(/;)-x;(/2)| is 4562 as highlighted here in blue and the absolute difference of point 2 from point 1 (i.e., 2-1) is
4562 and 511).
points
2-1
3-1
4-1
5-1
6-1
7-1
8-1
Window 1
X,
4562
8594
73904
90536
114249
120966
126423
x2
511
998
1643
2021
1900
2335
2635
points
5-4
6-4
7-4
8-4
9-4
10-4
11-4
Window 2
X,
16632
40345
47062
52519
64482
119758
167296
X2
378
257
692
992
1177
791
570
points
8-7
9-7
10-7
11-7
12-7
13-7
14-7
Window 3
X,
5457
17420
72696
120234
118323
131663
194397
x2
300
485
99
122
479
316
662
points
11-10
12-10
13-10
14-10
15-10
16-10
17-10
Window 4
X,
47538
45627
58967
121701
157303
211533
250858
x2
221
380
217
563
1661
2890
3663
points
14-13
15-13
16-13
17-13
18-13
19-13
20-13
Window 5
x,
62734
98336
152566
191891
218561
276265
329593
X2
346
1444
2673
3446
4169
4782
5257
points
17-16
18-16
19-16
20-16
21-16
22-16
23-16
Window 6
x,
39325
65995
123699
177027
194782
251975
327239
x2
773
1496
2109
2584
3304
3114
3611
points
20-19
21-19
22-19
23-19
24-19
25-19
26-19
Window 7
x,
53328
71083
128276
203540
202899
208985
257798
X2
475
1195
1005
1502
1696
2305
2199
-------
Table 6.3 - Amplitude (q(s)) for each window. These values were computed by first counting the number of points
in each state within each window, computing a probability for each state in each window (p(s)) and calculating the
amplitude, q(s) = Jp(s) for each state in each window.
Window
1
2
3
4
5
6
7
q(s)
i
0.6124
0.8660
0.7071
0.7071
0.5000
0.6124
0.6124
2
0.7906
0.5000
0.7071
0.5000
0.5000
0.6124
0.7071
3
0.5000
0.6124
0.5000
0.3536
4
0.3536
-------
Table 6.4 - Variables used to calculate FI for the Sail Luis Basin. These variables were selected from data used
to compute EFA. GNRP. and EmA. The data and their sources are discussed in Chapters 3, 4, and 5, respectively.
The variables include demographic, energy, land use. food and forest consumption, environmental and agricultural
production categories for the system.
Category
Demographic
Energy
Land type
Food and forest consumption
Variable
Personal income (thousands of dollars)
Population (persons)
Coal consumption (short tons)
Natural Gas Consumption (billion cubic feet)
Petroleum Consumption (thousand barrels)
Hydro-electric Consumption (million kilowatt hours; kWh)
Wood and waste Consumption (trillion British thermal units; BTU)
Solar Energy Absorbed (Joules per year; J/yr)
Rain Chemical Potential (J/yr)
Rain Geopotential on the Land (J/yr)
Snow Geopotential on Land (J/yr)
Rain Geopotential as Runoff (J/yr)
Snow Geopotential as Runoff (J/yr)
Built-up land (hectares; ha)
Arable land (ha)
Pasture (ha)
Forest including deforestation (ha)
bovine, buffalo (pounds per capita; Ibs/ca)
sheep, goat (Ibs/ca)
non-bovine (Ibs/ca)
milk (gal/ca)
cheese (Ibs/ca)
eggs (Ibs/ca)
fish (Ibs/ca)
cereals (Ibs/ca)
wheat (Ibs/ca)
vegetables (Ibs/ca)
maize (Ibs/ca)
fruit (Ibs/ca)
roots and tubers (Ibs/ca)
pulses (Ibs/ca)
coffee and tea (Ibs/ca)
cocoa (Ibs'ca)
oil seed (Ibs/ca)
fats (Ibs/ca)
sweetener consumption (Ibs/ca)
forest harvest (million board feet MBE)
-------
Category
Environmental
Food production
Variable
mean precipitation (centimeters; cm)
Change in ground water storage (acre-feet)
Change in surface water storage (acre-feet)
SLB CO, emissions
Evapotranspiration (J/yr)
Soil Erosion (tons/year)
bovine, buffalo production (kilograms; kg)
cereal production (kg)
wheat production (kg)
animal feed production (kg)
roots and tubers production (kg)
Potatoes All (Planted) (unitiacre)
Wheat Other Spring (Planted) (unitiacre)
Barley All (Planted) (unitiacre)
Hay Alfalfa (Dry) (Harvested) (unitiacre)
Oats (unitiacre)
Table 6.5 - Three-point mean Fisher information () for the overall system and the variables grouped by the
categories defined in Table 6.4.
Year
1986
1989
1992
1995
1998
2001
System
3.22
3.51
3.64
3.89
3.88
3.58
Consumption
3.34
3.93
3.66
4.13
4.48
4.28
Demographic
5.02
6.04
4.14
2.77
2.81
3.32
Energy
4.59
4.03
4.63
4.63
3.82
3.93
Environment
3.86
3.39
4.61
4.85
4.32
4.42
Land
4.82
5.58
6.32
5.12
6.00
4.73
Production
4.56
4.79
5.99
5.94
6.02
5.03
-------
1980
1985
1990
1995
2000
2005
winspace
hwin
Figure 6.1- Graphical representation of the sample settings for the Fisher information (FI) computation indicating
the size of the integration window (hwin) and window increment (winspace) in time steps used to move through the
data. In this example, hwin =10 and winspace = 5. The red star indicates the center point of the window denoting the
placement of FI.
-*
Iflr//) vft}\ <, AY (fnr all
variables 'in the time step)
Elseif #variables (per time
step) that satisfies the
criteria is >TL*#variables
M Bin in same state
Save point and use
No state
-i
. Goto next
point
Figure 6.2 - Flow diagram of the binning algorithm in the Fisher information computation. According to Equation 6.4,
if each variable \x.(t)- xft)\ is less than the size of states (Ax.), then the two points at time /' and y are indistinguishable
and are grouped (binned) in the same state.
-------
t 1200000
tO
T3
| 1000000
3
E 800000
600000
E
o
u
400000
i
8 200000
60000
50000
40000
£
30000 |
i
20000 Q-
10000
0
Year
Figure 6.3 - Population (xp green triangles) and Personal Income (x2, blue diamonds) values for the San Luis Basin
from 1980 to 2005. These variables were used as data for the example FI computation exercise provided in Section
6.2.4.
2-1
140000
20000
3-1
7-1
5-1
Figure 6.4 - Bulls-eye plot of binning points in state 1 for the simple example. Point 1 (pt: (x^tj), x2(tj)) = (310352,
38469)) is the center of this figure and the absolute difference from ptj are plotted on the corresponding axis, e.g.,
absolute difference 2-1 = (4562, 511). The values plotted are from Table 6.2. As described in Section 6.2.2, the level
of uncertainty (i.e., size of states (Ax)) was computed for each variable such that in this plot AXj and Ax2 correspond
to the red and blue line, respectively. The points are binned in a state when the difference is less than the level of
uncertainty (i.e., size of states).
-------
u.o
0.7-
0.6-
0.5-
|0.4-
0.3-
0.2-
0.1 -
0 -
_
_
~
1
_^
-
_
_
_
l
_
_
_
3
_
4 5
Window
6
1 —
-
_
_
7
D1
D2
D3
• 4
Figure 6.5 - Probability density (p(s)) for each window in the simple example (Section 6.2.4). These values were
computed by first counting the number of points in each state within each window and then computing a probability
for each state in each window. In window 3, four points were binned in both state 1 and state 2; therefore, the
probability was 50% for each state.
(A
0.9-
0.8-
0.7-
0.6-
0.5-
0.4-
0.3-
0.2-
0.1 -
• ' °
..,••'""
=0.61 /
/
/
/
/
/
/
•.
\
\
\ q(0)-q(2)
\
\
\
\
\
D2
State (s)
Figure 6.6 - Calculating gradients of the amplitude (q(s)) in window 1 of the simple example (Section 6.2.4).
Gradients were used as a convention for mimicking a discrete density function and were computed as qfqi+1 where /' is
the state. These values used in Equation 6.3 to compute FI.
-------
Original Fl
1984
1986
1988
1990 1992 1994
< Fl >, Smoothed Fl
1996
1998
2000
2002
B
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
Figure 6.7 - Fisher information for each window from the simple example: (A) Fisher information (Fl)
and (B) Smoothed Fl using three-point mean (), i.e., 1999 = (FI2002+FI1999+FI1996)/3 and 1990 =
(FT +FI, +FTV3.
San Luis Basin Fisher Information
c
o
I
Year
Figure 6.8 - Fisher information calculated for the San Luis Basin over a 26-year period (1980-2006). Each point
represents the Fl computed in each window and is reported in the center of that period (i.e., Fl; 1984 = 1980-1987,
1985 = 1981-1988, 1986 = 1982-1989, etc.).
-------
8.0000
7.0000 -
6.0000 -
5.0000 -
4.0000 -
3.0000
2.0000 -
1.0000 -
0.0000
San Luis Basin Mean Fisher Information
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Year
Figure 6.9 - Mean Fisher information () for the San Luis Basin. was computed as a three-point average of
the FI (Fig. 6.8) in order to smooth out short-term fluctuations and highlight trends, rather than fluctuations. The
is reported in the center of the three points (i.e., ; 2001 = 2002-2000, 1998 = 1999-1997, etc.).
0.00
San Luis Basin Mean FI (System and by Category)
•System —•— Consumption A Demographic 9 Energy X Environment • Land O Production
1986
1989
1992
1995
1998
2001
Time
Figure 6.10- Mean Fisher information of the San Luis Basin for the overall system and by category. Using this plot,
we compared changes in system mean FI () to the the variables grouped in categories. The is reported
in the center of the three points (e.g., 2001 = (FI2000 + FI2001 + FI2002)/3).
-------
0
1980
£20
c
CO
-510
c
0)
u
I
0
1980
SLB Fl by Category (Demographic)
1985 1990 1995 2000
Percent change in SLB Population
2005
1985 1990 1995 2000
Percent change in SLB Personal Income
2005
1985
1990 1995 2000 2005
Year
Figure 6.11- Percent change in the San Luis Basin demographic variables. The percent change in population
corresponded to the changes in dynamic order of the demographic category. The change in population decreased
initially, peaked in 1995, and decreased thereafter which corresponds to the changes in dynamic order for the
demographic category.
-------
Appendix 6-A: Derivation Of Fisher
Information
Fisher information was developed by the statistician
Ronald Fisher as a statistical measure of information in
data being used to fit a parameter (Fisher 1922). It is
formally defined by Frieden (1998, 2004) as:
dp(y\Q)l2
dQ
(6-A.l)
where, / is Fisher information andp(y\8) is the
probability density of observing a particular measured
value of a variable y in the presence of a parameter
0. From this equation, / is a measure of the amount
of information about 0 that is obtainable from the
measurement of y (we use / here to preserve the
historical context; Fisher 1922, Frieden, 1998, 2004,
Karunanithi et al. 2008, Mayer et al. 2007). For
example, if y has no information about 0, then the
derivative in Equation 6-A. 1 is zero and / is zero as well.
The parameter 0 can represent many items. With that in
mind, in order to transform this equation into a measure
of order for complex dynamic systems, let the parameter
0 be the mean of y over a particular period of time,
T:
(6-A.2)
where T is the integration period for all observations of
y. By substituting Equation 6-A.2 into (6-A.l), we have:
/(<>») =
dy \dp(y\}
p(y\)[_ d J (6-A.3)
Next, to represent the fluctuations iny around the mean
, we define a new variable s, such that s = y- .
According to elementary statistics, systems where the
variation around the mean is independent of the value of
the mean are said to be shift-invariant (Frieden, 2004)
p(y \)=p(y- < y >\< y >)=/>(*- < * >)=P*(*) (6.A.4a)
indicating that the probability distribution does not
depend on the value of the mean. Using the results
of Equation 6-A.4a and incorporating the chain rule
(6-A.4b):
dp dp ds dp
d ds d ds (6-A.4b)
where we have tacitly dropped the subscript "0" onp(s),
and Equation 6.A.I becomes:
j- ds \dp(s)
where p(s) is the probability density finding a particular
value of s. This is Equation 6.1 found in the main text
of Chapter 6. Now let s be a state of the system in
a Euclidean (linear) space defined by the observable
variables of the system and time. That is, a particular
state s is a region in a space, (i.e., linear phase space
where the dimensions are the observable variables of
the system and time). Then, p(s) is the likelihood of
observing a particular state, s.
From Equation 6-A.5, we note that / is proportional to
dp / ds. In the context of order, systems can exist within
two idealized extremes, perfect disorder and perfect
order, as discussed in the main text within a different
context. The perfect disorder case occurs when a system
is unbiased toward any particular state. In other words,
it has the same probability of being in one state as any
other state of the system (s=l: n), i.e., p(s) = p(l) = p(2)
=.. .p(n) and the probability density function (PDF) is flat
(Fig. 6-A. la) so that dp /'ds —>0. Accordingly, the system
lacks order (which in some contexts can be thought of
as predictability) and the resulting Fisher information
approaches zero (i.e., /—K)) (Path et al. 2003). Perfect
order occurs when repeated measurements of the system
result in the same state over time. This more structured
system has high order, is more predictable, and is biased
toward a particular state or states. Accordingly, the PDF
has a steep slope, dp /'ds —x» and Fisher information
approaches infinity (i.e., /—x») (Fig. 6-A.lb). However,
real systems typically function between these two system
extremes (e.g., Fig. 6-A.lc).
For completeness, we would like to point out that
one reasonably elegant means of evaluating Equation
6-A.5 is by adapting a statistical mechanics approach
and representing the system in its phase space. Here
the space coordinates are again the measurable system
variables, and the probability density for observing a
particular state (s) is proportional to the time the system
spends in state s, i.e., p(s) oo At(s). This method provides
us with a resulting expression that is a function of the
velocity (R '(t)) and acceleration (R "(/)) tangential to the
system path in its phase space (Mayer et al. 2007). The
expression used to compute the Fisher information under
this theory is:
(6-A.5)
(6-A.6)
However, Equation 6.A-6 requires the evaluation of
the first and second order derivatives tangential to the
system path defined by the system variables (Cabezas et
al. 2005, Path et al. 2003, Mayer et al. 2006). Although
this approach is appropriate for model systems for which
smooth data are readily available, the challenge is that it
-------
is extremely difficult to obtain high quality second order
derivatives from the noisy, sparse data sets characteristic
of real systems (Karunanithi et al. 2008). Accordingly,
the numerical approach (derived and discussed in
Chapter 6) was developed for calculating 1 to assess
the dynamic order of real systems without the need
to compute second derivatives. Further details of the
analytical and numerical derivation of/ can be found in
Mayer et al. (2007) and Karunanithi et al. (2008).
References
Cabezas, H.. Pawlowski. C.W., Mayer, A.L., Hoagland,
N.T. 2005. Simulated experiments with complex
sustainable systems: ecology and technology.
Resources, Conservation and Recycling 44, 279-291.
Path, B., Cabezas, H., Pawlowski C.W. 2003. Regime
changes in ecological systems: an information theory
approach. Journal of Theoretical Biology 222- 4, 517-
530.
Fisher. R.A. 1922. On the mathematical foundations
of theoretical statistics. Phil. Trans. Royal Society of
London 222, 309-368.
Frieden, B.R. 1998. Physics from Fisher Information: A
Unification, Cambridge University Press, Cambridge,
UK.
Frieden, B. R. 2004. Science from Fisher Information:
A Unification. Cambridge University Press,
Cambridge, UK.
Karunanithi A.T., Cabezas. H., Frieden, B.R.,
Pawlowski. C.W. 2008. Detection and assessment
of ecosystem regime shifts from Fisher information,
Ecology & Society 13(1), 22.
Mayer, A.L., Pawlowski C.W.. Cabezas. H. 2006.
Fisher Information and dynamic regime changes in
ecological systems. Ecological Modeling 195, 72-82.
Mayer, A.L., Pawlowski, C.W., Path, B.D., Cabezas,
H. 2007. Applications of Fisher information to the
management of sustainable environmental systems, in:
Frieden, B.R.. and Gatenby, R.A. (Eds.), Exploratory
Data Analysis Using Fisher Information. London:
Springer-Verlag, pp. 217-244.
Pawlowski, C.W., Cabezas, H. 2008. Identification
of regime shifts in time scries using neighborhood
statistics. Ecological Complexity 5, 30-36
-------
c) Medium Fisher
information ,
ds
a) Zero Fisher
information
b) High Fisher
information
dp
I cc-f-
ds
• oo
Figure 6-A. 1 - Fisher information is proportional to the probability of system states (Pawlowski and Cabezas 2008).
(a) Zero / results when a system has an equal probability of being in one state as any other state, resulting in a uniform
probability density function (PDF), such that dp / ds —K) and /—>0. (b) High / occurs when repeated measurements
of the system result in the same state over time (i.e., the system is more predictable resulting in high predictability),
a PDF with a steep slope, dp/ds —*» and Fisher information approaching infinity (i.e., /—*»). (c) However, real
systems typically function between these two extremes and exhibit medium /.
-------
Appendix 6-B: Sensitivity Analysis On
Integration Window Parameters
In section 6.2.4, we used a simple example to
demonstrate the FI computation steps. The FI result
was calculated using the integration window parameters
selected in accordance with the basic criteria (section
6.2.2) and contained seven values (one for each time
window) which were limited by the amount of data
available, hwin, and winspace (Table 6.1). However,
one of the challenges in computing FI is determining
optimal settings for the integration window parameters.
Accordingly, this appendix provides an exploration of
the effect of changing the value of hwin and winspace on
both the FI results and the resolution of results (number
of FI results produced from a computation) for the data
from the simple example. This exercise is not meant
to cover every possibility; however, it is intended to
(1) underscore the fact that care should be taken when
selecting the integration window parameters and (2)
provide key insights on the impact of the parameters on
the FI result.
The basic guidelines for setting the integration window
parameters are hwin > 8 and winspace < hwin (section
6.2.2). For this exercise, we used the data from the
simple example and examined four different values for
hwin (ranging from 8 to 11 time steps) and winspace
(ranging from 1 to 4 time steps). Graphing FI computed
from these settings provide some sense of the effect
changes in winspace and hwin on FI (Fig. 6-B. 1). Each
panel of Fig. 6-B. 1 reflects the FI result with hwin
constant and a different winspace for each computation.
For example, panel A contains the FI results given hwin
= 8 and winspace ranging from 1 to 4. The average
value of FI (avgFI) and standard deviation (SD) of FI
reported were computed from all the FI results in each
panel as a summary statistic. For example, the avgFI
and SD reported in panel A are the average value and
standard deviation of FI given hwin = 8 and winspace
from 1 to 4. In general, smaller winspace values
resulted in more fluctuations in FI, yet, the overall
impact of winspace on the value of FI was relatively
minor. However, looking at avgFI for each panel, as
hwin increased avgFI decreased. We performed similar
analysis on the data by holding winspace constant and
found that when assessing the variability of avgFI, the
standard deviation is 0.11 as winspace changes and 0.44
as hwin changes. Therefore, it appears that the selection
of hwin is the primary factor affecting FI. However, this
only just begins to answer the question of the impact of
the parameters on the FI result and provides minimal
insight regarding optimal settings for hwin or winspace
for this system.
One of the common approaches to assessing the
impact of controllable (independent) variables on
output responses includes evaluating the system one
factor at a time (OFAT). Not only is this approach
time consuming and costly, factor interactions are not
considered (Anderson 2005, Montgomery 1997). In lieu
of an OFAT approach, we opted to perform a sensitivity
analysis by designing a factorial experiment using hwin
and winspace as controllable factors and the average
value of FI (avgFI) and resolution of FI results (Npts)
as output response variables. A designed experiment
affords the ability to determine critical factors by
assessing the impact of controlled variable changes (and
their interactions) on output responses. Using Design
Expert 8 (StatEase, Inc., Minneapolis, MN) we created a
design matrix for the 42 full factorial design showing the
factor combinations and responses for each experiment
(Table 6-B.l). An analysis of variance (at a = 0.05)
of avgFI revealed that both hwin (F (3, 9) = 77.56,
P«0.05) and winspace (F (3, 9) = 4.64, P< 0.05) are
statistically significant factors (Table 6-B.2). However,
note that the sum of squares (SS) and mean squares
(MS) summarizes the amount of variability accounted
for by each factor and random error (Montgomery 1997;
NIST/SEMANTECH 2006). Accordingly, the results
indicate that more of the variation in avgFI comes from
changes in hwin. Similarly, we found that both factors
have a statistically significant impact on the resolution
of the FI results (number of FI results produced from a
computation); however, winspace is dominant (Table
6-B.3). Further, there was a strong negative correlation
(r = -0.9934, P = 0.0066, df= 2) between the avgFI
and hwin (i.e., as hwin increases, avgFI decreases)
and although hwin has little effect on the resolution of
FI results, smaller winspace values produce a higher
resolution of FI results (Figs. 6-B.2 and 6-B.3). From
the analysis, it appears that hwin is the primary factor
affecting the variability and amplitude of the avgFI
result and winspace is the key parameter affecting the
number of FI results produced from a computation
(Npts). The challenge at this point is determining
optimal settings for the integration window parameters
such that the deviation in the avgFI result is minimized
and the number of FI data points is maximized. In order
to determine the optimal settings, we sought a solution
to the multiple objective problem of (a) minimizing
the standard deviation of FI, the standard error of the
mean value of FI and standard error of the Npts and (b)
maximizing Npts (Table 6-B.4). Using the optimization
function within Design Expert 8 by StatEase, Inc., an
optimal solution was found at hwin = 8 and winspace =
1 (Fig. 6-B.4). Using the solution from the sensitivity
-------
analysis and numerical optimization, we calculated FI
and the three-point mean FI result, (Fig. 6-B.5).
These parameter settings were also used to compute FI
for the SLB system.
References
Montgomery, D.C. 1997. Design and Analysis of
Experiments. John Wiley and Sons, Inc., New York,
NY.
NIST/SEMATECH. 2006. e-Handbook of Statistical
Methods. Available online at http://www.itl.nist.
gov/div898/handbook/toolaids/pff/l-eda.pdf. Last
Accessed August 26, 2010.
Table 6-B. 1 - Design matrix for the sensitivity analysis of the simple example data: Determining the impact of
controllable factors, hwin and winspace on the response variables, average value of FI (avgFI) and the number of FI
results (Npts) for computation.
Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Factors
hwin
8
9
9
11
10
11
9
8
8
10
11
9
10
10
8
11
winspace
2
2
1
1
1
2
3
1
3
2
3
4
3
4
4
4
Response
avgFI
3.13511
2.74041
2.87719
2.22149
2.44636
2.13994
3.16116
3.18285
3.24848
2.3549
2.15498
2.90048
2.75425
2.44245
2.99323
2.12815
Npts
10
9
18
16
17
8
6
19
7
9
6
5
6
5
5
4
Table 6-B.2 -Analysis of Variance (ANOVA) forthe average value of FI (avgFI). The results reveal that both hwin (F
(3, 9) = 77.56, P«0.05) and winspace (F (3, 9) = 4.64, P< 0.05) were significant factors affecting avgFI. Moreover,
as an indicator of the amount of variability that is accounted for by factor effects (e.g., hwin or winspace) and random
error (i.e., residual), the MS values reflect that more of the variability was determined by hwin.
Source
Model
A-hwin
B-winspace
Residual
Cor Total
Sum of
Squares (SS)
2.4199
2.2833
0.1366
0.0883
2.5082
df
6
3
3
9
15
Mean
Square (MS)
0.4033
0.7611
0.0455
0.0098
F
value
41.1021
77.5644
4.6397
p-value
Prob > F
< 0.0001
< 0.0001
0.0317
significant
significant
significant
-------
Table 6-B.3 - ANOVA for the resolution of Fisher information results (Npts). Like the ANOVA for avgFI, both hwin
(F (3,9) = 8.33, P <0.05) and winspace (F(3,9) = 519, P «0.05) were statistically significant factors affecting Npts.
However, the mean square (MS) values indicate that majority of the variability in Npts was due to winspace.
Source
Model
A-hwin
B-winspace
Residual
Cor Total
Sum of
Squares (SS)
395.5
6.25
389.25
2.25
397.75
df
6
3
3
9
15
Mean
Square (MS)
65.92
2.08
129.75
0.25
F
Value
263.67
8.33
519
p-value
Prob > F
< 0.0001
0.0058
< 0.0001
significant
significant
significant
Table 6-B.4 - Criteria for determining the optimal value for hwin and winspace for the sample exercise. The goal of
the numerical optimization was to minimize the deviation in avgFI and maximize Npts, subject to (a) minimizing the
standard deviation of FI, the standard error of the mean value of FI and standard error of the Npts and (b) maximize
Npts. The highest importance was placed on maximizing Npts and minimizing the standard error of avgFI and Npts.
The optimization was performed in Design Expert 8 (StatEase, Inc.).
Name
A:hwin
B:winspace
StdErr(mFI)
Npts
StdErr(Npts)
std mFI
Goal
is in range
is in range
minimize
maximize
minimize
minimize
Constraints
Lower
Limit
8
1
0.065521169
4
0.330718914
0.421906393
Upper
Limit
11
4
0.065521169
19
0.330718914
1.178581952
Importance
3
3
4
5
4
3
-------
hwin
8
10
11
•winspace=1 winspace=2 winspace=3 winspace=4
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
avgFI = 2.50, SD = 0.175
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
avgFI = 2.16, SD = 0.042
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Figure 6-B. 1 - The effect of changing hwin on the FI result. The value of hwin changes from panel to panel (A-D),
while winspace varies within each panel. Therefore, each panel reflects the Fisher information result over time with
hwin constant and a different winspace for each computation: (A) hwin =8, winspace =1 to 4, (B) hwin =9, winspace
= 1 to 4, (C) hwin =10, winspace =1 to 4, (D) hwin =10, winspace =1 to 4. The average value (avgFI) and standard
deviation (SD) reported of FI in each panel were computed from all the FI results in each panel as a summary statistic.
For example, avgFI and SD in panel A are the mean and standard deviation of the FI values given hwin = 8 and
winspace = 1 to 4. Note that avgFI decreases as hwin increases and small winspace values result in more fluctuations
(e.g., winspace =1 is represented by the blue line).
-------
Interaction
Interaction
1.5-
A: hwin
O) 2.5-
ro
2 -
B: winspace
B
10
11
B: winspace
A: hwin
Figure 6-B .2 - Interaction plots for the average value of FI (avgFI). Plotting the value of avgFI as a function of
hwin and winspace affords the ability to evaluate the impact of both parameters on avgFI. (A) Each hwin value is
represented by a different color plot. With winspace varying along the x-axis, there is little variability in avgFI as
winspace increases. Conversely, note that as hwin increases (e.g., Al: hwin = 8, A4: hwin = 11), the value of avgFI
decreases. (B) The strong relationship between hwin and avgFI is also shown as hwin increases along the x-axis.
Interaction
Interaction
A: hwin
B: winspace
B
B: winspace
A: hwin
Figure 6-B.3 - Interaction plots for the number of FI results (Npts). Plotting the value of Npts as a function of hwin
and winspace affords the ability to evaluate the impact of both parameters on Npts. (A) Each computation with a
particular hwin value is represented by a different color plot. With winspace varying along the x-axis, there was a
great deal of variability in Npts. (B) However, as hwin increases along the x-axis, there are minor changes in Npts.
Further, there is a greater decrease in Npts as winspace goes from 1 to 2, than from 2 to 3 or 3 to 4.
-------
(0
'55
Q
B: winspace
A: hwin
Figure 6-B.4 - Results of the numerical optimization of hwin and winspace using data from the simple example.
Based on the optimization criteria of maximizing Npts and minimizing the deviation in avgFI, the optimal solution was
found (using Design Expert 8) with integration window parameter settings of hwin = 8 and winspace = 1.
Original Fl
* * * * *
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
< Fl >, Smoothed Fl
B
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Figure 6-B.5 - Final Fisher information (Fl) result from the simple example. (A) Fl result from the optimal parameter
settings (hwin = 8 and winspace =1). (B) Smoothed result: , three-point mean Fisher information.
-------
Appendix 6-C: Luis
- Utilizing MATLAB
(Gfisher.rn CalcFishMean.m)
to Fisher Information
Appendix 6-C.1: Using MATLAB Files to
Calculate Fisher Information via Command Line:
(GFisher.m, CalcAvgFish.m, SmoothFl.m)
The primary MATLAB files developed for this project
include GFisher m, CalcAvgFish m and Main_Fisher_
dala_proc.m (GUI) and AdvancedFishGUI.m (GUI).
Other supporting files arc Nfisherpdf.m, round2 m,
CloseGUI m and SmoothFl.m. In this section, we
present a general description for using MATLAB to
execute the functions (e.g., GFisher m, Calc AvgFish. m)
via command line to compute FI. Using the functions
in this way requires some knowledge of MATLAB;
however, this manual is not intended to be a MATLAB
tutorial. Conversely, the GUIs (i.e., Main_Fisher_data_
proc m and Advanced FishGUI m) arc not command
line functions and are the most user-friendly means
of computing FI. They will be described in Appendix
6-C.2:
The file GFisher m contains the code developed in
conjunction with the ongoing research within the
US EPA's Sustainable Environments Branch for
implementing the numerical approach to computing FI.
Given that m=number of time steps (e.g., years) and
n=number of characteristic variables, the command line
for this code is:
[FI,midt_win"| = GFisher (t.data.sost,hwin,winspace,TL.I
P) where the required inputs into the function are:
the time data for study of size m x 1
t:
data:
sost:
the time series data of variables
characterizing the state of the system
of size m x n
a row vector of the size of states for
each variable, size = lx n
hwin: the size of the time window
winspace: the moving window increment
TL: the tightening level (%)
IP: a toggle on or off (1 or 0) for requesting
an automatic plot of the FI result
The function outputs are:
FI: Fisher information for each time
window
midt_win: Midpoint of each time window
corresponding to the FI result
Prior to executing the code, the user must gather the
input data as described above, to include: time series
data of the variables that characterize the state of the
system (data), the time period (e.g., 1980-2005) that
the data represent time (t), the integration window
parameters (hwin and winspace) and the size of states
(sost) for each variable. Please refer to sections 6.2.2
and 6.2.4 for information on determining and gathering
this information.
Once the data are compiled, they may be entered into
MATLAB directly via command line or imported from
a saved file of a supported type (e.g., csv, Excel). The
data entry method is not critical when executing the
command line functions; however, data must be in a
particular format:
• The time series data of the variables characterizing
the system (data) should be configured such that
data for each variable are listed in one column
with each row containing the variable values for a
particular time step, producing a m* n matrix. For
example, if there are 5 variables and 20 time steps.
The data will be entered into a 20 x 5 matrix.
• The time data, (time, t) should be formatted such that
one time step value is entered in each row producing
a m x 1 column vector with each time period
corresponding to a datum point.
• The size of stales (sost) data should be formatted
such that there is a size of states value for each
variable in each column, producing a 1 x n row-
vector (see 6.2.2 for computing size of states).
For example, the 20 x 5 matrix mentioned above
contains 5 variables and would have 5 size of states
values, one for each variable (i.e., 1x5 row vector).
All of these inputs must be loaded into the workspace
prior to executing the function. In other words, properly
formatted data must be loaded into the workspace and
parameter values must be defined in the MATLAB
command window before entering the command.
While GFisher.m provides the FI result for one specified
tightening level (e.g., 100% = 100), Calc AvgFish m calls
GFisher.m to compute FI by incrementally changing the
tightening level and calculating the average of the FI
from the strict (TL = 100%) to relaxed tightening (sec
Section 6.2.2). The command line for this file is:
LAvgFisLPO_TL,mYJ= CalcAvgFish
(t,data,hwin.winspace,sost)
-------
The function inputs are similar to GFisher and outputs
are:
AvgFish: Average Fisher for each time window
POJTL: Tightening Level the system is in perfect
order. When there is a regime shift, it is an
indication of pervasiveness
mY: Mean year for each time window
The SmoothFI.m file is used to compute the mean FI
. a m point mean which smoothes the FI results to
focus the analysis on trends and not fluctuations. The
command line is: [FTm]=SmoothFT(m,t,FT). where m
is the number of points to be used in calculating the
average, t is the time vector and FI is the computed FI
result from GFisher.m, CalcAvgFish m or Main_Fisher_
data,_proc.m GUT. The output is 2 columns containing
the mean time and mean FI computed. In the SLB
project, m was set to a default of three in order to provide
a three-point mean FI.
6-C.2: Graphical User Interfaces
(GUI) for Calculating Fisher Information:
(Main_Fisher_data_proc.m and AdvFishGUI.m)
To simplify the calculation of FT and enhance user-
friendliness, a GUI was developed as an extension to
the San Luis Basin Sustainabilily Metrics Project (Fig.
6-C.l). The GUI is an interface for data entry, executing
code to compute FI, as well as, displaying and saving-
results from an FI analysis (Fig. 6-C. 1). It is designed to
simplify the procedure by allowing the user to select pre-
formatted data files, alter pre-set parameter values and
process data to calculate FI over numerous tightening
levels (from strict to relaxed tightening). The computed
FI is plotted in the GUI and the user is able to save the FI
computations and plot for further use.
Although MAT LAB can import multiple file types, the
code for the GUI was setup to accept data specifically
in Excel format. Accordingly, although the data files
are created using the data formatting guidelines (see
Appendix 6-C. 1), there are a few caveats:
* The Excel file containing the variable time series
data should be formatted as an mxn matrix with
a descriptive name that ends in dataxls (e.g..
SLBdata.xls).
* The time data should be saved in a separate Excel
file whose descriptive name ends intime.xls (e.g.,
SLBtime.xls) and contains the m x 1 time data, where
m is the number of time steps
• The size of states data should be saved in a separate
Excel file whose descriptive name ends in sost.xls
(e.g., SLBsost.xls) and contains the data in a 1x n
row vector, where n is the number of variables.
• There should be no column headings in these files.
Navigating through the GUI, the [Add/Remove
Files] menu tab contains controls that allow the user
to select the data files for computing FI. The [Add
Files] button is used to identify the location and add
the pre-formatted files. Note that because there is one
evaluation done at a time, only one time series, one size
of states, and one time series data file (e.g., SLBdata.
xls. SLBtime.xls, SLBsost.xls) should be loaded at a
time. The [Delete Files] button may be used to remove
unnecessary files. Once the data, time, and size of state
files have been selected, the user inputs the hwin and
winspace parameters, as well as. the number of points
for the smoothing the FI results. The values for these
parameters are preset to 8, 1, and 3 (respectively) for
this project. The [START] button executes the function
to perform the FI computation. Once this button is
selected, all buttons and figures are disabled until the
computation is complete and then a graph of the result
(FI and smoothed FI ()) is displayed in the Plot
Window. The [Save Plot] and [Save Calculations]
buttons are used to save the FI plot and numerical result
for further use. The [Save Calculations] button saves
the FI values, date and time of calculation, as well as
the smoothed mean FI results into an Excel file in the
following format:
Row 1 Coll: Date
Row2 Coll: Time (military)
Row3 Coll: Perfect order tightening level: PO_TL
Row4 Col2: Average FI results: [Year FI]
The smoothed results follow the average FI results
The [Reset] button allows the user to clear the figure and
parameter settings and the [Help] button hyperlinks to a
help file, which provides instructions on using the Main
GUI. The [Advanced] button opens the Advanced Fisher
GUI, which allows the user to assess the pervasiveness,
and intensity of a regime shift, if one exists (see Section
6.2.2 and Fig. 6-C.2). Such an examination is pertinent
only if a regime shift has been identified.
Given the same data and parameter settings that were
established in Hie Main GUI (Main_Fisher_data_
proc.m), clicking [Study] in the Advanced Fisher
GUI executes the computation of the FI for multiple
tightening levels and plots the result on two axes (axisl:
TL=100%-60% and axis2: 50%-10%) along with the
mean FI within the GUI interface. Through these plots,
the user is able to study the pervasiveness and intensity
of the regime shift as described in Section 6.2.2 and
Karunanithi et al. (2008). Pervasiveness indicates the
number of variables affected by the shift and is estimated
by determining the lowest level in which the drop in FI
is still present. This is based on finding the TL value at
-------
which the system is perfectly ordered (i.e., FI = 8) for
the entire period. The pervasiveness is calculated by
subtracting the perfect order tightening level (POTL)
from 100 and denotes the percentage of variables that
are impacted by the shift. The intensity relates to the
amplitude of the shift and is determined by how far the
FI drops between two stable regimes. A system with a
more intense shift would have a lower drop in FI before
settling into a new dynamic regime.
The main benefit of the graphical interface is that it
requires minimal user input and simplifies the evaluation
process for FI computation. In conjunction with this
project, MATLAB files for the GUIs were made into
executable files using the MATLAB Compiler. This
provides portability to locations where the MATLAB
software is not readily available; thereby making the
GUIs available for use on any desktop. However, these
executable files are not modifiable.
Main_Fisher_data_proc
Fisher information overtime
- Fisher information (FI)
* Smooth FI ()
\\AA.AD.EPA.GOV\ORD\CIN\USE\MAIN\R-L-Z
\\AA.AD.EPA.GOV\ORD\CIN\USE\MAIN\R-L-Z
\AA.AD,EPA.GOV\ORD\CIN\USE\MAIN\R-L-Z
1985 1990 1995 2000
Figure 6-C. 1 - GUI interface developed to compute FI from system data. The GUI is used for (1) importing pre-
formatted data files, (2) choosing parameter settings (3) computing and plotting FI, (4) saving the FI results and (5)
further study of regime shifts. The results plotted in this figure are from the simple example in Appendix 6-D.
-------
ERVASIVENESS AND INTENSITY OF REGIME SHIFT
Pervasiveness and Intensity (100%-60%)
Pervasiveness and Intensity (50%-0%J
1985
1990 1995
Year
2000
Figure 6-C.2 - The Advanced GUI was developed to study the pervasiveness and intensity of ecosystem regime
shifts by (1) evaluating and (2) plotting the FI at various tightening levels. (3) The perfect order tightening level
and pervasiveness are computed and shown in the text boxes. The results plotted are from the simple example as
computed in Appendix 6-D. Using this simple example, note that the pervasiveness is 100% and there is one regime
shift between two stable regimes.
-------
Appendix 6-D: Using MATLAB to
Compute Fl for the Simple Example
In this section, we describe the steps involved in using
the MATLAB GUI to analyze the data from the simple
example. The first step was to save the time series,
time, and size of states data into separate Excel files. In
accordance with the guidelines in Appendix 6-C.2, the
time series of the variables characterizing the system
and the time data (i.e., 1980 to 2005) plotted in Fig. 6.3
were saved as SLBExdataxls and SLBExtime.xls (Tables
6-D. 1 and 6-D.2). The size of states for each variable
as computed in section 6.2.4 (i.e., [86285.39 1639.83])
were saved in SLBExsost.xls (Table 6-D.3). Once the
data files were set up, we opened the Main_data_proc
GUI and clicked the [Add files] button to select the
formatted data files (SLBExtime.xls, SLBExdata.xls and
SLBExsost.xls) for computing FL We left the default
settings for the integration window parameters, hwin =
8, winspace = 1 and the number of smoothing points =
3 and pressed [START] to process the data and generate
a plot of the FI and smoothed FI () results (see Fig.
6-D. 1). After the data were processed, the plot and FI
calculations were saved using the [Save Plot] and [Save
Calculations] buttons.
Although, the data used in this example do not fully
characterize the sustainability aspects of the region
(contains only two demographic variables), for the
sake of this exercise, we assumed that these variables
characterize the system under study. Looking at Fig.
6-C.l, we noted a drop in FI "bookended" by two stable
regimes indicating the existence of a regime shift.
Thus, in order to study the pervasiveness and intensity
of the shift, we clicked the [Advanced] button to open
the AdvancedFish GUI (Fig. 6-C.2). After clicking
the [Study] button, the FI (green plot), as well as FI at
various tightening levels from strict (TL = 100%) to
completely relaxed tightening (TL=0%) are computed
and plotted. In addition, the perfect order tightening
level (POTL) and pervasiveness for the system were
shown in the textboxes. Typically, the POTL may be
observed graphically by noting the lowest tightening
level at which the features of the shift (drop in FI) are
still present. Note that there are only a few plots visible
because there are only two variables in this study. At
any rate, because the features of the shift remain intact
for the plots, the POTL = 0 and pervasiveness = 100%,
both variables are impacted by the shift.
Table 6-D.l - Excel file containing the time series of
the variables characterizing the system for the simple
example (SLBExdata.xls). In this exercise, there are
two variables (Xj and x2) such that each column contains
the time series data for one variable. As described in
Appendix 6-C. 1, the data file is displayed as an m x n
matrix with m = the number of time steps and n = the
number of variables.
x,
310352
314914
318946
384256
400888
424601
431318
436775
448738
504014
551552
549641
562981
625715
661317
715547
754872
781542
839246
892574
910329
967522
1042786
1042145
1048231
1097044
X2
38469
38980
39467
40112
40490
40369
40804
41104
41289
40903
40682
41283
41120
41466
42564
43793
44566
45289
45902
46377
47097
46907
47404
47598
48207
48101
-------
Table 6-D.2 - Excel file containing time data for simple
example (SLBExtime.xls). In this exercise, the time
period was from 1980 to 2005 and represents 26 time
steps. Accordingly, as described in Appendix 6-C.l, the
time file is displayed as an m x 1 matrix with m = the
number of time steps.
Table 6-D.3 - Excel file containing the size of state
(sost) data for simple example (SLBExsost.xls). In this
exercise, there were two variables (Xj and x2) and the size
of states was computed for each variable (AXj and Ax2).
The values are exhibited as a 1 x n matrix with n = the
number of variables.
Year
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
x,
86285.39
X2
1639.83
-------
Appendix 6-E: Fisher Information MATLAB Code
Code is available for download in the zipped file named "Appendix 6E - Fisher Information code.zip"
Gfisher.m
function [FI,midt_win,t_win]=GFisher(t,data,sost,hwin,winspace,TL,IP) %used with GUI
%This program calculates FI as a measure of dynamic order.
%It is one of the products of an on-going effort within the US
"/(Environmental Protection Agency's Sustainable
"/(Environments Branch under the direction of Heriberto Cabezas.
%Key contributors to this effort include:
%Theory, methodology and direction from Heriberto Cabezas as adapted from
%Ronald Fisher's work in theoretical statistics(1922); Chris Palowski for
%the initial coding and framework: recurnumFI4 m (6/17/2004).
%Methodology enhancements and corresponding source code
"/implementation by Arunprakash Karunanithi (3/13/2006) and additional
"/modifications by Tarsha Eason (2008-2009).
"/further details on the methodology may be found in:
%Karunanithi, AT. Cabezas H, Frieden BR and Pawlowski CW. 2008. Detection
%and Assessment of ecosystem regime shifts from FI. Ecology
%& Society. 13(1): 22.
%This program calculates the FI from time series data.
"/Pertinent parameters include:
% FI
% midt_win
% t_win
% t
% data
% sost
% hwin
% winspace
% TL
% IP
%profile on
Fisher information
mid point of each time window
time within each window
time data
Input data (matrix), size (#time steps, #variables)
size of states
window size(number of points in each calculation)
number of time steps we move the window (on)
t_win=[];
window=0;
for i=l:winspace:size(data,l) %start big loop to go through all the data
%Revised Method: #pts=hwin
lmin=min(i,size(data, 1));
lmax=min(i+hwin- 1 ,size(data, 1 )) ;
NP=size(data(lmin:lmax, 1), 1);
if NP== hwin
window=window+ 1 ;
[pdf,neighbour]=Nfisherpdf(data,lmin,lmax,sost,TL/100); "/calculate pdf for data
%pdf
%neighbour
q=sqrt(pdf(:,l)); %convert to amplitude of the pdf
"/modification
counter=0:
Q=0;
-------
neighbourQ=[];
forj=l:size(q,l)
ifq(j,l)~=0
countei=counter+l ;
Q(counter, 1 )=q(j , 1 ) ;
neighbourQ(counter, : )=neighbour(j , : )
end
end
%neighbourQ
%Q
%This portion of code re-arranges the states by assessing proximity. As
%such, it calculates the Euclidean distance (zz) of the points in the state.
%finds the smallest distance and orders the q vector by the Euclidean
%distance. This ensures that the states are indeed ordered by distance.
if size(neighbourQ, 1)>2
minimumneighbourQ=0;
tempneighbourQ=0 ;
tempQ=0;
z=[];
for ii=l:(size(neighbourQ,l)-l)
for jj=l :size(neighbourQ, 1)
ifjj>ii
z(iijrj)=sqrt(sum(neighbourQ(ii,:)-neighbourQ(jrj,:))A2);
else
z(ii,jrj)=5000000000;
end
end
minimumneighbourQ=min(z(ii, : )) ;
for kk=2:size(neighbourQ,l)
ifkk>ii
if z(ii,kk)==minimumneighbourQ
tempneighbourQ=neighbourQ(ii+ 1 , :) ;
tempQ=Q(ii+l);
neighbourQ(ii+l,:)=neighbourQ(kk,:);
Q(ii+l)=Q(kk);
neighbourQ(kk,:)=tempneighbourQ;
Q(kk)=tempQ;
end
end
end
end
end
QQ=[0;Q(l:end);0]; %adding points (beginning and end) for edge gradients
dq=diff(QQ)./l; %calculating dqs
%Calculates fisher for each window and places it in the middle of the %window
if (i+(hwin-winspace))<=size(data, 1)
t_win(lmin:lmax,window)=[t(lmin:lmax)]; %time within each window
FI(window)=[4*sum(dq.A2*l,l)]; %One fisher for each window
midt_win(window)=ceil(mean(nonzeros(t_win(:,window))));
%midt_win(window)= mean(nonzeros(t_win(:,window))); %forode calc
end
end
-------
end
%FI
ifIP==l %plot outputs
subplot(2,l,l), plot(t,data) %plots data
%subplot(2,l,l), plot(data(:,l),data(:,2)) %plots data for ode
title('Time series data');
subplot(2,1,2),plot(midt_win,FI,' -b*')
title('Fisher information for each time window');
axis([min(t) max(t) 0 8]);
xlabel(,Time'); ylabel(,Fisher information');
end
-------
NFisherpdf.m
function [pdf,neighbour]=NFisherpdf(data,lrmn,lmax,sizeofstates,TL)
%The methodolgy for calculating FI as a sustainability
%indicator as a measure of dynamic order has been an on-going interative
%effort within the US Environmental Protection Agency's Sustainable
%Environments Branch under the direction of Heriberto Cabezas.
%Key contributors to this effort include:
%Theory, methodology and direction from Heriberto Cabezas as adapted from
%Ronald Fisher's work in theoretical statistics(1922); Chris Palowski for
%the conceptual framework and coding of initial Matlab program, recurnumFI4.m
%(6/17/2004); Methodology enhancements and corresponding source code
%implementation by Arunprakash Karunanithi (3/13/2006) and additional
%functional and conceptual modifications by Tarsha Eason (2008-2009).
%Further details on the methodology can be found in: Karunanithi AT.
%Cabezas H, Frieden BR and Pawlowski CW. 2008. Detection and Assessment of
%ecosystem regime shifts from FI. Ecology & Society. 13(1): 22.
%The purpose of this program is to bin the points into states by
%calculating the difference (dist) of each variable vector per timestep
%and then counting the variable vector if dist<= size of the state AND the
%percentage of variables in the vector that meets that criteria is over the
%tightening level (TL). Further, the number of vectors that fit within a
%state are counted(rnum)and then used to calculate the pdf (rum/sum(rum)).
fp_tol=le-12; %correcting for floating point significant digits
% Initialize the vector which will hold the counter for points in a state
rnum=zeros(lmax-lmin+l, 1);
%Initialize array of points that will be counted in line with Imin and Imax.
%This dummy variable is used to ensure that the rows are not double
%counted.
pout=zeros(lmax-lmin+l, 1);
%
% now, find recurrence points about each in window and record their
% instances, ignoring points that have already been counted
for j=lmin:lmax %looping over the window in time steps
j;
count=0;
if pout(abs(j-lmin)+l,l)==0 %no double counting variable vector
pouttemp=zeros(size(pout));
for k=lmin:lmax
k;
Mcount=0;
for m=l:size(data,2) %entire column length(#of variables)
m;
dist=abs(data(j,m)-data(k,m));%for each variable 1 by 1
-------
%Dealing with floating point error
r_dist=round2(dist,fp_tol); %rounds it to certain precision
if r_dist<=(sizeofstates(l,m))
Mcount=Mcount+l;
end
end
%Ensures that the number of variables that have been counted (meet
%the criteria is >= TL*total number of variables and this variable
%vector has not been previously counted
if
Mcount>=TL*size(data,2))&(pout(abs(lmin-k)+l,l)==0)
count=count+l;
pouttemp(abs(lmin-k)+1,1 )=1;
end
end
rnum(abs(lmin-j)+l,l)=count; %number of points in state
neighbour(abs(lmin-j)+1,: )=data(j,:);
end
pout=pout|pouttemp; %update no double counting using logical OR
end
%Calculating pdf
if sum(rnum,l)==0
pdf=rnum;
else
pdf=rnum/sum(rnum, 1);
end
size 1=size(neighbour);
pdf;
-------
CalcAvgFish.m
function [AvgFish,PO_TL,mY]= CalcMeanFish(t,data,hwin,winspace,sost)
%This code was developed to calculate the mean value of the Fisher
%information by computing FI over various tightening levels to find the
%point above which perfect order is experienced. And then calculating a
%mean of the values for each time window. The program returns:
%AvgFish: Average Fisher information over TL from strict to relaxed
%PO_TL: Tightening Level the system is in perfect order. It is
% connected to pervasiveness (1-PO_TL)
%mY: Mean Year for each time window
% The code was developed by Tarsha Eason (2009)
%[AvgFish Y TL t_win]= Availdata(t,data,hwin,winspace,sost)
TL=[];FisherMat=[];
profile on
displayC Working....');
for 1=1:100
Tol=100-(i-l);
[FI,midt_win]=GFisher(t,data,sost,hwin,winspace,Tol,0);
FMat(:,i)=FI;
n=size(FI,2); %# of Fisher information Calculations
%Calculates average FI for each time window by seeking the tightening
%level at which all resulting fisher calculation for each widow is at its
%max = 8, perfect order. The average FI is calculated
%from the FisherMat which stores all of the Fisher results from each time
%window less the windows where all of the Fisher results are 8.
if sum(FMat(:,i))<8*n %Tightening level with all Fishers = 8, perfect order
%display('Yes');
FisherMat(:,i)=FMat(:,i);
TL=Tol; %Final value is lowest tightening level before all Fisher results are 8 (max FI)
%else
% display('End TL Calcs');
end
end %End Tol loop
if isempty (TL)==1
display('The system is in perfect order for this size of states at all tightening levels. Please feel free to make an
adjustment accordingly!!');
f = warndlg('The system is in perfect order for this size of states at all tightening levels. Please feel free to make an
adjustment accordingly!!', 'Warning Dialog');
PO_TL=100;
else
PO_TL=TL-1;
end
mY=midt win:
-------
%Provides an ouput when the system is competely orderly at all tightening
%levels. This results in no collection of Fisher values before the system
%is a maximum level. Accordingly, FisherMat is empty indicates either true
%perfect order or a need to adjust the size of states (smaller)
if isempty (FisherMat)== 1 ;
AvgFish=8*ones(size(FI));
else
%Calculates mean FI for each time window
AvgFish=mean(FisherMat ' ) ;
end
%Provides an ouput when the system is competely orderly at all tighten
%levels. Indicates either true perfect order or a need to adjust the size
%of states (smaller)
%Transposing (aesthetic)
AvgFish=AvgFish' ;
mY=midt_win';
subplot(2,l,l), plot(t,data) %plots data
%subplot(2,l,l), plot(data(:,l),data(:,2)) %plots data for ode
title(Time series data');
subplot(2,l,2), plot(mY,AvgFish)
axis([mY(l) mY(end) 0 9])
title('Fisher information over time');
xlabel(Time'); ylabel('Fisher information');
-------
SfnootfiF/.in
function [FIm]=SmoothFI(n,t,FI,plot)
% Computes , a "m" point mean of FI smoothing results to focus on trends
% and not fluctuations.
% The code was developed by Tarsha Eason (2010)
data=[tFI];
A=sortrows(data,-l);FIs=A(:,2);ts=A(:,l);
window=0; I=[];T=[];A=[];B=[]; FIm=[];
for i=l n:size(FI,l) %loop throughFIs
%Revised Method: #pts=hwin
lmin=min(i,size(FIs, 1));
lmax=min(i+n-l,size(FIs, 1));
NP=size(FIs(lmin:lmax,l),l);
ifNP==n
window=window+l ;
I(window)= mean(FIs(lmin:lmax));
T(window)=ceil(mean(ts(lmin: Imax))) ;
end
end
B=[TT];
FIm=sortrows(B, 1);
ifplot==l
subplot(2,l,l),plot(t,FI,'-*'); %plot original FI
title('Original FF); axis([t(l) t(end) 0 8]);
subplot(2,l,2),plot(FIm(:,l),FIm(:,2),'-*'); %plot
title(' Smoothed FI(< FI >)'); axis([t(l) t(end) 0 8]);
end
-------
GUI: Main_Fisher_data_proc m
%This program provides a graphical user interface (GUI) for performing
%Fisher information(FI)calculations. It is designed to simplify the
%procedure by allowing the user to select formatted data files, set
%parameter values and process the data which calculates FI over numerous
%tightening levels (from strict to relaxed tightening) providing an average
%FI result. The computed FI is plotted in the GUI, along with a plot of the
%FI results that have been "smoothed" using an "m" point average ().
%The user is able to save the FI computations and plot for further use. It %also provides access for the advanced
study of regime shifts, if they exist.
"/(Procedure for use
%
%Select the formatted data files (*time.xls, *data.xls and *sost.xls; see
%Appendix 6-B.2 for file formatting details) and press the
%' START' button to process the data and generate a plot of the
%time-averaged fisher. After the data is processed, the plot and
%calculations may be saved by using the 'Save Plot' and 'Save Calculations'
%buttons. The 'Advanced' button calls the AdvancedFisher GUI for further
% Fisher information analysis.
%Main_Fisher_data_proc was developed by Tarsha Eason (2008-20 10)as a
%graphical extension to work done from 2004 to 2008 by Cabezas, Pawloski.
%Karunithini and Eason to calculate FI in the Gfisher.m and
%Nfisherpdf.m code. Please feel free to refer to the source code of those
%files for additional references.
function varargout = Main_Fisher_data_proc(varargin)
gui_Singleton = 1;
gui_State = struct('gui_Name', mfilename, ...
'gui_Singleton', gui_Singleton, ...
'gui_OpeningFcn', @Main_Fisher_data_proc_OpeningFcn, ...
'gui_OutputFcn', @Main_Fisher_data_proc_OutputFcn, ...
'gui_LayoutFcn', [],...
'gui_Callback', []);
if nargin && ischar(varargin{ 1 })
gui_State.gui_Callback = str2func(varargin{l});
end
if nargout
[varargout{ 1 :nargout}] = gui_mainfcn(gui_State, varargin{:});
else
gui_mainfcn(gui_State, varargin{ : });
end
handles.processDataCompleted = 0; %
function Main_Fisher_data_proc_OpeningFcn(hObject, eventdata, handles, varargin)
handles.output = hObject;
%set(hObject, 'toolbar', 'figure'); %enables toolbar
%this variable used to prevent users from breaking the GUI
%the variable is set to 1 once the data has been processed
-------
%this command asks the user to confirm closing of GUI
set(handles.figure 1 , ' CloseRequestFcn' , 'closeGUF);
% Update handles structure
guidata(hObject, handles);
% UIWAIT makes Main_Fisher_data_proc wait for user response (see UIRESUME)
% uiwait(handles. figure 1);
% — Outputs from this function are returned to the command line.
function varargout = Main_Fisher_data_proc_OutputFcn(hObject, eventdata, handles)
% varargout cell array for returning output args (see VARARGOUT);
% hObject handle to figure
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
% Get default command line output from handles structure
varargout{l} = handles.output;
function inputFiles_listbox_Callback(hObject, eventdata, handles)
%no code needed for this callback, the listbox is only used as a visual
function addFiles_pushbutton_Callback(hObject, eventdata, handles)
%gets input file(s) from user
[input_file,pathname] = uigetfile( ...
{'*.xls', 'Data Files (*.xls)'; ...
'*.*', 'All Files (*.*)'},...
'Select files', ...
'MultiSelect', 'on');
%if file selection is cancelled, pathname should be zero
%and nothing should happen
if pathname == 0
return
end
%gets the current data file names inside the listbox
inputFileNames = get(handles.inputFiles_listbox,'String');
%if they only select one file, then the data will not be a cell
if iscell(input_file) == 0
%add the most recent data file selected to the cell containing
%all the data file names
inputFileNames{length(inputFileNames)+l} = fullfile(pathname,input_file);
-------
%else, data will be in cell format
else
%stores full file path into inputFileNames
for n = 1 :length(input_file)
inputFileNames{length(inputFileNames)+l } = fullfile(pathname,input_file{n});
end
end
%updates the gui to display all filenames in the listbox
set(handles.inputFiles_listbox, ' String',inputFileNames);
%make sure first file is always selected so it doesn't go out of range
%the GUI will break if this value is out of range
set(handles.inputFiles_listbox,' Value', 1);
% Update handles structure
guidata(hObject, handles);
%%
function deleteFiles_pushbutton_Callback(hObject, eventdata, handles)
%this function allows the user to delete files they accidentally
%added to the list box
%get the current list of file names from the listbox
inputFileNames = get(handles.inputFiles_listbox,'String');
%get the values for the selected file names
option = get(handles.inputFiles_listbox,'Value');
%is there is nothing to delete, nothing happens
if (isempty(option) == 1 || option(l) == 0 || isempty(inputFileNames))
return
end
%erases the contents of highlighted item in data array
inputFileNames(option) = [];
%updates the gui, erasing the selected item from the listbox
set(handles.inputFiles_listbox, ' String',inputFileNames);
%moves the highlighted item to an appropriate value or else will get error
if option(end) > length(inputFileNames)
set(handles.inputFiles_listbox,'Value',length(inputFileNames));
end
% Update handles structure
guidata(hObject, handles);
function edit_hwin_Callback(hObject, eventdata, handles)
% hObject handle to editjiwin (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
% Hints: get(hObject,'String') returns contents of inputText_hwin as text
% str2double(get(hObject,'String')) returns contents of edit_hwin as a double
-------
HW=str2num(get(hObject,'String'));
%checks to see if input is empty. If so, default input l_editText to zero
if(isempty(HW))
set(hObject,'String','0')
end
guidata(hObject, handles);
% — Executes during object creation, after setting all properties.
function edit_hwin_CreateFcn(hObject, eventdata, handles)
% hObject handle to editjiwin (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles empty - handles not created until after all CreateFcns called
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObject, 'BackgroundColor' , 'white ');
end
function edit_winspace_Callback(hObject, eventdata, handles)
% hObject handle to edit_winspace (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
% Hints: get(hObject,'String') returns contents of inputText_winspace as text
% str2double(get(hObject,'String')) returns contents of inputText_winspace as a double
WS=str2num(get(hObject,'String'));
%checks to see if input is empty. If so, default input l_editText to zero
if(isempty(WS))
set(hObject,'String','0')
end
guidata(hObject, handles);
% — Executes during object creation, after setting all properties.
function edit_winspace_CreateFcn(hObject, eventdata, handles)
% hObject handle to edit_winspace (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles empty - handles not created until after all CreateFcns called
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObject, 'BackgroundColor' , 'white ');
end
%%
function edit_Smoothpts_Callback(hObject, eventdata, handles)
% hObject handle to edit_winspace (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
-------
% Hints: get(hObject,'String') returns contents of inputText_winspace as text
% str2double(get(hObject,'String')) returns contents of inputText_winspace as a double
nS=str2num(get(hObject,'String'));
%checks to see if input is empty. If so, default inputl_editText to zero
if(isempty(nS))
set(hObj ect, ' String' , ' 0 ' )
end
guidata(hObject,handles);
% — Executes during object creation, after setting all properties.
function edit_Smoothpts_CreateFcn(hObject, eventdata, handles)
% hObject handle to edit_winspace (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles empty - handles not created until after all CreateFcns called
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObj ect, 'BackgroundColor ' , 'white ' ) ;
end
function reset_pushbutton_Callback(hObject, eventdata, handles)
%resets the GUI by clearing all relevant fields
handles.processDataCompleted = 0;
cla(handles.axesl, 'reset');
set(handles.edit_hwin, ' String' ,' 8 ');
set(handles.edit_winspace,'String','l');
set(handles.edit_Smoothpts, ' String' , ' 3 ');
%set(handles.inputFiles_listbox,'String',");
%set(handles.inputFiles_listbox,'Value',0);
guidata(hObject, handles);
function start_pushbutton_Callback(hObject, eventdata, handles)
inputFileNames = get(handles.inputFiles_listbox,'String');
%checks to see if the user selected any input files
%if not, nothing happens
if isempty(inputFileNames)
return
end
%disables the button while data is processing
disableButtons(handles) ;
refresh(Main_Fisher_data_proc) ;
%initialize the cell structures
handles.data = {};
handles.time = {};
sost=[];
%Read input data
-------
for x=l :length(inputFileNames)
A=inputFileNames{x};
a=size(A,2);
b=a-7;
if(A(b:a)=='data.xls')
%display('Loading Variable Data');
data=xlsread(inputFileNames{x});
%size(data)
elseif(A(b:a)=='time.xls')
%display('Loading Time Data');
time=xlsread(inputFileNames{x});
%size(time)
elseif(A(b:a)=='sost.xls')
%display('Loading Size of states Data');
sost=xlsread(inputFileNames{x});
%size(sost)
else
display('Not correct input filename. See Appendix 6-B.2');
end
end
%Calculate Fisher information
t=time;
handles.data=data;
handles.t=time;
Flag=isempty(sost); %test to see if *sost.xls file was uploaded
ifFlag==l
%display('Since no size of states file was added. The size of states will be calculated from the first 5 data points')
errordlgCSince no size of states file was added. The size of states will be calculated from the first 5 data points','File
Error');
sost=2*std(data(l:5,:));
end
handles. sost=sost;
%get hwin and winspace from input form
H=get(handles .edit_hwin,' String');
W=get(handles.edit_winspace,'String');
nS=get(handles.edit_Smoothpts,'String');
hwin= str2num(H);
winspace=str2num(W);
nSpts=str2num(nS);
handles hwin=hwin;
handles.winspace=winspace;
TL=[];FisherMat=[];
%Using search algorithm for finding the size of states
%[s_sost,region_year,diff_test]=statesize(4,hwin,winspace,data,time);
%sost=s sost:
-------
%Calculating lowest tightening level before all FI results are 8
%(max): PO_TL
for i= 1:100
Tol=100-(i-l);
[FI, midt_win, t_win]=GFisher(t,data,sost,hwin,winspace,Tol,0);
%YMat(:,i)=t_win;
%YMat(:,i)=t_win(i);
FMat(:,i)=FF;
n=size(FI,2); %# of FI Calculations
%Calculates FI for each time window by seeking the tightening
%level at which all resulting fisher calculation for each widow is at its
%max = 8, perfect order. The average fisher is calculated
%from the FisherMat which stores all of the FI results from each time
%window less the windows where all of the FI results are 8.
if sum(FMat(:,i))<8*n %Tightening level with all FI values = 8, perfect order
% display('Yes');
FisherMat(: ,i)=FMat(: ,i);
TL=Tol; %Final value is lowest tightening level before all FI results are 8
else
%display('End TL Calcs');
end
end %End Tol loop
ifisempty(TL)==l
display(The system is in perfect order for this size of states at all tightening levels. Please feel free to make an
adjustment accordingly!!');
f = warndlg('The system is in perfect order for this size of states at all tightening levels. Please feel free to make an
adjustment accordingly!!', 'Warning Dialog');
PO_TL=100;
else
PO_TL=TL-1;
end
handles.PO_TL=PO_TL;
%Provides an ouput when the system is competely orderly at all tighten
%levels. This results in no collection of FI values before the system is a
%maximum level. Accordingly, FisherMat is empty indicates either true
%perfect order or a need to adjust the size of states (smaller)
if isempty (FisherMat)== 1;
AvgFish=8*ones(size(FI));
else
%Calculates Avg FI for each time window
AvgFish=mean(FisherMat');
end
-------
T=zeros(size(t_win));
P=Tt_win;
t_win;
handles.AvgFish=AvgFish;
%AvgFish
for i=l :size(AvgFish,2)
AvgFI(:,i)=P(:,i)*AvgFish(i);
end
handles. AvgFI_win=AvgFI;
handles.PO_TL=PO_TL;
%PO_TL
%plot mean FI over time window
axes(handles.axes 1)
YMat=t(l:size(handles.AvgFI_win,l)); %time period
handles.YMat=YMat;
%%%%Plot Avg FI over time window%%%%%%%%%%%%%%%%%
%plot(YMat,handles.AvgFI_win)
%title('Average FI vs. Time');
%%%%%Plot average FI in center of window%%%%%%%%%%%%%%%%%
for j=l :size(t_win,2)
R(: j )=nonzeros(t_win(: j));
Z(:j)=AvgFish(])*ones(size(R,l),l);
Y(j)=ceil(mean(R(: j))); %mean year for each time window
end
plot(Y,AvgFish,'*b-')
handles.Year=Y;
title('Average FI overtime');
hold
n=nSpts; %n=3; %make sure to create an edit text to change this value
[FIm]=SmoothFI(n,Y',AvgFish',0);
handles.MnYr=FIm(:,l);
handles.MnFI=FIm(:,2);
plot(FIm(:,l),FIm(:,2),'*r-')
legend('Fisher information(FI)','Smoothed FI ()')
legend(handles.axesl,'boxoff')
axis([Y(l,l) Y(end) 0 8]); xlabel('Year'); ylabel('Fisher information');
enableButtons(handles);
handles.processDataCompleted = 1;
guidata(hObject, handles);
%%
function savePlot_pushbutton_Callback(hObject, eventdata, handles)
-------
%if the data hasn't been processed yet.
%nothing happens when this button is pressed
if (handles.processDataCompleted == 0)
return
end
disableButtons(handles) ;
refresh(Main_Fisher_data_proc) ;
savePlot(handles.axesl); %figure out how to save the legend?
enableButtons(handles);
guidata(hObject, handles);
function export_pushbutton_Callback(hObject, eventdata, handles)
%This function saves the mean FI values, date and time of
%calculation, as well as the smoothed mean FI results into an Excel file in
%the format:
%Rowl Coll: Date
%Row2 Coll: Time (military)
%Row3 Coll: Perfect order tightening level: PO_TL
%Row4 to Row4+size(FI) Col2: [YearFI]
%The Smoothed mean FI results follow
%if the data hasn't been processed yet.
%nothing happens when this button is pressed
if (handles.processDataCompleted == 0)
return
end
calctime=clock;
%stores savepath for results
[filename, pathname] = uiputfile ({'*.xls', 'DataFiles (*.xls)'; ...
'*.*', 'All Files (*.*)'},...
'Save file as', 'default');
%if user cancels save command, nothing happens
if isequal(filename,0) || isequal(pathname,0)
return
end
%saves the data
dlmwrite(fullfile(pathname, filename), [calctime(2),calctime(3),calctime( 1)] , 'delimiter' , '/' , ' -append');
dlmwrite(fullfile(pathname, filename), [calctime(4) calctime(5)], 'delimiter', ':', 'precision', '%4.2d', '-append');
dlmwrite(fullfile(pathname, filename),handles.PO_TL,'-append');
%dlmwrite(fullfile(pathname, filename), 'Mean FF, '-append')
dlmwrite(fullfile(pathname, filename),[handles.Year',handles.AvgFish'], '-append', 'delimiter', '\t','roffset',l)
dlmwrite(fullfile(pathname, filename),[handles.MnYr,handles.MnFI], '-append', 'delimiter', '\t','roffset' ,2)
-------
function disableButtons(handles)
set(handles.figure 1 , 'Pointer' , 'watch');
set(handles. start_pushbutton, 'Enable ' , ' off' ) ;
set(handles reset_pushbutton, 'Enable ' , ' off ' ) ;
set(handles. addFiles_pushbutton, 'Enable ' , ' off ' ) ;
set(handles. savePlot_pushbutton, 'Enable ' , ' off ' ) ;
set(handles.deleteFiles_pushbutton,'Enable','off');
set(handles. inputFiles_listbox, 'Enable ' , ' off ' ) ;
set(handles.export_pushbutton,'Enable','off');
set(handles. Advanceduser_pushbutton, 'Enable ' , ' off ' ) ;
set(handles.Help_pushbutton,'Enable','off');
function enableButtons(handles)
set(handles.figurel, 'Pointer', 'arrow');
set(handles. start_pushbutton, 'Enable ' , ' on' ) ;
set(handles reset_pushbutton, 'Enable ' , ' on' ) ;
set(handles. addFiles_pushbutton, 'Enable ',' on' );
set(handles. savePlot_pushbutton, 'Enable ' , ' on' ) ;
set(handles.deleteFiles_pushbutton,'Enable','on');
set(handles. inputFiles_listbox, 'Enable ' , ' on' ) ;
set(handles. export_pushbutton, 'Enable ' , ' on' ) ;
set(handles. Advanceduser_pushbutton, 'Enable ' , ' on' ) ;
set(handles.Help_pushbutton, 'Enable ' , 'on');
function inputFiles_listbox_CreateFcn(hObject, eventdata, handles)
% Hint: listbox controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObject, 'BackgroundColor' , 'white ');
end
% — Executes on button press in Advanceduser_pushbutton.
function Advanceduser_pushbutton_Callback(hObject, eventdata, handles)
% hObject handle to Advanceduser_pushbutton (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
AdvancedFisherGUI() %Call subGUI for further analysis
guidata(hObject, handles);
function Help_pushbutton_Callback(hObject, eventdata, handles)
% hObject handle to editjiwin (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
%testflag=250000
HelpPath = which('helpMain.html');
web(HelpPath); %opens GUI help file
guidata(hObject, handles);
-------
AdvancedFisherGUI m
function varargout = AdvancedFisherGUI(varargin)
%This program is called from the Fisher_data_proc GUI and provides a
%graphical user interface for studying the pervasiveness and intensity of
%regime shifts by plotting the Fisher information (FI) results for tightening
%levels (TL) from 100%-10% overlaid by the average FI. Pervasiveness indicates
%the number of variables affected by the shift and is determined by finding
%the TL value at which the system is perfectly ordered (i.e. FI is at its
%max (8) for the entire period. Pervasiveness is computed by subtracting
%the perfect order tightening level (POTL) from 100) and denotes the percentage
%of variables that were impacted. The intensity relates to the amplitude of
%the shift and is determined by how far the FI drops between two stable regimes.
%NOTE: Data files must be selected through the Fisher_data_proc GUI and
%processed prior to doing advanced analysis through this GUI.
%
%AdvancedFisher was developed by Tarsha Eason(2008-2010) as a graphical
%extension to work done from 2004 to 2008 by Cabezas, Pawlowski, Karuninithi
%and Eason to calculate FI in the Revfisher m, fisherpdf m code.
%Please feel free to refer to the source code of those files for additional
%references.
%Matlab GUIDE information
% ADVANCEDFISHERGUI M-file for AdvancedFisherGUI.fig
% ADVANCEDFISHERGUI, by itself, creates a new ADVANCEDFISHERGUI or raises
% the existing singleton*.
%
% H = ADVANCEDFISHERGUI returns the handle to a new ADVANCEDFISHERGUI or
% the handle to the existing singleton*.
%
%ADVANCEDFISHERGUI('CALLBACK',hObject,eventData,handles,...) calls the
%local function named CALLBACK in ADVANCEDFISHERGUI.M with the given input
%arguments.
%
%ADVANCEDFISHERGUI('Property', 'Value',...) creates a new ADVANCEDFISHERGUI
%or raises the existing singleton*. Starting from the left, property value
%pairs are applied to the GUI before AdvancedFisherGUI_OpeningFcn gets
%called. An unrecognized property name or invalid value makes property
%application stop. All inputs are passed to AdvancedFisherGUI_OpeningFcn
%via varargin.
% Last Modified by GUIDE v2.5 17-Jun-2010 12:44:09
% Begin initialization code - DO NOT EDIT
gui_Singleton = 1;
gui_State = struct('gui_Name', mfilename, ...
'gui_Singleton', gui_Singleton, ...
-------
'gui_OpeningFcn', @AdvancedFisherGUI_OpeningFcn,...
'gui_OutputFcn', @AdvancedFisherGUI_OutputFcn,...
'gui_LayoutFcn', [],...
'gui_Callback', []);
if nargin && ischar(varargin{ 1})
gui_State.gui_Callback = str2func(varargin{l});
end
if nargout
[varargout{l:nargout}] = gui_mainfcn(gui_State, varargin{:});
else
gui_mainfcn(gui_State, varargin{:});
end
% End initialization code - DO NOT EDIT
% — Executes just before AdvancedFisherGUI is made visible.
function AdvancedFisherGUI_OpeningFcn(hObject, eventdata, handles, varargin)
% This function has no output args, see OutputFcn.
% hObject handle to figure
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
% varargin command line arguments to AdvancedFisherGUI (see VARARGIN)
% Choose default command line output for AdvancedFisherGUI
handles.output = hObject;
%this command asks the user to confirm closing of GUI
set(handles.figure 1,' CloseRequestFcn','closeGUF);
%set(hObject,'toolbar','figure'); % activates toolbar
% Update handles structure
guidata(hObject, handles);
% UIWAIT makes AdvancedFisherGUI wait for user response (see UIRESUME)
% uiwait(handles.figure 1);
% — Outputs from this function are returned to the command line.
function varargout = AdvancedFisherGUI_OutputFcn(hObject, eventdata, handles)
% varargout cell array for returning output args (see VARARGOUT);
% hObject handle to figure
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
% Get default command line output from handles structure
varargout{l} = handles.output;
% — Executes during object creation, after setting all properties.
function inputText_hwin_CreateFcn(hObject, eventdata, handles)
% hObject handle to inputTextJiwin (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles empty - handles not created until after all CreateFcns called
-------
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObj ect, 'BackgroundColor', 'white');
end
function inputText_winspace_Callback(hObject, eventdata, handles)
% hObject handle to inputText_winspace (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
% Hints: get(hObject,'String') returns contents of inputText_winspace as text
% str2double(get(hObject,'String')) returns contents of inputText_winspace as a double
%WS=str2num(get(hObject,' String'));
%checks to see if input is empty. If so, default inputl_editText to zero
%if(isempty(WS))
% set(hObject,'String','0')
%end
guidata(hObject,handles);
% — Executes during object creation, after setting all properties.
function inputText_winspace_CreateFcn(hObject, eventdata, handles)
% hObject handle to inputText_winspace (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles empty - handles not created until after all CreateFcns called
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObj ect, 'BackgroundColor', 'white');
end
function inputText_TL_Callback(hObject, eventdata, handles)
% hObject handle to inputText_TL (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
% Hints: get(hObject,'String') returns contents of inputText_TL as text
% str2double(get(hObject,'String')) returns contents of inputText_TL as a double
%Tol=str2num(get(hObject,'String'));
%checks to see if input is empty. If so, default inputl_editText to zero
if(isempty(Tol))
set(hObj ect,' String',' 0')
end
guidata(hObject,handles);
% — Executes during object creation, after setting all properties.
function inputText_TL_CreateFcn(hObject, eventdata, handles)
% hObject handle to inputText_TL (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles empty - handles not created until after all CreateFcns called
-------
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObject, 'BackgroundColor', 'white');
end
%Calculates and plots FI different tightening levels
% — Executes on button press in FishTL_pushbutton.
function FishTL_pushbutton_Callback(hObject, eventdata, handles)
% hObject handle to FishTL_pushbutton (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
YMat=[];FMat=[];
mainFigureHandle=Main_Fisher_data_proc; %stores the figure handle of main GUI here
mainGUIdata = guidata(mainFigureHandle);
%Pulling data, time, AvgFish, and midyear from Main_Fisher_data_proc m
t=mainGUIdata.t;
data=mainGUIdata.data;
AvgFish=mainGUIdata.AvgFish;
midyear=mainGUIdata.Year;
PO_TL=mainGUIdata.PO_TL;
SOST=mainGUIdata. sost;
hwin=mainGUIdata.hwin;
winspace=mainGUIdata.winspace;
%Studying Pervasiveness and Intensity from Main_Fisher data_proc parameter
%settings
fori=l:10
Tol=100-(i-l)*10;
[FI, midt_win, t_win]=GFisher(t,data,SOST,hwin,winspace,Tol,0);
FMat(:,i)=FF;
YMat(: ,i)=midt_win;
end
axes(handles.axes3);plot(YMat(:,l),FMat(:,l),YMat(:,2),FM
5),FMat(:,5));
% Create legend
legendl = legend(handles.axes3,TL=100%\'TL=90%\'TL=80%','TL=70%','TL=60%');
set(legendl,'YColor',[l 1 l],'XColor',[l 1 1],...
'Location', 'NorthEast'
TontSize',8,...
TontName',' Arial');
-------
hold
plot(midyear, AvgFish, ' -gd' , 'Line Width' ,2);
title('Pervasiveness and Intensity (100%-60%)');
axis([YMat(l,l)YMat(size(YMat,l)) 0 10]); xlabel('Year'); ylabel('Fisher information');
hold off
axes(handles.axes4);plot(YMat(:,6),FMat(:,6),YMat(:J)JMat(:J),YMat(:,8),FMat(:,8),YMat(:,9)JMat(:,9),YMat(:,
10),FMat(:,10))
% Create legend
legend2 = legend(handles.axes4,'TL=50%','TL=40%','TL=30%','TL=20%','TL=10%');
set(legend2,'YColor',[l 1 l],'XColor',[l 1 1],...
'Location' , 'NorthEast' ,...
TontSize',8,...
TontName','Arial');
hold
plot(midy ear, AvgFish, ' -gd' , 'Line Width' ,2);
title('Pervasiveness and Intensity (50%-0%)');
axis([YMat(l,l)YMat(size(YMat,l)) 0 10]); xlabel('Year'); ylabel('Fisher information');
hold off
PVL=100-PO_TL;
set(handles.edit_POTL, ' String' ,num2str(PO_TL));
set(handles.edit_Pervasiveness, ' String',num2str(PVL))
%Save result of User Analysis
%dlmwrite('Results.xls',[hwin,winspace], '-append');
%dlmwrite('Results.xls',[YMat(:,l) AFish'], '-append', 'delimiter', '\t', 'precision', '%4.4f');
%dlmwrite('Results.xls', [midyear' AvgFish'], '-append', 'delimiter', '\t', 'precision', '%4.4f');
warning off
guidata(hObject, handles); %updates the handles
function edit_Pervasiveness_Callback(hObject, eventdata, handles)
% hObject handle to edit_Pervasiveness (see GCBO)
% eventdata reserved - to be denned in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
% Hints: get(hObject,'String') returns contents of edit_Pervasiveness as text
% str2double(get(hObject,'String')) returns contents of edit_Pervasiveness as a double
% — Executes during object creation, after setting all properties.
function edit_Pervasiveness_CreateFcn(hObject, eventdata, handles)
% hObject handle to edit_Pervasiveness (see GCBO)
% eventdata reserved - to be denned in a future version of MATLAB
% handles empty - handles not created until after all CreateFcns called
-------
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObject, 'BackgroundColor', 'white');
end
% — Executes on button press in RESET_Pushbutton.
function RESET_Pushbutton_Callback(hObject, eventdata, handles)
% hObject handle to RESET_Pushbutton (see GCBO)
% eventdata reserved - to be defined in a future version of MATLAB
% handles structure with handles and user data (see GUIDATA)
cla(handles.axes3,'reset');
cla(handles.axes4,'reset');
set(handles.inputText_TL,'String',100)
set(handles.inputText_hwin,'String',10)
set(handles.inputText_winspace,'String',5)
set(handles.edit_Pervasiveness,'String',0);
set(handles.edit_POTL,' String',0);
%handles.buttonCounter = 0; %reset buttoncounter
guidata(hObject, handles); %updates the handles
function edit_POTL_Callback(hObject, eventdata, handles)
% hObject handle to edit_POTL (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
% Hints: get(hObject,'String') returns contents of edit_POTL as text
% str2double(get(hObject,'String')) returns contents of edit_POTL as a double
% — Executes during object creation, after setting all properties.
function edit_POTL_CreateFcn(hObject, eventdata, handles)
% hObject handle to edit_POTL (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles empty - handles not created until after all CreateFcns called
% Hint: edit controls usually have a white background on Windows.
% See ISPC and COMPUTER.
if ispc && isequal(get(hObject,'BackgroundColor'), get(0,'defaultUicontrolBackgroundColor'))
set(hObject, 'BackgroundColor', 'white');
end
% — Executes on button press in Help_pushbutton.
function Help_pushbutton_Callback(hObject, eventdata, handles)
% hObject handle to Help_pushbutton (see GCBO)
% eventdata reserved - to be defined in a future version of MATL AB
% handles structure with handles and user data (see GUIDATA)
HelpPath = which('herpAdvGUI2.html');
web(HelpPath);
guidata(hObject, handles);
-------
rouii(l2 m : Floating Point Correction
function z = round2(x,y)
%ROUND2 rounds number to nearest multiple of arbitrary precision.
%Z = ROUND2(X,Y) rounds X to nearest multiple of Y.
%Floating point work around designed by Robert Bemis 15 Dec 2003 and updated
%(29 Jun 2006): http://www.mathworks.com/matlabcentral/fileexcnange/4261
%% defensive programming
error(nargchk(2,2,nargin))
ror(nargoutchk(0, l,nargout))
if prod(size(y))>l
error('n must be scalar')
end
z = round(x/y)*y;
CloseGULm
function closeGUI
selection = questdlg('Do you want to close the GUI?',...
'Close Request Function',...
'Yes','No','Yes');
switch selection.
case 'Yes'.
delete(gcf)
case 'No'
return
end
-------
7.0
and
7.1 Introduction
In this report, we have presented the results of a five-year
project that estimated four metrics of sustainability for
the San Luis Basin (SLB). The SLB is an agricultural
region in south-central Colorado, adjacent to New
Mexico, with a large portion of land publicly owned.
It is a relatively uncomplicated region that has not
experienced a substantial amount of change during the
26 years we examined. This lack of change provided an
opportunity to develop a methodology for calculating
sustainability metrics and to examine a large geographic
region with relatively little diversity in land uses (i.e..
there are few major industries or land uses to overly
complicate the calculation of metrics) and economic
drivers. For example, the San Luis Valley Development
Resources Group (SLVDRG; 2007) finds the top three
land uses in 2002 were: I) rangcland (43.6%); 2)
forest land (39.6%); and 3) agricultural land (11.8%).
Urban land is only 0.2% of the area in the six county
Valley.11 Because individual metrics do not capture all
aspects of a system, multiple metrics were estimated
to examine the SLB's movement toward or away from
sustainability. We assembled a detailed dataset that
met the three research objectives presented in Chapter
1. We successfully estimated the metrics, even though
data for some of the variables were not available at
the county level. In the previous four chapters, we
described the individual methodologies and presented
the metric results. Those chapters will enable the reader
to understand each metric in terms of what they capture
and the meaning of the results. Even though we chose
a less complicated region, we still faced a number of
hurdles while developing the methodology. This chapter
examines the sustainability metrics together, using the
results of the previous four chapters, to decipher whether
the SLB is moving toward or away from sustainability.
It describes the strengths and limitations of the overall
approach (rather than the strengths and limitations of
the individual metric approaches, which can be found
in Chapters 3-6). We present future research directions
generated from this research project and review our
research goal.
11 The San Luis Valley Development Resources Group defines the Val-
ley as the following six counties: Alamosa, Conejos, Costilla, Mineral,
Rio Grande, and Saguache. This differs from our definition of the SLB
by not including Hinsdale County.
7.2 Summary of Calculations
Based on the results of Ecological Footprint Analysis
(EFA), we found that even though there was a surplus
in bioproductive land, the ecological remainder was
decreasing due to population growth in the SLB (Chapter
3). If the trend continues, the ecological remainder
(a reserve in this region) will eventually become an
ecological deficit. Hence, we concluded that the SLB
is moving away from sustainability based on the EFA.
The estimation of Green Net Regional Product (GNRP)
provides a one-sided test of sustainability and reveals
when a system is not sustainable (see Chapter 4). The
results for GNRP provided no definitive evidence
that the system was moving away from sustainability.
Emergy Analysis (Em A) used the fraction of renewable
emergy used to total emergy used as a metric for
sustainability, but it was limited by data availability
(Chapter 5). Unlike the other metrics, a complete EmA
was estimated only for the 11 -year period of 1995-2005.
Overall, there was a general decrease in the fraction
of renewable emergy used to total emergy used. Early
in this period, EmA suggested the SLB was moving
away from sustainability. More recently, there was an
increase in the fraction of renewable emergy used to
total emergy used, indicating the SLB has been moving
toward sustainability since 2003. Finally, the results
of the Fisher information (FI) assessment revealed the
system exhibited no regime shifts for the period of the
study (Chapter 6). Although the system exhibited small
changes in dynamic order during the study period, it
was relatively stable. Further, with the slight decreasing
trend in dynamic order at the end of the study period,
we noted the SLB may be starting to move away from
sustainability. These results are summarized in Table
7.1.
7.3 for Luis
In Chapter 1. we stated that for the SLB to be on a
sustainable path, the following criteria must be satisfied:
1) ecological balance (i.e., remainder) is increasing (or
the ecological deficit is decreasing) with time and 2) the
ratio of renewable emergy used to total emergy used is
trending to one. FI and GNRP are one-sided tests and
cannot technically show if the system is moving toward
sustainability. Criteria for these one-sided tests indicate
if the region is moving away from sustainability when:
3) FI is steadily decreasing with time (Karunanithi et
al. 2008); and 4) GNRP is decreasing with time (i.e..
-------
Pezzey et al. 2006). For this study, we define a system
to be moving away from sustainability if ecological
balance. GNRP (or human welfare), the ratio of
renewable emergy used to total emergy used, or FI are
declining over time. Because each metric captures
different aspects of a system, and all aspects are vital to
the maintenance of a system, losing any one component
would result in problems with the system and, we
would consider the system to be moving away from
sustainability.
Interpretation of these results could be examined
statistically (e.g., Baslianoni et al. 2008, Shmelev and
Rodrigucz-Labajos 2009) or graphically (e.g., Hanlcy ct
al. 1999, Lammers et al. 2008, Pezzey et al. 2006). In
this case, given the difference in the number of years
calculated, interpreting the results statistically is not
likely to provide useful conclusions (although FI uses
data for the entire 26-year period, it produces only six
data points). Furthermore, we cannot satisfactorily
account for variation in each of the variables used in
any of the metrics. Hence, we lack statistical analyses
for trends and uncertainty. To explore further, we
graphed the results using a similar scale. The ecological
remainder showed a downward trend whereas GNRP
was increasing from 1980-2005 (Fig. 7.1). First, we can
examine how EFA and GNRP might relate because these
metrics were estimated for the entire 26-year period. It
is not surprising that GNRP has an inverse relationship
with EFA (as represented by ecological remainder)
because increased economic activity often accompanies
a corresponding increase in the human impact on the
environment or resource use. Although GNRP attempts
to account for environmental degradation, the effect
on human welfare might be mitigated by increases in
physical capital. The indirect relationship between
GNRP and EFA is consistent with Nourry (2008) who
examined the sustainability of France. Other studies
have found little relationship between GNRP and
EFA (e.g., Hanley et al. 1999, Nourry 2008). This is
consistent with our calculation of EFA and GNRP after
1994.
When we only look at 1995-2005 (11 years used for
EmA), we find that the fraction of renewable emergy
used to total emergy used was declining early in the time
period and began to increase again in the last couple of
years. Ecological Balance continued to decline during
these 11 years. Ecological Balance may be positively
correlated to total emergy used. In addition, total
emergy used per person, is decreasing slightly (Fig.
5.4). Ecological footprint (ef) and total emergy used per
person are supposed to lead to comparable results, but
similar to Bastianoni et al. (2008), there appears to be no
similar trend.
FI is difficult to include in this conversation on a year-
by-year basis because the window settings established
to capture the dynamic behavior of the system does not
result in a FI value for each year. Further, as noted in the
Sustainable Regimes Hypothesis, the method requires
that a mean FI () be computed to capture the trends
in the dynamic order. As such, only six values are
reported for the 26 year period. From the analysis, we
found the system FI increased slightly until it peaked
during the period centering 1995, and decreased slightly
thereafter. Although there were changes in the FI over
time, it was relatively steady and there was no indication
of a regime shift for the period of the study. Upon
examining the relationship between the for the
overall system and the variables by category, we found
the peak in system's during the period centering
1995 occurred at the same time as for the energy
and environmental categories. In addition, we found the
system had a negative correlation with the demographic
category. However, the overall system, consumption
(food and forest), land use, and agricultural production
categories indicated movement away from sustainability
by the end of the study period.
Interpreting the results holistically, we find evidence that
the SLB may be moving away from sustainability. All
four metrics are capturing different aspects or answering
different sustainability questions (e.g., Hanley et al.
1999), and EFA and EmA indicate movement away
from sustainability. EFA and EmA helped us reach this
conclusion based on the criteria listed above and the
graphs just described (Fig 7.1). If, as Neumayer (2010)
points out, weak suslainabilily is an important step in
the right direction, then, according to GNRP and FI, the
SLB was moving in the right direction. According to the
results of EFA and EmA. the region does not meet the
requirements of strong sustainability and was moving
in the wrong direction. If we correctly identified each
metric as strong and weak measures of sustainability, it
is interesting to realize the weak measures indicate the
region is stable or moving toward sustainability, whereas
the strong measures show the region is moving away
from sustainability. These results support Neumayer's
(2010) idea that, although not a long-term solution.
weak sustainability may be a step in the right direction.
However, we recognize the estimates are not perfect
and have identified the limitations of each metric in
their respective chapter. The next sections focus on the
strengths and limitations we identified using the SLB
project methodology.
-------
7.4 of
Knowing if a region is trending toward or away from
sustainabiiity is quite useful and interest in this type of
information is growing (e.g., Graymore et al. 2008).
We believe that our methodology -our first attempt-
adequately characterizes and evaluates the movement
toward or away from sustainabiiity of a regional
system. The methodology suggests the multidisciplinary
approach adequately assesses whether a region is
moving toward or away from sustainabiiity. Whereas
no methodology measures every aspect of the system.
we argue that each metric identifies changes to major
components of the system and that together these metrics
adequately represent the complexity of a regional
system.
The metrics are data intensive and the data requirements
are often prohibitive and lead to difficulty in calculating
and interpreting the results. However, we were able to
overcome several of the shortcomings by collecting and
adapting data for many of the variables at count}'', state,
and national levels during much of the 26-year period.
thereby successfully characterizing the SLB (discussed
in previous sections). We were tasked with developing
a methodology that used readily accessible data to
estimate four metrics. We believe we were successful
in completing this task. All data used in this approach
could be found on the World Wide Web and through
personal communication with experts and stakeholders
in the SLB. Table 2.1 summarizes data used this project.
It identifies which data were used and by which metrics,
the source of the data, and the number of years the data
were available. More detailed tables on data and their
sources are located in Chapters 3-6 for each individual
metric. The majority of data were available from Federal
government agencies including the Department of
Commerce, Bureau of Economic Analysis; Department
of Energy, Energy Information Administration; and the
US Department of Agriculture including the Agricultural
Research Service, the Forest Sendee, and the National
Agricultural Statistics Survey. Data from Global
Insight, Inc., used to compute Emergy, although readily
accessible, were not free.
Another strength of this methodology is the use of
multiple metrics to cover multiple attributes of the
environmental system including ecological, economic,
energetic, and system order. Sustainabiiity is not
limited to one attribute of a region and multiple metrics
are necessary to cover the three pillars (economic,
environmental, and social aspects) of sustainabiiity
(e.g., Graymore et al. 2008, Hanley et al. 1999).
Because we used a multidisciplinary approach to
estimate the metrics, we were able to take advantage
of the economies of scale (i.e., specialization) to
share information, knowledge, and data to support the
calculation of the four metrics. For example, as data
for wind erosion were being collected and analyzed,
the individuals calculating GNRP were able to explain
and share the process for the other metric groups
that used wind erosion (rather than estimating it on
their own). Once data were available for one metric,
they were available for all the metrics. Assigning the
collection of specific variables to individuals provided
an approach that limited contacting various experts,
thereby saving time (and money) of the researchers and
entities that possessed data. This sharing of data and
assigning responsibilities for collecting data reduced the
duplication of efforts and streamlined data collection
activities.
Another objective was to determine if data were
available for a time-series analysis, thereby examining
trends. Our initial goal was to calculate metrics and
examine a regional system from 1980 to 2005. This
required 26 years of data for numerous variables. EmA
was limited by data suggesting results are truly only
available from 1995-2005. Because data were available
for a number of variables over a number of years, we
successfully calculated four metrics and were able to
visually test for trends. This indicates that useful data
in electronic format can be easily obtained. Alone
this finding is a valuable conclusion to our study and
suggests that other useful metrics may be estimated
with relatively little direct costs to local and regional
governments.
Finally, one reason for the apparent success of this
project in addressing the multidisciphnary nature of
sustainabiiity was including stakeholders in the process
of identifying components of this system that were
responsible for the trends of the metrics through time.
We recognize that future regional analyses should
include stakeholders in the process from the beginning,
in order to address directly aspects of sustainabiiity that
stakeholders identify as important. It helps identify
appropriate metrics that can quantify those aspects which
are most useful to interested parties (i.e., the product
is wanted by the stakeholders). Getting participation
from individuals who were interested in the work and
were potential users of the information proved quite
useful in developing and improving the methodology' in
this study. For example, interacting with stakeholders
early on assured that results from the metric calculations
coincided with their observations of the SLB.
-------
7.5 of
The majority of the limitations of this methodology
relate to data needs, data availability, and data
manipulation. Because many of the data needed for
these metrics typically are not collected at the county
level, we needed to rescale or convert state or US data
to the county or regional level. Additional difficulties
included measuring trends and making the methodology
transferable to other regions besides the SLB. We
discuss these in turn.
One of the objectives of this research was to develop a
methodology that used readily accessible data. Although
this objective was met, not all of the data were free.
Global Insight. Inc. data were needed to estimate some
components of EmA. Import and export data, down to
transportation type and quantities of different goods,
are desirable for EmA, but difficult to estimate without
detailed data. Imports and exports to a system can be
estimated by alternative means with some sacrifice
in accuracy (Odum 1996). However, the cost of the
Global Insight, Inc. data («$20K for 11 years) may
preclude other regions from calculating this metric
with the accuracy obtained in this study. It should be
noted the import and export data were limited by the
years available (1995-2005; see Table 2.1). Hence, the
difficulty we had examining the results for the 26-year
research period. Future research could compare the
resulting trends from excluding these type data, thereby
eliminating the added expense.
We took a number of steps to overcome the issue of data
intensive calculations required for these suslainability
metrics. In addition, we addressed the lack of region-
specific data. Simplifying assumptions were needed
in order to estimate the metrics, but they are common
for these types of calculations. Note the assumptions
become limitations if they somehow misrepresent
the region of study. In this study, we assumed the
SLB could be represented by state and/or US data
with the appropriate rescaling (i.e., estimated on a per
capita basis), and we tested some of these rcscalings
(e.g.. Appendix 3-A). We attempted to confirm the
assumptions were appropriate through discussions with
stakeholders and by reviewing other data sources. For
example, to estimate the Net Regional Product (NRP)
for the SLB, initially we used Colorado data, but local
stakeholders felt that both Colorado and New Mexico
influenced the economy. Using this information, we
averaged the NRP calculation for the SLB using both
Colorado data and New Mexico data. In cases of
missing data, interpolation was necessary to ensure a
value could be computed for every year and because
complete data were needed to proceed with calculations
of EFA, EmA, and GNRP. Linear interpolation is a
weakness if it was not appropriate. There are a number
of examples described in Chapters 3-6 where we
manipulated some state or US data in order to proceed
with the calculations. The weakness here lies in whether
these manipulations yielded adequately representative
data for the SLB. It is probable that some manipulations
better represented the SLB than others.
The team initially thought it could estimate confidence
intervals and analyze trends for each of the metrics
in order to understand uncertainty in the data. The
proposed approach was to use bootstrapping, but we
found it difficult to proceed given the state of the data
(i.e., missing years, interpolated values, and lack of
variation). From the bootstrap distribution, an estimate
of bias, standard error, and confidence intervals are
produced (Dixon 2001), thereby permitting hypothesis
testing. Confidence intervals were difficult to estimate
because there is some amount of error associated with
data collection, but this error is rarely reported with the
datasct. Furthermore, the metrics arc a manipulated
quantity rather than the individual variables. This
manipulation has an unknown impact on the error.
Additionally, each variable in all metrics will have some
unknown variation for each year, especially when we
assume that a parameter is constant across years. Using
a bootstrap approach to resample the metric would
miss this type of uncertainty. Therefore, we were not
confident that an uncertainty analysis would yield useful
results given the project methodology and our estimates
could be criticized for lacking a measure of uncertainty.
Although trend analysis is quite common in
environmental studies and market studies (e.g., Burn
and Elnur 2002, Hirsch et al. 1982, Hirsch et al.
1991), we were only able to find a statistic of trends
in one paper examining sustainability metrics (Sato
and Samreth 2008), but it lacked references to address
its legitimacy. Initially we considered using Mann-
Kendall to test for trends. Mann-Kendall test is a
non-parametric test that concentrates on the relative
value (i.e., positive or negative) of the observations
rather than the magnitude of the observations and is
frequently used for trend detection (Burn and Elnur
2002, Hirsch etal. 1982). Although Mann-Kendall
is non-parametric and, therefore, free of assumptions
about the distribution of the data, it does assume data
are not serially correlated. To calculate each metric
on a yearly basis, we did a linear interpolation to fill
variables for missing years. This interpolation resulted
in a strong autocorrelation between points. Thus, we
were not confident in the resulting/>-value from the
test (i.e., p-value overestimated the significance). The
-------
majority of published studies on sustainability metrics
appear to examine trends visually through graphs of the
metric calculation over time (e.g., Chen and Chen 2007,
Haberl et al. 2001, Hanley et al. 1999, Lammers et al.
2008, Pezzey et al. 2006,'wackernagel et al. 2004b)
which was the approach we ultimately used for this
project. Nonetheless, we recognize the methodological
improvement that would result if a measure of
uncertainty could be included and the benefit of testing
for statistical significance in apparent trends.
Multiple metrics were estimated to cover all aspects of
sustainability. However, in some instances, different
metrics used several of the same variables, suggesting
a potential correspondence. For example, wind erosion
is used in GNRP, EmA, and FI (e.g., see Table 2.1). A
relationship between metrics may not be a weakness per
se, but it limits any correlation calculations. Rather than
being independent metrics, we have estimated metrics
that were already linked and duplicated certain aspects
of the system, but we do not think this is a problem. For
example, although population drives the EFA calculation,
it plays a minor role in GNRP.
The methodology presented in this report was general in
structure and was useful in evaluating the sustainability
of the SLB. However, certain approaches, variables,
data, and calculations can only represent the SLB
specifically. Accordingly, other regions will require in
depth study prior to estimating the metrics. Because
different regions have different issues in regards to their
sustainability, the specific methodology presented here
is not simply a generic recipe for sustainability metric
calculations. However, it is useful in a general sense.
For example, GNRP used soil erosion groundwater
level, and CO2 emissions to document the depreciation of
natural capital. A new region may have concerns related
to mining, forestry, and surface water quality, requiring
different data and calculations. A different region might
not be able to adopt the approach used to estimate soil
erosion, but it could examine how it was calculated and
apply that to its own issues. The approach will have to
be adapted to capture the new depreciation of natural
capital and the literature and methodology presented here
will provide a certain amount of direction. On the other
hand, there is complete transportability- in the ability to
assess the dynamic order of other regional systems as FI
may be computed for any system given the availability
of pertinent data characterizing the state of the system.
Stakeholders expressed an interest in examining how
the metrics change using different scenarios. They saw
a need for scenario or futures analysis in part because
of a delay between the time that an event occurs and the
time when data on the event are released to the public.
For example, CO, emissions data only become available
three years after the year when the emissions occurred.
In other words, the calculations are always three years
behind the present year making the metric calculations
somewhat limited in their ability to provide decision
support for the present. Some stakeholders suggested
that the lag could be managed with an alternative futures
component to the methodology. However, an objective
of this report was to describe an approach to examine
sustainability' over time, and not an approach to look into
alternative futures. Kok (2009) presents a framework
for considering the interrelationships between the types
of variables considered in our metrics. An alternative
futures approach would require significant modeling
efforts that were not part of the project's objectives. The
information that would be created through this modeling
would be very useful for the SLB or any regional
system. Regional plans could be evaluated on the basis
of their sustainability. Development scenarios could be
created and then modeled using the metric calculations
and used to see how the metric changes under different
scenarios. The results could provide local decision
makers information that results in a movement
toward sustainability'. Our results provide a different
decision support tool that requires more of an adaptive
management strategy. The metrics reveal a region's
existing movement toward or away from sustainability
and identify components of the system that require
attention. Continued monitoring of the system and
quantifying the resulting trend can influence additional
management decisions.
Our final concern relates to our goal of producing a
straightforward, relatively inexpensive methodology
that is simple to use and interpret. We cannot state
with confidence that we met Hie goal because it would
be difficult to test any one of these components. Some
anecdotal evidence would suggest that the metrics are
not as simple to use or interpret as we initially projected.
Calculating FI was automated by developing a graphical
user interface and spreadsheets were created for the
remaining three metrics. However, understanding
each metric and what variables should be included for
different regions may prove somewhat intimidating for
land managers and decision makers. Although we have
scientific criteria for interpreting the metrics, it may
not provide enough insight into what the results really
mean. The four metrics are based on sound science, but
the weakness may occur with the ability to succinctly
synthesize what they tell decision makers about a region
like the SLB. Therefore, we suggest this as one of the
recommendations for future research activities below.
-------
7.6 for Future
on
Metrics
We offer the following recommendations for future
research based on the results of our study:
I) Examine previous management decisions in the
SLB to see if and how the metrics changed. For
example, the Great Sand Dunes National Park and
Preserve was created in 2000 and increased the
National Monument land area by almost four times
(NFS 2007). This recommendation would help
to determine if any of the changes in the metrics
could be linked back to the management decision
to increase the public land area.
2) Continue calculating the metrics in the SLB for
subsequent years to examine how sensitive the
metrics are to management decisions.
3) Use stakeholder meetings to determine major
sustainability issues in the region and better
link the metrics on those issues and concerns.
Determine if there are other established metrics
that may better assess sustainability of a region
by addressing known issues. This relationship
with stakeholders will help ensure the metrics are
capable of capturing changes pertinent to issues
of public concern. The meetings can be used to
gauge whether stakeholders and decision makers
can understand the results of the metrics. More
research can focus on simplifying the interpretation
of results for various groups of people: managers,
regulators, landowners, educators, policy makers.
etc.
4) Develop trend analyses and/or approaches for
estimating confidence intervals for individual
metrics. Although a trend may appear obvious,
recognizing and identifying the variation in the
data and resulting metric will better identify
significant trends. One example might be using
econometric methods (e.g., Greene 1993, Hay et
al. 2002) or sensitivity analyses.
5) Develop models of alternative future scenarios and
estimate multiple metrics (Baker et al. 2004, Kok
and van Delden 2009, Yeh and Chu 2004) that are
sensitive to relevant changes in a system or region.
For example, how would the installation of solar
farms affect the metrics?
6) Test these multiple metrics in other regions and
determine if the metrics are indeed capturing the
trend in sustainability of regional systems.
7) Examine the correlation among alternative metrics
to determine whether certain metrics could be
dropped from analysis (i.e.. one metric moves in
a similar or opposite pattern to another over time)
in order to reduce the resources needed (e.g.,
Bastianoni et al. 2008 used Principal Components
Analysis).
8) Consider (lie use of other scientific approaches,
rather than limiting analysis to visual examination
of graphs, for deciphering the holistic results of
multiple metrics, such as multicriteria methods
(e.g., Shmelev and Rodriguez-Labajos 2009);
data envelopment analysis (e.g., Dcspotis 2005,
Kortelainen 2008); mathematical programming
(e.g.. Zhou et al. 2007); or other integrated
sustainability metrics (e.g., Mayer 2008).
7.7 Conclusions
Although we met the three objectives of this research
project, at this time we cannot definitely state we met
the overall goal of the project. Recall from Chapter 1,
the goal was to produce a straightforward, relatively
inexpensive methodology that was simple to use
and interpret. We believe the method is relatively
inexpensive because we have identified the majority
of data sources for these types of calculations leading
to reduced time and resources necessary for others
to compute the metrics. How straightforward the
methodology is would be based on anecdotal evidence
and discussions with stakeholders in the region. We
do think we partially achieved the goal for some of the
metrics, but are utilizing feedback from end users to
fine-tune the descriptions and explanations to enable
better understanding by decision makers. We recognize
it may require an intermediary to compute some of the
metrics, but we want decision makers to understand
the methodology sufficiently so they see the benefit
and importance of calculating the metrics and they
understand what the results mean in order to aid in future
decision-making.
We think we made progress in this regard by producing
a methodology that successfully characterizes whether
a region is moving toward or away from sustainability.
However, the next major accomplishment will be
translating the information for decision makers. Applying
one or more of the eight recommendations will help us
do just that.
-------
7.8
Baker, J.P., Hulse, D.W., Gregory, S.V., White, D., Van
Sickle, I, Berger, P.A., Dole, D., Schumaker, N.H.
2004. Alternative Futures for the Willamette River
Basin, Oregon. Ecological Applications 14, 313-324.
Bastianoni, S., Pulsclli, R, Focardi, S., Tiezzi,
E., Gramatica. P. 2008. Correlations and
complementarities in data and methods through
principal components analysis (PCA) applied to
the results of the SPIn-Eco project. Journal of
Environmental Management 86, 419-426.
Burn, D. H., Elnur, M.A. 2002. Detecting hydrologic
trends and variability7. Journal of Hydrology 255. 107-
122.
Chen, B., Chen, G.Q. 2006. Ecological footprint
accounting based on emergy--a case study of the
Chinese society. Ecological Modelling 198. 101-114.
Despotis, D.K. 2005. A reassessment of the human
development index via data envelopment analysis.
Journal of the Operational Research Society7 56. 969-
980.
Dixon, P.M. 2001. The bootstrap and the jackknife:
describing (lie precision of ecological indices,
in: Scheincr, S.M., Gurevitch, J. (Eds.). Design
and Analysis of Ecological Experiments. Oxford
University Press, New York. pp. 267-288.
Graymore, M., Sipe. N., Rickson, R. 2008. Regional
sustainability: How useful are current tools of
sustainability assessment at the regional scale?
Ecological Economics 67, 362-372.
Greene. W. 1993. Econometric Analysis. Prentice Hall,
New Jersey.
Haberl, H., Erb, K., Krausmann, F. 2001. How to
calculate and interpret ecological footprints for long
periods of time: the case of Austria 1926-1995.
Ecological Economics 38, 25-45.
Hanley, N., Moffatt, I., Faichney, R., Wilson, M. 1999.
Measuring sustainability: A time scries of alternative
indicators for Scotland. Ecological Economics 28,
55-73.
Hirsch R.M., Alexander, R.B., Smith, R.A. 1991.
Selection of Methods for the Detection and Estimation
of Trends in Water Quality. Water Resources Research
27, 803-813.
Hirsch, R. M., Slack, J.R., Smith, R.A. 1982.
Techniques of trend analysis for monthly water quality
data. Water Resources Research 18, 107-121.
Kok, K. 2009. The potential of Fuzzy Cognitive Maps
for semi-quantitative scenario development, with an
example from Brazil. Global Environmental Change
19, 122-133.
Kok, K., van Delden, H. 2009. Combining two
approaches of integrated scenario development to
combat desertification in the Guadalentin watershed.
Spain. Environment and Planning B: Planning and
Design 36, 49-66.
Kortelainen. M. 2008. Dynamic environmental
performance analysis: a Malmquist index approach.
Ecological Economics 64, 701-715.
Lammers, A.. Moles, R., Walsh, C, Huijbregts, M.A.J.
2008. Ireland's footprint: a time series for 1983-2001.
Land Use Policy 25, 53-58.
Mayer, A. 2008. Strengths and weaknesses of common
sustainability indices for multidimensional systems.
Environment International 34. 277-291.
NPS (National Park Service). 2007. Final Summary
of General Management Plan/Wilderness Study/
Environmental Impact Statement. US Department of
the Interior. Mosca, CO.
Nouny, M. 2008. Measuring sustainable development:
Some empirical evidence for France from eight
alternative indicators. Ecological Economics 67, 441-
456.
Odum, H.T. 1996. Environmental Accounting: Emcrgy
and Environmental Decision Making. John Wiley &
Sons. New York, 370 pp.
Pezzey, J., Hanley, N., Turner, K., Tinch, D. 2006.
Comparing augmented sustainability measures for
Scotland: Is there a mismatch? Ecological Economics
57, 60-74.
San Luis Valley Development Resources Group. 2007.
Comprehensive economic development strategy.
Accessed at: http://www.slvdrg.org/ceds.php on June
17, 2009.
Sato, M., Samreth, S. 2008. Assessing sustainable
development by genuine saving indicator from
multidimensional perspectives. Unpublished
manuscript. Munich Personal RePEc Archive MPRA
Paper No. 9996, posted 13 August 2008.
Shmelev, S., Rodriguez-Labajos, B. 2009. Dynamic
multidimensional assessment of sustainability at
the macro level: the case of Austria. Ecological
Economics 68. 2560-2573.
-------
Wackernagel M, Monfreda C, Erb K-H, Haberl H,
SchulzNB. 2004. Ecological footprint time series
of Austria, the Philippines, and South Korea for
1961-1999: comparing the conventional approach to
an 'actual land area' approach. Land Use Policy 21,
261-269.
Yell. S.C., Chu, Y.P. 2004. Economic-environmental
scenario simulations to measure sustainable
development in Taiwan. Advances in Architecture
Scries 18, 483-492.
Zhou, P., Ang, B.W., Poh, K.L. 2007. A mathematical
programming approach to constructing composite
indicators. Ecological Economics 62, 291-297.
Table 7.1™ Summary of four metrics calculated for the San Luis Basin, Colorado.
Metric
Ecological Footprint
Green Net Regional Product
Emergy
Fisher information
Paradigm
Strong sustainability
Weak sustainability
Strong sustainability
Weak sustainability
Summary
Movement away from sustainability, although gradually
during the 26-year period
No definitive evidence to suggest movement away from
susta inability-
Movement away from sustainability, although showing
improvement recently
Relative stability, slight movement away from sustainability
at the end of the study period
-------
Ecological Balance
I*q
GNRP
Fraction of Locally Renewable Emergy
Fisher Information
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
Figure 7.1 - Graphs of Ecological Balance, Green Net Regional Product (GNRP), Fraction of Locally Renewable
Emergy, and mean Fisher Information for 1980-2005.
-------
&EPA
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |