&EPA
United States
Environmental Protection
Agency
Policy, Planning,
And Evaluation
(2122)
EPA 230-R-95-003
May 1995
Climate Change And The
Boston Area Water Supply
,T3k ^
(r
Ilk
: ^ £~~~ * V **•«*
--»*
V
*^l%,, ^ *" if
.-^.- /,7., ' ^ ^i.|
.-•
^<^^
**1^
-------
-------
Climate Chang
and
• ' • !•
Boston Area Water Supply
By Paul Kirshen
and
Neil Fennessey
May 1995
The United States Environmental
Office of Policy, Planriin
A Report Prepared For:
Protection Agency
ing and Evaluation
Climate Change Division
EPA 230-R-95-003
-------
ACKNOWLEDGEMENTS
Neil Leary of the U.S. Environmental Protection Agency (EPA) and his
colleagues at the EPA are thanked for their extensive review comments. The authors also would
like to thank Joel Smith, formerly of the EPA, for his support on this effort. In addition, they
would like to thank Stephen Estes-Smargiassi, Daniel Nvule, and Joseph Smith of the
Massachusetts Water Resources Authority (MWRA) for their many efforts. The help of Eric
Anderson of the US National Weather Service in obtaining and implementing the rainfall/runoff
model is highly appreciated. Dennis Joseph of the National Center for Atmospheric Research
(NCAR) promptly and efficiently supplied meteorological and GCM data. Peter Eagleson and
Dara Entekhabi (both of the Massachusetts Institute of Technology) provided useful discussions
and references in regard to Ep and Et calculations. Harry Schwarz, David Major, and Thomas
Stevens were also very helpful. The authors also appreciate the comments of the reviewers. The
authors would also like to thank the following for their contributions; Bettina Burbank, Kathleen
O'Connor, Louis Diminico, Sharon Diminico, and Andrew Taylor. This work was performed
under contract for the U.S. Environmental Protection Agency, contract number 68-W2-0018.
ICF, Inc., the primary contractor with the EPA for this project, provided administrative and other
support.
-------
Table of Contents
Acknowledgements ........
List of Figures
List of Tables
Preface ........
Executive Summary ......
1: Introduction .......
2: Study Area
3: Safe Yield Model
4: Calibration and Verification of Runoff Model
5: Pot-
-------
-------
SI
S2
2.1
3.1
3.2
4.1
4.2
4.3
4.4a
4.4b
4.5
4.6a
4.6b
5.1
6.1
6.2
A.1
A.2
A.3
List of Figures
Boston Area Water Supply for Normal and 2xCO2 Climates
Sensitivity of Boston Area Water Supply to Precipitation at
Temperature Changes
Study Area
Ware River Monthly Transfer Volume
Ware River Monthly Transfer Days
Structure of Snow Model
Sacramento Soil Moisture Accounting Model
Ware River Verification
Ware River Entire Series (1950 - 64)
Ware River Entire Series (1965 - 79)
Connecticut River Verification
Connecticut River Entire Series (1950 - 64)
Connecticut River Entire Series (1965 - 79)
Comparison of Estimated Ep and Measured Pan Evaporatioi
GISS and GFDL Impacts upon Mean Monthly Flows of Wa
OSU and UKMO Impacts upon Mean Monthly Flows of W
Temperature Comparison
Windspeed Comparison
Cloud Cover Comparison
iv
xi
id
xii
14
17
. 18
23
24
25
26
27
28
29
30
i 38
re River 46
are River 47
88
89
90
-------
A.4
A.5
Relative Humidity Comparison 91
Precipitation Comparison , 92
-------
List of Tables
SI
1.1
1.2
4.1
5.1
6.1
6.2
7.1
7.2
7.3
7.4
7.5
7.6
7.7
A.1
A.2
A.3
A.4
A.5
A.6
A.7
Summary of Climate Scenarios "••••• x
Summary of Scenarios 9
'i
Summary of Results .....' 10
Comparison of Historic and Simulated Time Series for Present Climate 22
Monthly Forest and Surface Water Albedo ; 39
i
GCM Variables Perturbed j 44
I
Summary of Results 1 45
MWRA System Demands 1 56
Total MWRA System Safe Yield j 57
Additional Sources of Supply Under 2 x CO2 1 58
Financial Analysis - No Climate Change, e = -0.10 59
Financial Analysis - Climate Change, e = -0.10 ' 60
Financial Analysis - No Climate Change, e = -0.10, Reduced Costs 61
Financial Analysis - Climate Change, e = -0.10, Reduced Costs 62
'!
Reported GCM Temperatures ; 76
if
i
Reported GCM Windspeed j 77
Reported GCM Atmospheric Pressure • 78
Reported GCM Total Cloud Cover i 79
Reported GCM Specific Humidity or Mixing Ratio 1 80
Calculated GCM Relative Humidity •....'.. 81
j
Reported GCM Precipitation ,| 82
-------
A.8
A.9
A.10
A.ll
A.12
Observed Air Temperature 83
Observed Windspeed • 84
Observed Possible Sunshine 85
Observed Relative Humidity 86
Observed Precipitation , 8/7
VI1.
-------
PREFACE
Rising atmospheric concentrations of greenhouse gases threaten to warm the! earth and alter the
global climate system. Global climate change poses wide ranging risks to the environment,
human welfare, and human health. To advance our knowledge of the risks, the Climate Change
Division of the United States Environmental Protection Agency has conducted and supported a
number of studies to examine different categories of climate change effects. Climate Change and
Boston Area Water Supply reports results from one of the studies supported by our office.
,(
One source of risk posed by global climate change is the potential for climate change to alter the
availability of water. Water runoff into streams and reservoirs is controlled, in part, by the
amount of precipitation falling to earth and the amount of water returning to the atmosphere via
evaporation and transpiration. The geographic distribution, seasonal patterns;! and variability of
precipitation, evaporation, and transpiration will change as the global climate system changes.
Whether or not these changes will increase or decrease the available supply of water will vary by
location and season. Where water supplies are diminished, competition for this vital resource will
intensify and water quality may be degraded. Residential, commercial, industrial and agricultural
water users may face higher water costs and reduced water consumption. In addition, instream
uses of water such as the support offish populations and water recreation may also suffer.
I
In this study, Paul Kirshen and Neil Fennessey investigate the potential impacts of global
climate change on the municipal water supply of the Boston Metropolitan Area and potential
responses to supply changes. The case study finds that Boston area water supply is highly
sensitive to climatic change and that the costs of responses are potentially high for some
scenarios. The results provide useful insights regarding strategies for adapting to climate change
and suggest that climate change risks for municipal water supplies warrant concern and further
consideration. t
The study was performed for the Environmental Protection Agency under contract with ICF,
Incorporated. An earlier version of the report was reviewed by EPA staff and by experts from
outside the agency. Participants in the review process are thanked for their efforts to improve the
report. The final report is a report to the EPA and does not necessarily reflect the views or
policies of the EPA.
Neil Leary, Project Manager
Climate Change Division
Office of Policy, Plannin
U.S. Environmental Protection
Vlll
g and Evaluation
Agency
-------
-------
EXECUTIVE SUMMARY
Overview
In this study, the potential effects of global
climate change on the municipal water
supply of the Boston Metropolitan Area are
investigated. The Boston area receives 80
percent of its water from the Quabbin and
Wachusett Reservoirs, which are operated
by the Massachusetts Water Resources
Authority (MWRA). This source of water is
found to be highly sensitive to climatic
changes that are within the range of changes
that climate scientists believe could occur
during the coming century. Estimated
impacts on the amount of water that can be
reliably supplied by the MWRA range from
a loss of one-half to a gain of one-third for
the climate change scenarios evaluated in
the study.
Water supply changes of these magnitudes
have important implications for the need to
develop new sources of water supply and the
cost of water to water users. In the absence
of climate change effects, regional growth in
water demand is projected to exceed the
reliable supply of water from the existing
reservoir system and investments to expand
water supply will be required to satisfy the
growing demand. If climate change
decreases water runoff in the region, the
need to add to the capacity of the system
will be magnified. The result would be
higher costs to water users in the Boston
area. In one scenario, added capital costs
exceed $700 million by the latter half of the
next century.1
In contrast, if climate change increases water
runoff into the MWRA reservoirs, the need
for additional supplies and their costs will be
offset either partially or entirely. Capital cost
savings over coming decades could amount
to $15 million relative to a scenario of no
climate change. j
Methodology of the Study
i'
Estimates of the reliable water supply
available from the MWRA to the Boston
area are made for fifteen different climate
scenarios. One of the scenarios represents a
baseline climate which is presumed to
exclude the effects of global climate change.
It is based on local weather observations for
the 1950 to 1979 period. The other scenarios
represent changed climates for a world with
an enhanced greenhouse effect. Table SI
provides an overview of the climate
scenarios examined in the study.
F
Four of the changed climate scenarios are
derived from the simulations of Global
Circulation Models (GCMs) and represent
the possible effects of doiibling the
concentration of carbon dioxide in the
atmosphere from the preindustrial
concentration. The GCMs provide scenarios
of regional changes in temperatures,
precipitation, relative humidity, cloud cover,
solar radiation and wind speed that are
internally consistent with current
understanding of how the' global climate
system behaves.2 |
1 All cost estimates are reported in 1990 dollars.
2 The GCM scenarios used in this exercise project
warming that is near the upper end of the range of
warming that many climate scientists expect to occur
-------
Table S1: Summary of Climate Scenarios
Scenario
Baseline
GISS
GFDL
OSU
UKMO
A,+/-B%
Description
Historical observations for the
period 1950-1979.
2xC02 equilibrium climate
projection of the Goddard Institute
for Space Studies (GISS) General
Circulation Model, 1982.
2xC02 equilibrium climate
projection of the Geophysical Fluid
Dynamics Laboratory (GFDL)
General Circulation Model, 1988.
2xC02 equilibrium climate
projection of the Oregon State
University (OSU) General
Circulation Model, 1984/85.
2xCOa equilibrium climate
projection of the United Kingdom
Meteorological Office (UKMO)
General Circulation Model, 1986.
Sensitivity scenarios in which
temperature is increased A° C and
precpitation is increased or
decreased B%.
Global
Temp.
Change
(°C)
NA
+ 4.2
+ 4.0
+ 2.8
+ 5.2
NA
Regional
Temp.
Change
(°C)
NA
+ 3.7
+ 4.9
+ 3.1
+ 8.3
+2.0 to
+4.0
Regional
Precip.
Change
(%)
NA
- 1.6
- 7.6
+ 13.0
+ 23.0
-20 to
+20
The other climate scenarios are sensitivity
scenarios that impose specified changes in
temperatures and precipitation relative to
baseline climate. The scenarios incorporate
regional warming of 2° and 4° C with
precipitation changes of plus and minus 0,
10 and 20 percent. Other climate variables
are held constant hi these scenarios.3
by the end of the next century. Climate scientists are
not agreed on the probabilities to be assigned to any
particular projection within the expected range, nor
to the possibility that warming might fall outside the
expected range.
3 The 2° and 4° C warming of the sensitivity
scenarios reach roughly from the middle to the upper
end of the range of expected warming for the next
century. Warming of 2° and 4° C correspond to
warming of 3.6° and 7.2° F respectively.
Streamflows into the
Quabbin and Wachusett
reservoirs are estimated for
each climate scenario. The
estimates are made using
well known, commonly
applied hydrologic models
that have been calibrated
and validated for the local
watersheds. Given climatic
inputs, the hydrologic
models calculate rainfall,
snow, snowmelt, storage of
water in the soil,
evapotranspiration
(evaporation and
transpiration) of water to
the atmosphere, runoff, and
Streamflows into the
reservoirs.
Management of these
Streamflows to supply the
Boston area are modeled using the MWRA's
Safe Yield Model. Given the calculated
Streamflows from the hydrologic models, the
safe yield model calculates the amount of
water that can be supplied with a reliability
of 98.5 percent by the system to meet
Boston area water demands. This amount of
water is the safe yield of the system. The
calculation takes into account evaporation
losses from reservoirs, minimum flow and
flood control requirements, and the net '•
volume of water stored in reservoirs.
Impacts on Water Supply
The safe yield of the MWRA system for
baseline climate is calculated to be 306
million gallons per day (mgd). Together
with water yields from other non-MWRA
sources of water, the Boston area has
adequate water to meet its current water
-------
demands of roughly 360 mgd under baseline
climate conditions. Figure SI displays the
safe yield for baseline climate alongside the
calculated safe yields for GCM-based
climate scenarios.
In two of the GCM-based climate scenarios,
the GISS and GFDL scenarios, the local
climate becomes drier as precipitation
decreases and higher temperatures increase
water losses to evapotranspiration.
Consequently, streamflows decline and the
safe yield of water supply from the MWRA
declines. Safe yield is projected to decline
23 percent and 43 percent relative to the
baseline for the GISS and GFDL climate
scenarios respectively. Under these
conditions, water supply would be
insufficient to satisfy current water
consumption in the Boston area.
In contrast, streamflows and safe yield
increase in the two other GCM-based
climate scenarios, the OSU and UKMO
Boston Area Water Supply
for Baseline and 2xCO2 Climates
Baseline
DG1SS
H)GFDL
.OSU
UKMO
Figure Si: Estimated safe yield of water from Boston area reservoirs for baseline climate
and GCM projections of future climate for a doubling of carbon dioxide m the
atmosphere above the preindustrial concentration.
scenarios. In the OSU scenario, a 7 percent
increase in safe yield is Largely due to
increased local precipitation. In the UKMO
scenario, precipitation also increases but
another major factor increasing streamflow
and safe yield is increased relative humidity
in summer and fall. With higher relative
humidity, evapotranspiration losses are
reduced during these periods despite higher
temperatures. The projected result for the
UKMO scenario is a 38 percent increase in
safe yield. ;
Figure S2 displays the results for the
sensitivity scenarios. As expected, safe yield
of the MWRA reservoir system is found to
be negatively correlated, with warming and
positively correlated with precipitation.
j.
If there is no change in precipitation,
warming of 2° and 4° C! reduce safe yield by
9 and 18 percent respectively. If warming is
combined with decreases in local
precipitation, the losses; are magnified. In the
case of a 20 percent decline
in precipitation, safe yield
would be cut by half from
the normal safe yield of
306 nlgd.
If local precipitation
increaises., however, losses
from '[vanning would be
ameliorated and the net
effect could be an increase
in safe yield. A 10 perbent
increase in precipitation
would offset the effects of
4° C wanning almost
entirely. A 20 percent
increase in precipitation
would more than offset the
effects of 4° C warming
-------
and substantially increase safe yield.
Potential Responses and Costs
The above results indicate that
global climate change could
substantially affect the water supply
of the Boston area. If the water
supply available from the existing
system is diminished, water users
may be required to reduce their
water consumption and/or new
sources of water supply may need to
be developed. These impacts
represent but one of the risks posed
by climate change to human welfare.
Responding to the multiple risks
posed by climate change is a
challenge that must combine two
general strategies. The first is to
reduce the risk of damages by
limiting or slowing climate change.
For example, by conserving energy we can
reduce emissions of the greenhouse gases
that drive global climate change. The second
general strategy is to adapt so as to limit the
damages that may result from climatic
changes that are not averted. The study's
authors, with assistance from the MWRA
have examined potential adaptive responses
for the MWRA water supply system.
Adaptive responses are examined for a
single climate scenario, the GISS scenario.
The GISS scenario is selected for analysis
because it represents a case in which climate
change is calculated to have a negative
effect on local water supply and the authors
are interested in examining potential
responses to a supply reduction. Because of
the focus on a single scenario, the evaluation
of adaptive response options is merely
Sensitivity of Boston Area Water Supply to
Precipitation and Temperature Changes
Gg2o C Warming
O^oC Warming
-60
-20%
-10% o% 10%
Percentage Change Precipitation
20%
illustrative of the types of responses that
might be made. A more complete analysis
would need to explicitly take into account
the uncertainty of global and regional
climatic changes, the uncertainty of their
impacts, and the implications of these
uncertainties for water planning.
To evaluate potential responses to the GISS
climate scenario and their costs, the authors
first assess future water demand and supply
in the absence of climate change. Based
upon modest expected population and
economic growth, and an aggressive water
conservation program, demand is projected
to grow by only 5 percent from 1990 to
2050. Under normal climatic conditions and
current prices, future demand would exceed
the safe yield of the existing water supply
system by 11 mgd.
xii
-------
The MWRA has identified a number of local
water supply sources that could be
developed to supply this deficit under
normal climate. The estimated costs of the
new sources are $ 15 million in capital costs
and $ 1 million in annual operation and
maintenance costs.
In contrast, assuming that the GISS climate
projections for a doubling of carbon dioxide
in the atmosphere are realized by the year
2050, the authors estimate that the water
deficit in that year would be nearly 100
mgd.4 Working with MWRA staff, the
authors identify a number of responses that
would eliminate the projected deficit if
implemented. They include development of
local water sources, reactivation of an
existing but currently unused reservoir on
the Sudbury River, a water conservation
program for industrial, commercial and
institutional water users, and, finally,
diverting water from the Merrimack River to
Boston area water users. .
The total capital costs of these responses are
nearly $740 million and annual operation
and maintenance costs exceed $20 million.
Climate change, under the GISS scenario,
would raise costs by more than $700 million
for capital needs and $20 million for annual
operations and maintenance relative to the
scenario of no climate change. Adding these
costs to current water system costs would
raise a typical household's annual
4 The warming projected for the GISS scenario
exceeds the range of warming expected by most
climate scientists for the year 2050. The estimated
water deficit includes a 70 mgd reduction in the safe
yield of the MWRA reservoirs, a 10 mgd estimated
reduction in water supply from other non-MWRA
sources, and an estimated 4 percent increase in water
demand as a consequence of climate change.
expenditure on water and sewer services
$45, or 25 percent, hi the year 2050.
Conclusions i
'!
Whether or not these responses and their
costs are warranted is not yet clear. The
specifics of how global and regional climate
will change, and the time path these changes
will follow, are uncertain. As demonstrated
by the hydrologic analyses' of this study, the
uncertainties about climate imply a wide
range of potential impacts on the MWRA
water supply. As illustrated for the GISS
scenario, climate change may require
substantial additions to the water supply
capacity of the MWRA and/or substantial
water conservation efforts jn the region. The
GISS scenario, however, is but one of many
possible outcomes. ;
Further uncertainty about future water needs
is introduced by expected'cost increases for
Boston area water users that are unrelated to
climate change. Costs for hew sewage
treatment facilities may raise future
combined water and sewer rates by 50
percent. Water users are likely to respond to
these rate increases by reducing water
consumption. The study demonstrates that
future needs for new water supplies as a
consequence of climate change may be
substantially reduced or even eliminated,
depending upon how strongly higher rates
dampen future water demand.
The MWRA's present strategy regarding the
risks posed by climate change is to learn
more before acting. Potentially costly
commitments to develop new water supply
sources or to conserve water will be delayed
until more evidence regarding the likelihood
of their need accumulates. This strategy
Xlll
-------
appears to be an appropriate one for the
Boston area at this time for a variety of
reasons.
First, the pace of global climate change is
unlikely to be so rapid as to result in local
water deficits in the near term given existing
water supply infrastructure and modest
expected regional growth in population and
economic activity. Second, expected water
and sewer rate increases may dampen water
demand substantially, thereby lessening the
risk that climate change would result in a
water deficit if supply capacity is not
increased. Third, the MWRA currently
engages in extensive water supply planning
activities. Other than the MWRA's
participation in this study, their planning
activities have not explicitly addressed risks
posed by climate change. Nonetheless, these
activities can help the region to adapt to
climate change and prevent future water
supply deficits.
The planning activities of the MWRA have
identified possible water sources for future
development and have encouraged their
protection in order to preserve them as
future options. Should future events and
improved understanding of climate science
and impacts indicate that the risk of regional
water deficits is unacceptably high, the
planning efforts of the MWRA will have
laid important groundwork for timely and
flexible responses. For regions which have
not yet done so, the identification and
protection of future water supply options
represent important first steps in a climate
change adaptation strategy that can help to
lessen the consequences of adverse impacts.
XIV
-------
: INTRODUCTION
The MWRA is an independent, ^t'yoUe rte weirn
This system is aiso
Contents of Report . , ^ ^
-------
yield, (6) the financial
,
d
*e possiWe inpae,, of fte - Mping to estimate
Summaiy of Methodology
n Table ,
-------
„„ the sensitivity of the streamflow and
As be can seen in Table 1.2, **
the increases in Et and Ep. An <
* rise impacts upon
„* eases mitigated the
ai cases, increased safe y.eld. If
plified. The worst ease occurred rf
°P
less than the base case. ,,„ _^CT season in each
. , -nQ
WMe there were increases in evapotranspiratron dun^
be sisnifrcantW decreased.
^
mg/ulder, he ffl^-sirauiation period.
approximately 50 months
than close to 100 percent
-------
°f
^
demand.
ncreases
using
s
f
of the
price elasticity of
had to be
-
as
percent.
determined
716 r» ^ TT "~ ocuc yieia oi
yield would decrS 2^ 6ngineering
dwater vi«M^?* 2° percuent « from
^ would remain the same In
mcreased under the GISS
~"~ factors were not
n-gd»d;;;roSes^™frettiro^
-------
the
GISS 2050 demand of
implementation of
pro
entaton ^ _ ^
205°-
g^=S3S3»-"-
-------
-~
Forth ~"v 1U ine ^lection of
For the c demand was
-s,fa™:5£?£^££^ ^"«SKTT per EH
a^, «•» ^* 4e need fe^sss^* ^ a4"££^r ».
-------
• iQQft were equivalent to those in 1987
under the
scenano
percent
no
climate change scenario.
to price even
change could
with low values of elasticity. Even wi
result in increased water supply costs
«f MWRA Policy Response
Summary of MWKA * • „,:„ immcts of possible
. climate
effeCtS'" - A- OT ITS cU.lv.'CUAy ^^ B1
-------
a
endorsement are necessary before w, governmental, profesioi } T°Val °f
-s""
of
~
Needs
As shown by the MWRA>O
^srt-s-r;^
•
Policy making
change.
»
can be provided
then, ,c '
ne what
to
-------
Table 1.1
Summary of Scenarios
Name
Description
Base Present hydrologic conditions as represented by the climate variables of air
temperature, precipitation, incident solar radiation, wind peed, and
relative humidity.
GISS Perturbation of climate variables according to 2 x CO2 results of GISS
model. i
i
GFDL Same as above except GFDL model.
OSU Same as above except OSU model.
UKMO Same as above except UKMO model.
A B% Base case with A degree Celsius increase in air temperature, B % change
in precipitation. For example, 2,+20% is the Base Case with 2 degree
increase in temperature, 20% increase in precipitation.
Incr. re GISS scenario with decreased potential evapotranspiraticn because of
higher canopy resistance.
Ext. Sea. GISS scenario with longer growing season.
-------
Table 1.2
Summary of Results
Note: Reported by percent of average change compared to Base
™
Case except for
. , j- i "• "JLWJ.. j.^, 10 utuicasiily
enrichment. "Ext. Sea" is increasing growing season. More
RUn Et Ep Precip Temp
information is in Section 6.
Flow Yield
Base
GISS
GFDL
OSU
UKMO
2,0%
2, +20%
2,-20%
4,0%
4, +20
4, -20%
4, '-10%
Incr. rc
Ext . Sea .
+20
+57
+23
+10
+12
+12
+12
+24
+24
+24
+24
+24
+5
NA
+17
+41
+13
+32
+6
+ 6
+6
+11
+11
+11 •
+11
+11
+17
+17
o
-1.6
-7.6
+13
+23
0
+20
-20
+ 0
+20
-20
' +10
-10
-1.6
-1.6
+3.67
+4.9
+3.11
+ 8.27
+2
+2
+2
+4
+4
+4
+4
+4
+3.67
+3.67
%
~
-16
-33
+ 6
+30
-8
+23
-39
-15
+15
-44
+ 0.2
-30
-10
NA
mgd
306
236
173
328
421
278
379
161
250
355
139
302
196
262
195
10
-------
2: STUDY AREA
Introduction i ' •..
Boston its neighboring communities, and some communities in western and
Legislature, jyiosi ui u oimr>1ied^ The MWRA service territory includes 46
SSrsrss SMKS: sawtttf- -
study.
In 1987 water demand on the MWRA system was 336 millions of gallons
The aggressive nonstructural approach to meet water supply jeeds^™ the
••* =^r^:^^
long-term water supply requirements of the MWRA.
As reported inMWRA (1990, page 3), over the three
MWRA, among other related activities,:
"- found and repaired MWRA pipes leaking 5 mgd;
- found 25 mgd in leaks in 5000 miles of community
- developed, and issued for comments, regulations
11
year period, the
lines;
requiring leak detection
-------
and accurate metering;
school children" Sjgnou,
°f
«* educated
water."
- helped change the Massachusetts Plumbing Code to require low flow toilets;
- installed water saving devices in over 7000 homes...;
" **** ""^^ °f industrial> Commercial, and institutional
users to save
of developing
ssar
<>> «"J^. *e feasibility
cos, of these
structeal alternatives
trying "all reasonable means of
new source; only if mat failed after a a
plannmg route of augmentation be c^nsMe
planning to review these policies hmshnt „
«th the expansion of m -
, ooT '"""IT1 Water "&>»
' ' Pag° 8)' The MWRA is
' * ^ anticil3ates
12
-------
Quabbin Reservoir is located on the Swift River. It collects water from 186
square miles of Swift River drainage as well as water transferred to the reservoir from the
Ware River watershed by the Quabbin Aqueduct (see additional information below). The
reservoir has a total storage volume of 412 billion gallons and an active storage volume of
256 billion gallons. The active storage volume, or the minimum volume the reservoir is
permitted to obtain, is necessary to maintain the water quality of the water, supply releases.
Quabbin has a mandated minimum downstream flow release! to the Swift
River of 20 mgd for fishery maintenance. During the months of June through November,
the minimum is increased to 45 mgd if the flow on the Connecticut River, (to which the
Swift Sver is tributary, see Figure 2.1) at Montague is less than 4900 cfs The minimum
reZtse il increased to 71 mgd if the Montague flow is below 4650 cfs. The average flow at
Montague during June through November is 7,421 cfs. ,
Water diverted from the Ware River, the next river basin to the east of the
Swift River basin, can be transferred either to Quabbin Reservoir or Wachusett Reservoir.
The decision is based upon the time of year and storage in each reservoir. The diversion
structure controls 96.8 square miles of the Ware River basm Water may only be.taken
during the period of October 15 through June 14 when the flow exceeds 138.5 cfs^This is
referred to as flood skimming. The average monthly flow of the Ware Riyer during this
period is 229 cfs. I
Wachusett Reservoir receives inflow from the Nashua River watershed and
water transferred from the Ware River. Its storage volume is 65 billion gallons. It also has a
minimum downstream release. I
Spills from the reservoirs under flood conditions are rare ajid the impacts are
minor. ]
j
Hydrology }
,1
The United States Geological Survey (1986) reports the average annual
precipitation over the Swift and Nashua watersheds is approximately 42 mches. It is
distributed relatively evenly throughout the year. There is snow from ^^^
December through March. The average annual runoff is approximately 23 inches with a
peak flow in the Spring from snow melt (see Ware River hydrograph, Figure 4.3).
There is approximately 40 inches of precipitation per year; over the
Connecticut River Basin above Montague with an average annual runoff, of 23 mches (see
Connecticut River hydrograph, Figure 4.5). i
13
-------
(0
OQ
>-
OQ
CO
<
0)
Cn
CO
D
<8
14
-------
3: SAFE YIELD MODEL
Introduction
The MWRA's Safe Yield Model is described in MWRA (1989). It uses
reservoir simulation techniques to determine the average annual yield that can be supplied
from th « water supply system with a specified reliability. The average monMy ,
that satisfies the reliability constraint is referred to as the safe yield. The model
onamonthly time step! during each time step the model read* the streamflows^
the system, calculates the net reservoir volume changes due to direct precipitation
orat on determines minimum flow release requirements from Ouabbin reservoir
«on st°eamflow at Montague, and determines the amount of flood skimming (see
d"on a^^ nTof Sect on 2) permissible from the Ware River and the Wachusett flow
re eTse Then reservoir volume balances are calculated and a determination is made if that
±th's^eman1 cln be met. The reservoir volumes are then adjusted tc; reflect the amount
ofDemand supplied If the entire monthly demand can not be met, a failure is recorded. The
^ystmTeSTs measured by the ratio of successful supply months to the total number
of months simulated.
At the option of the user, the model can also simulate drought management
scenarios These scenarios attempt to model how the MWRA might respond to a drought
situation That is when the MWRA noticed that the system storage had: decreased to a
s? —>' * ^ start to rati°n water suppiy reieases svhat *e r T^ a
be fpreadTver several months instead of just one. The amount of rationing («^J^^
Liction of the demand) depends upon the volume of the combined stooge in Quabbm and
WaSusett reservoirs. If the decreased demand can not be met, then a failure is recorded.
Data Requirements
In the Safe Yield Model, monthly demand is specified
average annual demand. They are derived from historic data and are:
a a percentage of the
January
February
March
April
May
June
97 %
96 %
95 %
94 %
97 %
105 %
July 112 %
August in %
September 104 %
October 98 %
November 96 %
December 95 %
Generally, when the MWRA uses the model for studies
hydrologic data for the period 1930 to 1979 (which contains the wors
These data include monthly precipitation on the reservoirs, monthly *
River at the Colbrook gage (the location of the diversion and also i
streamflow into Quabbin reservoir on the Swift River using the ratio
used
15
they use historic
drought of record).
streamflow of the Ware
to estimate monthly
drainage areas), the
-------
cm be traced
St.
16
-------
I
Ware River Transfer Volume as a Functior
Average Monthly Streamflow
T = -255.3 + 0.75Q + 4.43E-2Q2 - 3.16E
0 100 200 300 400 500 600 700
Mean Monthly Flow (cfs)
Figure 3.1 Ware River Monthly Transfer
17
of
-5Q
3
800 900
Volume
-------
c
o
CO
>s
o
Q
w
C
O
0
Ware River Transfer Days as a Function of
Average Monthly Streamflpw
T = -11.2 + 0.223Q -3.84E-4Q2 +2.09E-7Q3
100 200 300 400 500 600 700 800 900
Mean Monthly Flow (cfs)
Figure 3.2 Ware River Monthly Transfer Days
18
-------
4:
CALIBRATION AND VERIFICATION OF RUNOFF MODEL
Introduction ;
As described in Section 3, Safe Yield Model, it is necessary to be able to
— -di^^
climate change scenarios were used in the runoff model to simulate possible impacts upon
steaSlow Ihis section briefly describes the runoff model, and the procedures and results
of tr^ calibration and verification. The resulting simulated flows for present conditions are
referred to as Base Case flows.
Rainfall-Runoff Model
•(-
The rainfall-runoff model consists of two major modules; the National
Weather Service River Forecast System (NWSRFS) snow accunaulatioB and ablation model
CAnderson 1973) and the Sacramento Soil Moisture Model (Burnash et al 1973)
FOKraAN computer codes of each module are available from the US National Wether
sSv^a^pring, Maryland. The project team wrote drivers for the codes tha allowed
for ™t7y of Spu'data anrretrieval and display of the output. Each module is well known
and used in operational and research hydrology. j
The snow melt model is summarized in Figure 4.1. It uses an energy balance
to calculate snowmelt during rain-on-snow periods and a f^^J^L (24
—^^^^
^ed extent of snow as water-equivalence changes and the temperature winch
rain from snow. With these data the model internally determrnes the areal-extent
nd heat deficit of the snow pack at each time step, and the resultmg melt
and.rain in the sub-basin. I
The Sacramento soil moisture model represents the passage of the daily (as in
the case of the snow model, the time step for the soil moisture model; was 24 hours) ram
and melt over the soil surface or through the soil into water bodies such as rivers. I . i
XowT n Figure 4.2. It effectively models direct runoff, interflow, and slower responding
SseZw Evapotranspiration is possible from both upper and lower soil ^- M*£
parameters include maximum storage amounts of tension and free water ni .each ^zone, ^and
Irates of passage and transfer between zones based upon storage volumes. The output of the
model is runoff.
19
approach
-------
assumption did not liS^S k^™ md ve"ficati™ of the models, this
Calibration and Verification for the Ware River
temperature vataes were used L *
Since the United States Geological Snrvpv m<^rlc!^ i, • . • i
of record, i, ^ ^ed tous^ *lffi^TP??e" °frc°Ord'
ass - » ==-
values and constant monthly values of nofeS « + g iteration, parameter
20
-------
^transpiration model which — the
wind speed, sunshine, and ^^^^n^&o^ determined from the
calibrated with the monthly values of P^ evap ^ ^^ ^ ^ ^ gectlon
initial calibration of the runoff model. ^is P™ d 4 4b the monthly streamfiow
5, Evapotranspiration. As can be seen m F §^ 4'4a where ^ closeriess of the values of
ts£ ^?f HEtr Therefore the modd
represents monthly flows in the Ware River basin. .
Calibration and Verification for the Connecticut River
used i
Similar procedures
also a USGS gage, were followed as
area was modeled as one sub-basin,
Daily precipitation and temperature
Newport, VT, Chelsea, VT C
missing values for the period of
Seel Cavendish, VT, where there ^are
temperature were determined and ^.
for the period 1960 - 1970 and verified for the
verification are in Figure 4.5. Using the sme
verification to provide -%""$ ^ Cast flow )
record for the period 1950 - 1 979 (B ase Case
mo/el were the iaverage of values at
^^ ^.^ were
^st square regression relations
e n and
. The model was calibrated
. 1979. The results for the
. ^ ^ Ware ^^
^1949, a simulated monthly
was developed. As in the case of the
for the growing season
represents the flows in the
Use of Simulated Flows in Safe Yield Model
A fin, cnec, on
Safe Yield Model with historical
period 1950 - 1979 and then
derived w«r, the procedures ahove
same periodi The evaporate and
data f e
21
-------
Table 4 1
Present
Ware Hist.
Ware Simul.
169.2
169.0
std. Dev (cfs) skewness
148.3
166.8
Correlation Coefficient between Hist,
1.4
1.5
Conn. Hist.
Conn. Simul.
14550.
14029.
and Simul. = 0.8118
Correction Coefficient betwsen
11810.
11852.
1.5
1.7
= 0.8451
22
-------
Precipitation
Air Temperature
Ta > Pxtemp - Rain
Ta <» Pxtemp - Snow
Legend
x*»».
Input
Rain Or Snow
Heat Exchange
Snow-Air Interface
Areal Extent
Of Snow Cover
Liquid Water A
Storage )
Snowpack Heat Storage
Transmission of Excess Water
Through the Paick
Snowpack Outflow
Figure 4.1 Structure of Snow Model
23
-------
LLLL
Evapotransfjiration
.Demand. :?
Precipitation Input
n
tension Water
uzw ^#^Free Water
UZFW
4.2 sac^ento soil. Moisture Acoounting
24
-------
(sgo) H£»IVHDSIC[
25
a
o
0)
0)
(0
0)
-------
Figure 4.4a Ware River Entire Series (1950 - 64)
or
LJ
LJ
Q
E
D
(SJO)
30Vd3AV
26
-------
Figure 4.4b Ware River Entire Series (1965 - 79)
UJ
LJ
O
LJ
D
Q
UJ
00
4-
(SJD) 39aVHDSIQ 3OVd3AV
27
-------
(0
o
0)
0)
U
0)
u
in
to
fc,
HOHVHOSIQ
28
-------
3
CJ
Figure 4.6a Connecticut River Entire Series (1950 - 64)
o
o
to
o
LO
o
O
CN
O
'it-
ID
cn
I;
-------
LJ
h-
Z)
CJ
O
UJ
O
CJ
Figure 4.6b Connecticut River Entire Series (1965 - 79)
o
CM
en
!--'
' en
00
01
r-.
en
ui
r^
en
en
r-O
r^
en
Q 0
UJ ii
CM
en
LO
U3
en
(spuosnot)j_)
(SJD) 3OdVHOSia
30
-------
5: POTENTIAL EVAPOTRANSPIRATION AND EVAPORATION
Introduction
In a humid climate, such as New England, actual evapotranspiration,
Et', may account
for up to one half of the long-term average annual water budget. Thus, in a
hydroclimatologic sensitivity analysis of a water supply system, careful attention to Et, the
potential evapotranspiration, is necessary. In this study, the Penman-Monteith equation is
employed to estimate Et and the Penman equation is used to estimate the potential (free
surface) evaporation, Ep. With these methods, daily values of Et could be generated to be
used in the Runoff Model and monthly values of Ep for use in the Safe Yield Model.
Potential Evapotranspiration ;
The primary reason for differentiating between potential evaporation,; Ep, and potential
evapotranspiration, Et, is that the diffusion of water vapor into the atmosphere follows very
different physical pathways from vegetation (transpiration) than it does from a free water
surface. Shuttleworth (1979) cites Penman (1948) who defines Ep as "the quantity of water
evaporated from an idealized, extensive free water surface per unit area, per unit time under
existing atmospheric conditions". Shuttleworth (1979) cites Gangopadhyaya et al (1966)
who defines Et as "the maximum quantity of water capable of being lost, as water vapor, hi
a given climate, by a continuous, extensive stretch of vegetation covering the whole ground
when the soil is kept saturated". His definition of Et therefore recognizes the combined
process of transpiration by vegetation and evaporation from saturated bare soil. Estimating
Et is more difficult than Ep because several vegetation species specific model parameters
are required. Furthermore, ways to estimate these parameters are still being developed.
There are several reasons why the Penman-Monteith energy budget equation was
chosen to estimate potential evapotranspiration. First, the Penman-Monteith equation "big
leaf model is presently used by a number of General Circulation Models (GCMs) to
estimate the flux of energy and moisture between the atmosphere and the land surface/water
surface boundaries as discussed by Milly (1991). Secondly, the model is; composed of a
number of the GCM prognostic variables, thereby lending itself to easy perturbation by
climate change scenarios. Lastly, the model is derived from the energy Conservation
equations and therefore it is generally considered to be universally applicable. The Penman-
Monteith evapotranspiration model (Monteith, 1965) is shown below as Eq. 5.1.
A(Rn - G)
pCp(e° (z)
e(z))/ra
XEt =
+ 7 ('1 + rc/ra)
where X is the latent heat of vaporization of water; A is the gradient of
31
(5.1)
:he saturation vapor
-------
presswe-temperature function; Rn is the net radiation; G is the soil heat flux; p is the air
pSsure of th± SP^flCtheat f the * at constant Assure; e°(z) is the satoated vapor
pressure of the ar, a function of air temperature, measured at height z; e(z) is the vapor
pressure of the air measured at height z; ra is the aerodynamic reLtanU to water vapo
diffusion into the atmospheric boundary layer; y is the psychrometric constant; and Tcls
tfie vegetation canopy resistance to water vapor transfer. Notation is summarized in
Appendix B.
mini lk^tians.of ^ Penman-Monteith equation is its data requirements. At a
mimmum, the mode requires air temperature, windspeed, solar radiation and the saturation
vapor pressure deficit. The surrogate of solar radiation employed in the present study is
empirical clear sky estimates of solar radiation combined with observed cloud cover or
percent of possible sunshine. The surrogate for the saturation vapor pressure deficit, e°(z) -
e(z), usua ly estimated from air and dewpoint temperatures, is relative humidity For this
study, which spans 1950 through 1979, average monthly observations of air temperature
windspeed, percent of possible sunshine and relative humidity were obtained for several'
First Order Summary of the Day NOAA climate stations.
In Eq. 5.1, the net radiation, Rn, is described by Eq. 5.2, which is shown below.
Rn = (l-o!) Rs - Rb
(5.2)
where Rs is the shortwave (solar) radiation, a is the surface reflectivity or albedo- and Rb
is the longwave backscatter radiation. In this study the values of a determined by' Brest
(1987) are employed. He derived mean monthly values of albedo for coniferous and
Re^tUno»7?reSt!1and ^Ozen surface water bodies ** the region around Hartford, CT.
Brest (1987) conducted ground-truth radiometer measurements against Lands* satellite
band 4 and 7 observations. The forest albedo values used here assumes an equal mix of
deciduous and conifers in the study watersheds. Therefore, the forest albedo used is an
average of Brest's (1987) deciduous and confer forest values, shown in Table 5 1 It£ also
assumed that the forest vegetation will not evolve sufficiently to alter the albedo within the
next 100 years, as discussed by Nicholson (1988).
A rr mcominS shortwave radiation, Rs, is determined using the methods
described by Heerman et al (1985) and Fritz and MacDonald (1949). The net longwave
back radiation, Rb, which is a function of air temperature and relative humidity is
estimated using methods described by Wright and Jensen (1972) Wright (1982) and
Lmsley Kohler and Paulus (1982). Relative humidity, which is not a variable output from
General Circulation Models (GCM, see Section 6), is derived from the GCM variables of
specific humidity and mixing ratio. This is described more in Appendix A.
In the estimation of Et, the term G in Eq. 5.1, represents the soil heat flux During the
spring and summer months in the New England climate, energy that might otherwise be
32
-------
available for evaporating soil moisture instead heats the soil. Similarly,'the flux of energy
from the soil to the atmosphere during the fall months (soil cooling) will contribute to the
evaporation of soil moisture. The soil heat flux, G, is estimated using a! single soil layer
model described by American Society of Civil Engineers (ASCE, 1990);
The right-hand term of the numerator of Eq. 5.1 is referred to as the sensible heat flux.
The saturation vapor pressure deficit is estimated by Eq. 5.3. ;
e°(z) - e(z) = e°(z,Ta)(l - S/100)
where S is the relative humidity; and Ta is the air temperature.
(5.3)
j
In Eq. 5.3, the saturated vapor pressure, e°(z,Ta) measured at height; z is estimated by
the methods described by Tetens (1930) and Murray (1967). }
In Eq. 5.1, the slope of the saturation vapor pressure temperature curve,
by the methods described by Tetens (1930) and Murray (1967). The latent
vaporization of water ,A,5 is estimated using the method described by Harrison
psychrometric constant, y, is estimated by Eq. 5.4, shown below.
J =
Cp Pa
0.662 X
In Eq. 5.4, the specific heat of moist air, Cp, is assumed to equal 1.
reported by Brutsaert (1982). Functions describing the estimation of the
pressure, Pa, and the air density, p, (of Eq. 5.1) are reported by List (1984)
(5.4)
013 kJ/kg°K as
atmospheric
re and ra in.Eq. 5,1 are respectively dependent upon characteristics
and the surface windspeed drag conditions. In this study, it is assumed
of the study areas can be appropriately modeled as an equal mix of
conifer forests as discussed above.
33
A, is estimated
heat of
(1963). The
of the vegetation
ihat the vegetation
eastern deciduous and
-------
Rauner (1976), pp. 257, defines the low level drag coefficient CaM by Eq. 5.5.
I u* )
CaM = 1 | (5.5)
U(z)
such that
ra =
CaM U(z)
(5.6)
where U* is the friction velocity, defined by Monteith (1973), and U(z) is the windspeed
measured at height z.
Rauner (1976) reports values of CaM of 0.019 and 0.046 for a birch and pine forest
respectively for windspeeds in the range of the present study observed values. These drag
coefficient values are chosen to represent those of a deciduous and conifer forest.
Therefore, based on the reasonable assumption that the study watersheds are comprised of
an equal mix of this forest vegetation, Eq. 5.6 is simplified to Eq. 5.7.
ra =
37.5
U
(5.7)
Eq. 5.7 illustrates an important assumption regarding windspeed. Although Rauner
(1976) measured windspeed profiles above the forest canopy, this sort of observation is
unavailable for this study. Therefore, it has been assumed that the First Order Climate
station windspeed observations, although measured at the 2 meter height, may be suitably
employed hi Eq. 5.7.
Detailed discussion about empirical estimates of the canopy resistance, re, has been left
for last. Ways to estimate this Et parameter, either empirically or through physically based
models of vegetation, are still being developed. Furthermore, the focus of most research
has been on estimating re for cultivated crop irrigation requirement purposes, such as
alfalfa. Monteith (1965) assumes that the ratio of rc/ra equals 15 for forest canopies, and
Eagleson (1984) suggests that rc/ra approximately equals 10.
34
-------
Unfortunately, no long-term empirical measurements of re for forest; canopies have been
reported in the literature. This is due in part to the difficulty in measuring forest
evaporation rates as discussed by Byrne et al (1988). Verma et al (1986) report mid-day
values of 75-160 s/m over a six day period in August for a fully leafed deciduous forest in
eastern Tennessee. McNaughton and Black (1973), pp. 1587, report a re value of 75 s/m
over an eighteen day period in July for a Douglas Fir forest in British Columbia. To
circumvent the difficulty 'of defining a precise value of the canopy resistance, re, in this
study, it was decided to employ it as a runoff model calibration parameter. Monthly values
of re were adjusted so that the long-term monthly average of the daily Penrnan-Monteith
model simulations for Et using historical data agreed with the monthly Values of Et
determined in the calibration of the Runoff Model during the growing season (April through
October for the Ware Basin, May through October for the Connecticut Basin). As a check
on this procedure, the monthly values of re determined for both the Ware'and Connecticut
Basins were within one or two orders of magnitude of the diverse values reported in
literature.
Potential Evaporation ;i
Eq. 5.1 for Et reduces to the potential free surface evaporation, Ep, by substituting the
water .body heat flux G' for the soil heat flux G, and by equating the canopy resistance, re,
to zero. This is the well known Penman equation. Energy that might otherwise be
available for evaporating water heats the water column of the reservoir. 4 As with soil in the
New England climate, One would expect this to occur in the spring and 'early summer.
Similarly, the flux of energy from the reservoir to the atmosphere as it cools during the fall
will contribute to the evaporation of water. In this study, however, time did not permit a
detailed examination of the reservoir energy budgets. Instead it is assumed that the
reservoirs are fully mixed and isothermal. !
t
For a fully mixed, vertically isothermal surface water body, G' will jequal zero and
evaporation may occur at the potential rate, Ep. Clearly, the aerodynamic drag will be
different over water than over a forest canopy. In this study, the form of the reservoir
surface aerodynamic resistance is defined as that reported by Eagleson (1970), pp. 230,
shown below as Eq. 5.8.
In2 (z/z°)
ra =
In2 (200/O.OQi;
(0.41) 2 U(z)
35
(5.8)
-------
where z and z° are the measurement and surface roughness heights (here in centimeters)
and K is Von Karmen's coefficient.
Solving for the values in Eq. 5.8 results in Eq. 5.9.
ra =
886
U
[5.9)
Comparison to Observed Data
Long-term monthly average values of potential free-surface evaporation determined in
this study for the present climate were compared with the Class A evaporation pan
measurements reported by Farnsworth and Thompson (1982), for Norfolk, CT. and
Lakeport, N.H. The monthly values reported here is the average monthly value of both
stations. The available records are for only May through October because the water in the
pans typically freezes during the other months of the year. Monthly values of temperature,
windspeed, possible sunshine and relative humidity, which are all used as input variables to
the Ep model, are from the Hartford, CT. and Concord, N.H. NOAA stations. The Ep
results reported are an average of both Hartford and Concord Ep time series. Figure 5.1
shows that the long-term monthly estimates of Ep are within 0.5 and 1.0 mm/day of the
evaporation pan values. The generated Ep values were multiplied by 0.70 to estimate
reservoir evaporation in the Safe Yield model.
Generation of Ep and Et Time Series
The Connecticut River Et series was generated using the Concord, N.H. NOAA data.
While there is a climate station in Burlington, VT., located near Lake Champlain, it was felt
that this large water body might influence the data needed for the present study in an
unrepresentative way. Furthermore, the Green Mountains of Vermont separate the Hudson
river basin, where Lake Champlain is located, from the Connecticut River basin. The Ware
River Et and the MWRA Quabbin and Wachusett reservoir Ep time series were generated
using a composite monthly average of the Concord, N.H and Hartford, CT. data.
36
-------
Table 5.1 -
^^ * «• C S l/V^ -t~ d "v* TV
JyT *-» «_ o_ "i * ¥ «• t- C. i. ^^
"^"^ ^ ^- CtXiCDT-J ^1 ^ ^* ^ TTI
. ergreen^) Mean(%)
Jan. 10 0 :
9.0
FSb-
March 12 o 10'2
10'7 11.5
•"P^ll i /i n
-t. *x • U ^ -_
12 5
^ •«.,-
JUnS 17'4
July 17-7
"*.
Sept. 14.?
Oct. 12 g
11.6
Nov. 10 _ . ... 12-1
Dec ' 1°-1 10'5
J— 'CL- . T /*\ (-»
10.0
9 . 3
. • - 9.7
37
^^^•^
Lbedo
Water (%)
- — _____
3.1
2.8
' 2.4
2.0
1.7
1.6
1.7
2.0
2.3 -
2.7
3.0
3.2
-------
o
T5
1
c
o
"-4— '
o
I
V
o
CL
0
5 1
4
3
2
1
UJ o
F M A M J J A S 0 N
Month
D
38
-------
6: POTENTIAL CLIMATE CHANGE IMPACTS
i
Introduction
Since there is considerable uncertainty on the exact meteorological impacts that mav
occur under climate change, the US EPA requested that the project team ruTa set of
scenanos that may occur if the equivalent of CO2 doubling occurs. The s "ario include
changes predicted by several General Circulation Models (GCM) and separate S>i ra7y
sensitivity analyses to precipitation, temperature, and other variables. In this section the
data^rom EI&T™* ^ ?*? *"" ** °CMs "* COmP™ to present ctate
data from the study area are given in Appendix A), and then the scenario procedures and
impacts are presented. Since the MWRA reservoir system is the major uppy±ce mlhe
section'ln S^ "Vf * in*Bch.of climate cha<^ ^ * yield are'deterZ d nfh s
Se v°eld of the n^ ' T ^ glVCn °f ^ lmpaCtS °f °ne °f the GCM sc^rios upon
the yield of the other supply sources in the study area. !
)
General Circulation Models I
GCM's are complex numerical models of atmospheric circulation that model
variables such as winds, precipitation, temperature, radiation transfer, cW over air
pressure, and humidity. Because of their complexity, they use large scale grid ly stems - a
typical grid size is 4 degrees latitude by 8 degrees longitude. Within a #d parameters
a^al" Tragedr°^,the ^spheric, land, and sea conditions with! me™T
advantage of using GCM's over sensitivity analysis of selected variables to study the
(SbVr Ay £ 1 pS). *"» ™ ** **"« "***** ** ^
Results from several GCMs were obtained with grid squares centered near 42°
degrees north latitude and 72° degrees west longitude, the approximate centroM^f the
Sun^t"" ,ThCSe G™S*Vec^y investigated the response of the climate to a
doubling of ataiosphenc carbon dioxide concentration. The GCM output data sets were
^mf ^OSpheriC ReS6arCh <*<**) M are ^Sd by
avable rru * I' thfSe ^ thC m°St rCCent GCM data sets NCAR has
available. The GCMs used in this study are described below. • i
(GISS) is described ^ Han^n et al
r A 65 meter deeP slab ocean' ™d*™ atmospheric
The grid square resolution is 7.83° latitude by 10° longitude. The GISS GCM is
fte only model among the four described here that accounts for partial coverage of ocea^
££ a^ f 7thm ' SinglC fid SqUare" The m°del results used in ** p'-ent srud^
that are for a grid square centered at 43.04° by 70°. The model runs were made in 1982
The Geophysical Fluid Dynamics Laboratory GCM (GFDL) is _
Wetherald and Manabe (1991). The Q-flux R-15 GISS GCM model has"
described by
a slab ocean, two
39
-------
soil and nine atmospheric layers with a grid square resolution of 4.44° latitude by 7.5°
longitude. The model results provided for the present study are for a grid square centered
at 42.22° by 75°. The model runs were made in 1988.
The United Kingdom Meteorological Office (UKMO) GCM is described by Wilson
and Mitchell (1987). This model has a 50 meter thick slab ocean and eleven atmospheric
layers with a grid square resolution of 5° latitude by 7.5° longitude. The run results
provided for the present study are from a grid square centered in northeastern
Massachusetts at 42.5° by 71.25°. The model runs were completed in 1986.
The Oregon State University (OSU) GCM is described by Schlesinger and Zhao (1988).
It has two atmospheric layers with a 60 meter deep slab ocean and a surface boundary with
4 0° latitude by 5.0° longitude resolution. The run results provided for the present study
are from a grid square centered in south-central Massachusetts at 42° by 72°. The model
runs used to generate these results were conducted in 1984^ and 1985.
As stated previously, the GCM data used in this study and comparisons to present
climate are in Appendix A.
Procedures
The procedures used for each scenario were generally similar. The US EPA
considers the period of 1950 to 1980 as representative of present climate conditions.
Therefore since the reservoir Safe Yield model requires streamflow, precipitation, and
evaporation (Ep) time series, the objective was to determine how these time series might
change unde the various doubling of CO2 (or 2 x CO2) scenarios and impact the safe yield
of98 5 ^rcent reliability. Because of hydrologic data and Safe Yield model limitations, the
actuai period used in this study was October 1950 to September 1979. While an improved
estimate of safe yield may have been obtained using a longer period of simulation than 30
years, the 30 year period is certainly adequate to determine the types of impacts climate
change may have on the safe yield of the reservoir system.
The time series variables necessary to generate streamflow are precipitation
temperature, and potential evapotranspiration (Et). As discussed earlier, both Et and Ep can
be determined from incident solar radiation, temperature, wind speed, and relative humidity
(which can be determined by either specific humidity or mixing ratio - see Appendix A).
Therefore it must be determined how the driving variables may change under 2 x CO2.
Since the GCMs do not agree well with the present climate but may be general y
representative, it is reasonable to adjust each of the measured present values of fce dimng
variables by the monthly ratios of each corresponding 2 x CO2 scenario value to the 1 x
C02 scenario value. The one exception is temperature; there the present temperature is
increased by the monthly absolute temperature change predicted by the GCM. These
methods have been used in previous studies (for example, in those summarized by Smith
and Tirpak (1989a)).
40
-------
The actual variables used from
the GCMs are in
longwi
already
GISS
relative humidity, and have higher sor
m an average increase in Et of 20 pe?cen° Th
streamflow for the Ware River compared
impact in the Connecticut River is
of 16 percent is the increase in Et
5.1. The longwave
ave back radiation
perturbed. The data
GISS
feSS "^ have sliSMy
" d.OUd ^ ™s resul«
l '" aVerage monthly
ShoWn in fiem «•!- The
*•
results are summarized in Table 6.2
GFDL
the Ware River.
radiation with less cloud cover These SZ, u
significantly higher than those 'o^pSS^^I
Ware River streamflow. As shown in Table filf ]?
safe yield is decreased by 43 percent ' Streamflows
osu " •
warmer, be
have more
sh°WS tKe imPacts on
33 percent less and the
have
.
for the Ware River, tee is
There ,s a srmilar impact on Connecticut
» -ease of six percent in streamflows
UKMO
flow, A
a- sTyield
have ,
>**»*•***> 6.2 sho'ws
™ter thiW "* base
Percent and safe yield by 33 percent.
41
-------
mder the UKMO scenario because
in evapotranspiration durmg the
in *e o*er OCMs
humidity in the UKMO
in fall and winter.
Sensitivity to Temperature and Precipitation Changes
Another series of analyses examined
systems to combinations of changes m
sr
ranged from none
the soil dries.
«, it, Table 6 2 the
As be can seen in Table 62, me
degrees and precipitation ^^.
precipitation were to increase ten to
to four degree temperature rise.
Unfortunately, some "f
precipitation wiU ^remair i
-
worst case occurs if temperature increases 4
w scenarios in this set, if
mitigate the impacts of a two
would actually increase.
) and several GCMs indicate that
the northeastern United States. As the
« m accompanying changes in other
impacts on streamflow and safe yield will be
severe.
Sensitivity to Canopy Resistance
Rosenberg et al (1990) present a comprehensive dis
of vegetation growing in an ^^ ™£^^ Wise
scenario, the canopy resistance (re) could increase oyz P^ ^ .^^ in transpiration
transpiration due to stomatal narrowing would be gr ^^ ^ may
due to increased leaf areas. They also reportJ^*^^. ^ team examined t
not actually occur in fields and fofst* m^* „ .bility by'increasing the monthly values of
sensitivity of the water supply system to this> possi y ^^ ^ safe yield for the
re by 22 percent, and then determining the resulting , (compared to 20
GISS scenario. As shown m T^^^^^tte safe yield were higher than in
percent in the other GISS scenario) and^^^^iched" CO2, some of -
Sensitivity of Length of Growing Season
42
-------
While there were se «
scenarios, there was no mog o *° ~son each of the previous
Base Case Et values. Cornparedto tL fct GISS s ^ &St °ISS scenar>° Tth?
of 12 percent in the Ware basin Jito^^'^*' ^ ** * an inc^se in Et
decrease of 14 percent in the Ware basin £d?ne ^•P01f!CtlCUt basin, a streamflow
yield decrease of 17 percent from 236 m Jto ! Q.^ f the Connec^t basin, and Tsafe
are increases of a few months in the toltl ^ ^ Table 6'2>- Thereto if therf
significantly decreased. g OWmg Season' ^eamflows and; safe yields will I be
Reliability of 306 mgd
sunuiation period. The reliability of
83.4 percent; failures oecur in approrima
penod H 'drought management S
percent. As discussed in the description of
Cement practices inc.de
°f
"Ot be met OTCr
*e OISS sce-«rio
to only 86
close
utilities such as electricity - always
Comparisons to Other Studies
«-' suppy system
to te »e otTer
studies.
Basin. For example, in the scen™o
Percent precipitation decrease, they
decrease of 39 percent The impj of fc GISS
a flow decrease of 25 to 39 percent. FoTme Ware
decrease was 16 percent. Schaake (1990) c
the southeastern United States. Rosenberg e
to 4o found in other
" "* *"" °f ^ P*™e R?*r
S temf ratule i^ease and a 20
f™ ^"^ ™s ^ showed a
r ? D^WaK Kv« *>*** was
C°Imect'c'« ^insi me flow
due
43
-------
Table 6.1
GCM
Variables Perturbed
GCM
a)
3
d
.Oregon State University
Absolute Change in Temperature
ratio
o .
SSdent solar radiation ratio
e) Total cloud cover _ratio
f) Modified soil heat tiux vau
2. United Kingdom Meteorology Office GCM
a) Absolute Change
b) Windspeed ratio
in
Temperature
3.Goddard institute for Space Studies GCM
a)
Absolute Change in Temperatur
otspec
incident solar rado.ati.on
o
d
e Total
f) Mo
cover ratio
a
dified soil heat tiux
humidity
. Geophysical
Fluid Dynamics Laboratory
GCM
a)
I',
absolute Change in Temperature
X??oPoftnf m?xing ratio
incident solar radiation
0*
44
-------
Table 6.2
Summary of Results
Note: Reported by percent of average change for Ware and Connectic
Base Case except for temperature, which is increase in temperature (d
Yield, which is mgd. Yield is the safe yield of the MWRA reservoir s
2 degree increase and 20 percent increase in precipitation. "Incr re" i«
re due to C02 enrichment. "Ext. Sea" is increasing growing season.
Run Et
%
Base
GISS +20
GFDL +57
OSU +23
UKMO +10
2,0% +12
2, +20% +12
2, -20% +12
4,0% , +24
4, +20 +24
4, -20% +24
4, +10% +24
4, -10% +24
Incr. re +5
Ext . Sea . NA
Ep
%
+17
+41
+13
+32
+6
. +6
+ 6
+ 11
+11
• +11
+11
+11
+17
+17
Precip
%
-1.6
-7.6
+13
+23
0
+20
-20
+ 0
+20
-20
+10
-10
-1.6
-1.6
45
Temp
•Gel.
+3.67
+4.9
+3.11
+8.27
+ 2
+2
+2
+4
+4
+4
+4 , .
+4
+3.76 -
+3.67 I
Lit Rivers compared to
egrees Celsius) and
ystem. 2,+20% means
increasing value of
Flow Yield
% mgd
306
-16 236
-33 173
+6 328
+30 421
-8 278
+23 379
-39 16l'
-15 250
M5 355
-44 139
K).2 302
-30 196
-10 262
JA 195
-------
(fl
8
rH
fa
>1
c
flj
o
as
o ^
Q.O)
3 >
•H
(0 «
O 0)
(0 U
1
a
O
VO
0)
!-l
(ga) 3OHVHDSIQ
46
-------
30HVHDSIQ
47
o
to
i-f
fa
>i
r-t
JS
•P
c
_ «
W
•P
O
I
H
•o
in
§
VO
a>
^i
3
D>
•H
EL,
-------
7: MWRA POLICY RESPONSE
Introduction ^ ^e possibie impacts
The previous work in ^^ j^H^Y system. As was shown,
s£££5« dr rr a"^^n u-^™
3rtXs^rs^r?^^*^am-ta*po-M'
centuries.
General Methodology ^ scenario The first
Two analyses are done to ^££fi^£te both the 1 X CO2
Water Demands
and
increases by only live p
-
48
-------
Yields
a 3S '
Water Supply Sources and Costs l
49
-------
ed firs, by the MWRA because^ — ^ J *• «»-«*».«
=?
MWRA service territory (i-e., it does not
the northwest). 9 be $m5
As can be seen in Table 7.3, the ^^T^^U costs are
coital aquifers for water supply). n
„ climate change ***%%££%?££** supplSd from
a,so significantly impacted.
^
of
the future costs
In mis analysis, possible £. ror of
50
-------
households is the number of h™ eLds 8' ^ * °f
of households and industrial users. Since wroSiate^ ST* "^ " *°
projected water demands are from resident cSta 1% Per°en' °f present m
households is twice the actual number oft™, T u c- ^ """"vatent ™mber of
users consists of both water su^d^Ttl fc J?° 7* 1"» °f mOSt
calculating the water rate. The water rate aTt h ' H?!I combmed c°sts is used in
MWRA is wholesale service!) ^ added to " '"nmiunity costs, (The
of
through 1999 and it is assuTJ ^ ±* "^ t0 *"' rate
reagents, their annual costs „*£%
anajysis
assumed that the supply shortage undette OISS V, T' fr°m I99° to 2050' r«
present value of 368 mgd to 294 S and to ,L ^° f^ '^^ from *=
mcreases linearly by five percent "vt to s^erinT^ °f ^^^ h°USeMds
demand increase without climate chalfand^tkT ( 1 P™611' is *e present P™Je*d
price from 1990 to 2060 (see Table™!)) <=°™derat,on of demand sensitivity ,o
the
01SS scenario dtd adlorS
379 mgd. I, is assumed tha, VafrtSESS
3-(HmgoV60 years) with ad.ustmen, elcy
2°5°
du!
wa
demand per EH by the number of EHs Tdth™ T1*1*1"8 uthe Previously determined
demand under climate change, (?) ™ c« of^e^T8 f Vi"Ue by ^ "^o^
rojects in Table 7.3, (4) ft! resuittog w^er rate^EH '? ^^ * **°* *»
r use m the next time step based upon the DercertL, i, (? *" Wata demand Per E«
price elasticity. P Percentage change in EH annual water rate and
i- '"•'
51
-------
necessary.
Ms value. It is assumed that the value
literature for smaller changes in price and
respond to price increases by making easy
(less elasticity) unless new technologies are
demand would be reduced bj '>
25 percent (elasticity of -0.40).
cJps of Engineers, 1991) reviewed the literature
price elasticity of -0.20 to -0.40
SS
conservative.
Results of Financial Analysis
* Water users
dura ^ adjustments
^^ ^
few years if price increased
Course Workbook (US Army
an annual residential average
rf _Q JQ tQ
values in New England of -
value of -0,0 is certainly
1991 MWRA annual wholesale
The cost is approximately
institutional customers. Since there
territory, *ere are 1,600,000 .
of water supply and sewage
is different than the published rate
uses an estimated niber
households in the service
efore the average annual cost
is assumed to equal thi, (This
other differences, the MWRA
f 9Q 000 llons per year per
demand was 361 mgd
percent or 80,000 over to^J^^^to Lpected 1995 demand,
In 1995, the expected number of EH sis i 1,«K ^ Q safe yidd of 368
using the 1990 EH demand, is 362.4 mgd. S nee ^the 1 x ^ ^ time
mgd (see Table 7.2) exceeds ^this ^^^1^^995 MWRA wholesale water and
Acting to data P^^^^ community costs of $160.7 M
sewer rate requirement are $33 7.4 M. Aa Compafed to the 1990 value of $253
^^^
demand will be reduced by 2.2 percent to 220.5 gpd.
52
-------
355.8 mgd a777t •
popular in million of equivalent holeSf Jl iT" feBowing inf°™«ion; the
(Demand) the water demand derived from "e ER? /P rf ^ demand P61 h™^old
Summer demand during climate change 7nc Dmo? T" > ^ increm«« increase m
Pmd), the total available supply (Suppl llw T °f ^ ^ deman* (Tot
supply (Surplus), the tota! annual ccToi ''sutlv the:total «™and and the
2-9 percent to 214.2 gpd/EH. P nt TIus results ln a demand decrease of
below the
safe
exected
«**»*
2050 of
Per EH to 214 mgd per EH.
- ,
Per'°d' demand « «duced from 225 6 mgd
, mere is
is implemented. Up uffl
-rease in EH'S exceeds the increas
""
2045
°f <*^ change After
»e Merimack
"
, e major reason for thi
of 393
«ayed until
(instead of its entire safe yield of 48
«* Merrimackj Diversion to be
approxlmatel>' '5 mgd fluough 2050
»s«s ta the
53
-------
,0 the MWA and tot the
costs for water supply and wastewater
augmenting deficits caused by
to extra costs for
^ ^ ^ so
to climate change that there is really
-, however, that sensitivity to price must
be considered in these types of analyses.
Sensitivity Analysis on Price Sensitivity
decreased
Analyses are also
199C >
*
thss
elasticity in these analyses.
** major deficit
need for consideration of pnce
As noted Piously
.
.re
^
uni^ to areasludTallhl^A- wMch „ -penencm
associated sewer rates because of except^ -dM*t or ^ ftnded
wastewater treatment .rJrastracture^ost wast™ Before to investigate the
by the federal government under the Clean «a» ^^ if its costs were more typical of
fmancial impacts that would have^ "^^^ ^t the MWRA's fixed annual costs
other communities, the analyses are repe « ^sum JS ^^ Me ^ ^e as
' tota, — cost is S27,S M and the
unit cost is $170 per EH.
^e resu,,s for the scenario
The OISS scenario resuits are
supply defied result to *ady rate -e-s
instead of the
in Tab,e
?? „
infrastructure costs induded, under
arelf pSeTagher than the no climate change scenano.
the
the GISS
The
of
more to an area without me eximuiu^, ^ ^-~ that'any demand projections are
infrastructure costs of the MWRA- Uwy' ag icg elasticities, however
sensitive to price even with low values ofeteh^v ^ Additionai sensttmty
climate change could result m fc^d^efSus levels of system reliability
54
-------
of MWRA Staff to Recife, of Stlldy
inthe project,
change on the yield of their syaem Ts *° P0"a*' to"acts of c1
(1991) in his paper ^^ .4 £ SS^^""1™ by Estes-Sroargiassi
money, and the practical consideration of Sn?f "^ G'Ven ^ ^
unhkely to (now) propose substantial^! ZZ , TgS> W a' "* MWRA
we have, and continue to make consLS" ££to ° Change'
* no, (now) make major po.ty
response are
dn-ect «to
flinds for basic maintenance or
""
necessary
°btainin8
for
b
posable local sources located and is
for example, helping member comm
'esf™1;js««) The MWRA also knows :
to could be quickly implemented i
because the MWRA believes they
has
feasMit^
?* n° Other Iar8^ale
someudemand 'eduction measures
"' been imPlemented to date
sc,ent,r5c endorsement are necess^
of chmate change on urban water^
md
face;
after it has. Before that perio
normal variations in weaker, .
due to a new extreme even, or belief in the scieX
""«
*>
°collrred mtil ™ to 40
"**
55
-------
Demand
Fully
Supplied
Partially
Supplied
Other
Unaccounted
For (MWRA)
TOTAL
Table 7.1
MWRA System Demands (MGD)
1990
2050
(1XC02)
242
89
10
20
361
248
104
12
15
379
2050
(2XC02)
258
108
12
15
393
56
-------
Source
Quabbin System
Local Groundwater
Local Surface Water
TOTAL
Table 7.2
MBRA system
1 X CO2
300
20
48
368
57
2 x CO2
236
20
38
294
-------
Source
Local
Sources
Sudbury
River
Consrvt
Merrimack
River
TOTAL
Table 7.3
Additional Sources of Supply
$1990
5^fd Capital Cost
TMGD) (§miiiioas)
7.5
16
30
48
99
40
90
600
~
737.5
2 X C02
Annual Oper & Maint
($millions)
0.5
1.6"
0.0
20.0
—
22.1
58
-------
Table 7.4
Financial Analysis - NO
Year pop
MEH
1990
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
Year
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
^^^M
1.6000
1.6066
1.6132
1.6198
1.6264
1.6330
1.6396
1.6462
1.6528
1.6594
1.6660
1.6726
1.6792
Demand
gpd/EH
~ 225.6 ^
225.6
220.5
214.2
214.3
214.3
214.4
214.5
214.6
214.7
214.8
214.9
214.9
Surplus Cost
mgd
7.0
5.6
12.2
21.1
19.5
18. 0
16.4
14.9
13.3
11.7
10.2
8.6
7.1
^^^^MM
$M
405.2
498.1
644 . 6
644.6
644.6
644.6
644.6
644.6
644.6
644.6
644.6
644.6
644.6
•^^^^^_^_^_
Climate
Change - e
_ Pop Dmd me Dmd Tot Dm
' mgd mgd mgd
~ — 3~6l — FT
355.8
346.9
354.7
356.3
357.8
359.4
360.9
Rate
$/EH
•
253.3
310.0
399.6
398.0
396.3
394.7
393.1
391.6
390.0
388.5
386.9
385.4
383,. 9
59
.
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Rate Chg
•
o.ooo
0.224
0.289
-0. 004
-0.004
-0.004
-0.004
-0.004
-0.004
-0.004
-0.004
-0. 004
-0.004
1 • __
362.4
355.8
348.5
350.0
353 .'l
354.7
356.3
357.8
359.4
360.9
Dmd Chg
• .
0. 0000
-0.0224
-0.0289
0.0004 .
0. 0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
0.0004
= -o.io
3 Supply
mgd
, ^ ^^
~~~368.0
368.0,
368.0
368.0
368.0
368.0
368.0
368.0
368.0
368.0
368.0
368.0
368.0
New Dmd
mgd
: . —
225.6
220.5
214.2
214.3
214.3
214.4
214.5
214.6 '
214.7
214.8
214.9
214 . 9
215.0
-------
Table 7.5
Financial Analysis - Climate Change - e - -0.10
Year
.
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
Pop
MEH
•
1.6000
1.6066
1.6132
1.6198
1.6264
1.6330
1.6396
1.6462
1.6528
1.6594
1.6660
1.6726
1.6792
Demand Pop
gpd/El
, ~
225,6
225.6
220.5
214.2
214.2
214.1
214.2
214.3
214.3
214.4
214.4
214.5
212.4
a
Year Surplus Cost
.
mg
•
1990 7.
1995 (1.
2000 0.
2005 3.
2010 (4.
2015 7.
2020 (0.
2025 (7.
2030 (7.
2035 (7.
2040 (7
2045 (7
2050 36
0
7)
5
3
4)
5
3)
,7)
.6)
.6)
.6)
.8)
.3
$M
405.
498.
645.
645.
650.
650.
650
652
654
656
655
723
723
2
5
0
0
.5
.5
.6
.3
.1
.0
.4
.3
.3
Dmd Incr
__ j
mgd
- — - — -
361.0
362.4
355.7
346.9
348.4
349.7
351.3
352.8
354.3
355.7
357.2
358.8
356.6
Dmd
mi
ILL1
o
\j .
1.
1.
1.
1.
1.
1.
1.
1.
1
1
1
1
Rate Rate
$/EH
•
253.3
310.3
399.8
398.2
400 .0
398.3
396.8
396.2
395.8
395.3
393.4
432.4
430.7
0.
0.
0.
(0,
0
(0
(0
(0
(0
(0
(0
0
(0
ad
yu.
— , —
o
2
1
1
1
1
,1
.1
.1
.1
.1
.1
.1
Chg
-
000
.225
.289
.004)
.004
.004)
.004)
.001)
.001)
.001)
.005)
.099
.004)
Tot.
me
361
D
jd
j
_ — •
.0
363 .6
356.9
348.0
349.5
350.8
352.4
353 .9
355.4
356.8
358.3
359.9
357.7
Dmd
Chg
• —
0 .0000
(0.0225)
(0 .0289)
0.0004
(0.0004)
0 .0004
0 .0004
0.0001
0.0001
0 .0001
0.0005
(0.0099)
0 .0004
supply
mgd
368.0
361.9
357.4
351.3
345.1
358.3
352.1
346.2
347.8
349.2
350.7
352.1
394.0
New, Dmd
mgd
225.
220.
214.
214.
214.
214
214
214
214
214
214
212
212
6
5
2
,2
.1
.2
.3
.3
.4
.4
.5
.4
.5
60
-------
Year
•
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
Year
" —
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
^H
Table 7.6
Financial Analysis - No climate
Reduced Costs
P°P Demand Pop Dmd Incr
TWC'TT
A i.Cj.n. f*mr^ /TPTLT
^J^ / ^-j-ti rncrd
^•*' mere
Change - e
Dmd Tot l
3
1 mgd
1.6000 225 6 -ae
1.6066 225.'6 -gfo'? °'° 361.,
1.6132 225.7 aS-'f °'° 362.,
1-6198 225.8 3«-* °-° 364.]
1.6264 225.9 3" '7 °'° 365.^
1'6330 226.0 369-n °'° 367.4
1-6396 226.0 37o'° °'° 369.0
1.6462 226.1 I12l °'° 370.6
1.6528 226.2 373-? °'° 372.2
1.6594 226.2 373.8 0.0 373.8
1.6660 226.3 3770 °'0 375.4
1.6726 226.4 370'° °'° 377.0
*••*•>**• 226.4 387o8;26 0-0 378.6
u .
Surplus cost R,t-0 „
. u Kate Rate Chg
mg $M . $/EH
£ i-l iT^T
2.3 271 a nf8-5 (°-004)
0.6 27l1 JfZ-8 (0-004
(1'0) 2?2.1 i66'6 ,(°-°04)
^•6) 272 3 iff'f 0.003)
'-•" "32:06 ?- <"-00°33,)
£« : S-1 -°;-')
(1-6) 273.1 J6|-| (0-003)
(1'6) "4- '"..I i^oll
61
^^1^^^.^^^^^
u 380.2
Dmd Chg
o.oooo
0.0004
0.0004
0.0004
0.0004
0.0003
0.0003
0.0003
0.0003
0. 0003
0.0003 •
0.0003 \
0.0003 j
' = -o.io
md Supply
mgd
~ — .
3 368.0
1 368.0
- 368.0
' 368.0
368.0
368.0
369.0
370.6
372.2
373.8
375.4
377.0
378.6
New Dmd
rngd
• .
225.6
225.7
225.8
225.9
226.0
226.0
226.1
226.2
226.2
226.3
226.4
226.4
26.5
-------
Table 7.7
• i zmalvsis - Climate Change - e = -0-10
Financial Analysis
Reduced Costs
Tot D Supply
year
year
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
1.6000
1.6066
1.6132
1.6198
1.6264
1.6330
1.6396
1.6462
1.6528
1.6594
1.6660
— *- i-» ^> f?
225.6
225.6
225.3
225.4
225.4
225.4
225.3
225.2
225.2
220 .0
361.
362.
366.
369.
372
373
0
4
6
.5
.3
.7
Year Surplus Cost
369.B
Rate
n o
\j , \j
1.2
1.2
1.1
1.1
1.1
1.1
1.1
1.1
1.1
1.1
I i
_L . -1-
1.1
361.0
363.6
365.2
366.1
367.7
369.3
370.6
372.0
373 .4
374.8
367.6
369.2
371.0
368.0
361.9
357.4 .
370.6
364.4
361.6
363.1
364.5
365.9
367.3
409.1
403.0
396.8
Rate Chg Dtnd Chg New Dmd
mgd
.
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035
2040
2045
2050
•
7.0
(1.7)
(7.8)
4.5
(3.3)
(7.7)
(7.5)
(7.6)
(7.5)
'(7.6)
41.5
33 .7
25 . 8
_ —
271.8
272.2
277.7
277.7
278.5
280.3
282.2
284.0
285.8
353.7
353.1
351.9
350 7
J _ J v • *
'
169.9
169.4
172.1
171.4
171.2
171.7
172.1
172.5
172.9
213.1
212.0
210.4
208.8
On n n
. u u u
(0.003)
0.016
(0 .004)
(0.001)
0.003
0.003
0.003
0.002
0.233
(0 .005)
(0.007)
(0.007)
o.oooo
0.0003
(0.0016)
0 .0004
0.0001
(0.0003)
(0.0003)
(0.0003)
(0.0002)
(0.0233)
0 .0005
0.0007 ,
0.0007
225.6
225.7
225.3
225.4
225.4
225.4
225.3
225.2
225.2
220.0
220.1
220.2
220.4
62
-------
8: CONCLUSIONS AND ADDITIONAL
Conclusions
RESEARCH
of similar studies. Some of these limitations include:
i
- Limitations of GCMs. !
- Disagreement among the GCMs. |
I
- The methodology to apply results of GCMs to river basin studies.
characteristics
""
°'
«"
climate change.
response to dS
" ASSUmptk>nS in the Et' EP-
™dels tha, may! be vidated under
"" *" °f™**
in
- Inconsistencies in doing sensitivity analysis on just precWtation and
temperature when other important variables will also be^pacted bySt chan^l
possible
"**
- Uncertainty in price elasticity
.
decreased the impacts are even more severe. Impacts are only
^-zzzz -~
63
-------
streamflow but also because flow maintenance requirements mean less of the flows are
available; the low flow season under these types of climate change scenarios is of longer
dumtion and has lower streamflows. Therefore, it appears that climate change couW have
significant detrimental impacts upon streamflows and reservoir yields in the northeast.
The OSU and UKMO GCM scenarios, however, indicate that precipitation in
New England may increase. The increases are significant enough to offset the increases in
ev^omtSrf and evapotranspiration and result in increases in streamflows and system safe
yield.
The contradictory results of the GCMs are not surprising and have been
reported upon by others (for example, Stone, 1992). This again indicates the present
uncertainly in climate change impacts.
Even though further analysis of the impacts of a moderately negative 2 x
C02 scenario (the GISS scenario) upon the MWRA supply system ^wed feat present
would be grossly inadequate and the costs of adaption would be high, the MWRA
tTy notSdng any direct action to adapt to the possible impacts of climate change
yl™^pl«*ofgM climate change are too uncertain and the costs of^direct
action (that is either increasing conservation and/or supply) are too high The MWRA
howeve^will continue to remain aware of possible climate change and formulate actve
ooHcTes when they think it is necessary. They also believe that the climate changes will be
grldual and ley will have time to react; it is possible for them to react to a perceived need
within five to ten years.
The MWRA has the luxury of this response because its already extensive
water supply planning activities and options.
The MWRA's response of "wait and see" is similar to the that of other water
manage s. The relevant conclusions of Schwarz and Dillard's (1990) survey of urban
vemaSgL from cities across the United States were that (1) water %>*£*>£*
see climate change as a "cause for immediate major concern (pg. 365) (2) removal of
some of the uncertainty by scientific consensus and governmental, .professonal and
scientific endorsement are necessary before water managers will act, and (3 ™st ™Pacts
of climate change on urban water can be mitigated, albeit at significant cost (pg. 366).
Additional Research
As all researchers in the area of climate change and water resources have
stated a major scientific research need is to improve the predictive ability of GCMs
enhance to hydrologic components, and determine how best to apply their results to river
InteSis'of policy, Fiering and Matalas (1990) state that better methods are
64
-------
needed on decision-making
As shown by the MWRA
.
65
-------
REFERENCES
Reading, MA., 643 pp
299 PP'
Bconimics, Washington, DC
Boulder, CO., 69. pp.
Eagleson, PA, 1984. Hydroclimatology 1.
66
satellite observations"
Generalized Streamflow
,.Biosphere- .
-------
Technology, Cambridge, MA.
Resources
Estes-Smargiassi S A 1QQ? P
Authority, Boston, ili October0"8'
Farnsworth, R.K., and Thompson EH 1989
Evaporation for the United States! NO^TR
Fiering, M.B. and Matalas
and US. Water
Boston's
•
Massachusetts Water Resources
A™' pan
' '
Climate
Fritz, S. and MacDonald, JH 1949 »*, ,
and Ventiiating, Vol. 46, pp. ftjf Average «* Cation in tte United States" Heating
»»»,
67
-------
RL
of S.ec.edCOiatoate Mode, Runs,
Evaluation, Washington, ,
, « i TT H 1982 Hydrology for Engineers, 3rd edition
Linsley, R.K., Kohler M.. «* Pautas, IL.H, 1982. Hy
New York, McGraw-Hill, 508 pp. „,,„,„ 539
i • «i tables 6th revised edition, Washington, 539
List, RJ, 1984. Smithsonian meteorological tables, btti r
PP' "Sensitivity of evapotranspiration in
dioxide", Climatic Change, Vol
1990, Boston, MA.
.
,
Tournal of Climate, Vol. 5, No. 3, pp ^
journdi ux „„ onS-234 in The state and
-*>
L, London, 241
• l of Applied
Mete
,
orology, Vol. 6, pp. 203-204.
68
-------
District Commission, Boston, MA.
Rauner, J.L., 1976. "Deciduous Forests" no 241
Vol. 2, Edited by J.L. Monteith, '
Trea
' P«5»red for Metropolitan
of
» «>= Atmosphere,
C02
Schlesinger, M.E., and Zhao
002 as simulated by the
Vol. 2, No. 5, pp. 459-495.
C
by d°ubled
layer ocean model", J. of Climate,
Schwa., H.E., ,992. Personal
,, 197, .
commmication
Global Climate
and
69
-------
Technology, 12 pp.
Review, February/March.
Tetens O 1930. "Uber einige meteorologische Begriffe", Z. Geophy,, Vo. 6, pp. 297-309.
United Sfc.es Army Corps of Engineers .
Ft. Belvoir, Virginia.
SX SS SKSB1
Matt DR and Clement, R.J., 1986. "Eddy
Vol36,pp. 71-91.
CricUm«5 CM v I/1C***•»•** . -
model", Journal of Climate, (in review).
, TFB 1987 "A doubled CO2 climate sensitivity experunent
' ling a simple ocean", 1 Geophysical Research, Vol.
^No^a Pp7 13,315-13,343. !
Wright JL 198Z "New Evaporation Crop Coefficient, ASCE , Irrigation and
SageDWsion, Vol. 108.IR2, pp. 57-74.
"Peak water requirements of crops in southern Idaho , ,
" ... „,-.„, „,,. 193-201. /
70
-------
<-™ ^.^. APPENDIX A:
GCM DATA AND COMPARISONS TO PRESENT
Introduction
CLIMATE
""*- both
how the GCM data for specific hidy d mixle ™f °? ^ * disCUSsion of
"** tO CaI°Ulate relative
Pr6SSUre terttl in
y mxe
humidity. Relative humidity is need to cdcuTate^L
Et and Ep equations (see Eq 5
comparison of the
GCM DATA
"
VaIU6S' Table
3 reports air pressure; Table 1-4 rrtotal H '
Estimation of Relative Humidity
The mixing ratio, w, defined by Bras (1990), pp. 85, is shown
below as Eq. A-l.
w =
0.622 e(z)
P - e(z)
where P is the atmospheric pressure (kPa, note that 1 kPa equals 10 millil
>y Bras (1990), pp. 85, is shown below as
millibars)
Eq.
qh =
0.622 e(z)
- 0.378e(z)
71
(A-l)
. (A-2)
-------
Given values of w, Eq. A-l can be inverted to solve for the corresponding relative
humidity as shown below by Eq. A-3.
100
(A-3)
S =
;o (Q.622/P) (1/w + 1/0.6220)
Similarly, given values of qh, Eq. A-2 can be inverted to solve for the corresponding
relative humidity as shown below by Eq. A-4.
100
(A-4)
S =
e° (0.622/P) (1/qh + 0.378/0.622)
Bv usine Eqs A-3 and A-4, monthly values of GCM relative humidity can be
By using uqs. A ^ cuiu , j here ls sometmng
Comparison to Present Climate
TiTalkstein Study
While GCMs function at scales considerably larger than the proj^t^area and '
A study was recently completed comparing the outputs of GCM's for present
72
-------
to
season except Winter, ^^T^^^^ *?" f°< *"
greater. The average actual present precipiS i^/m^a "
climate
precipitation is approximately 0 8 m d ss
present c,imate in Spring J[ Fall, The 2±?
°SU
.2!^
an seasons with an
Project Study
fa
>* °'5 to M *>»>*•**
' C°mPariS°n ** ^ **•
in Tables A-8 through A- 12. y W' The;se data are
v^ Temperature ]
area temperatures a xO2 ba e CCM , *
GCM monthly temperature fare somewhat M mPratoes
and somewhaf low'r STZT
somewhat ,ower than those observed
-n-
°bserved
«Pper plot, the OSU
73
-------
Windspeed
Windspeed is a factor used to determine the aerodynamic -11mtrates
Windspeed. In the upper plot, the OSU GCM monimy T™ ^ monthly windspeed is
- -
r.lnud Cover
fe^ — >- °f
possible — -d°a substantially lower than that observed
?Mns^
Ser months, but under predicts the observed summer monthly values.
.lative Humidity
The relative humidity term is employed to estimate the saturation vapor pressure
deficit ^dSelongwrbackscatter radLion, as discussed earlier Figure A-4 illustrates
tfld^en^b:twgeen the observed monthly — Jj^g
74
-------
relative humidity values seriously disagree with the observed values
Precipitation
during the entire year.
Conclusions
GCM are d
75
-------
Table A-l
Reported GCM Temperatures (°C)
Mon.
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
OSU
lxC02 2xC02
0.08 4.36
0.67 4.28
2.48 6.27
6.53 8.63
9.43 12.59
13.78 17.11
16.43 19.37
16.68 19.71
13.69 16.89
8.54 12.08
5.13 7.15
2.74 5.10
UKMO
GISS
1XC02 2XC02 1XC02 2xCO2
-12.68 -0.64 -0.02 4.97
-9.6
-6.25
3.14
5.53
1.18 9.61
7.35 14.70
15.70 23.23
22.54 27.35
21.42 26.31
15.32 20.27
4.54 11.51
-4.29 3.96
-10.20 -0.75
1.96 6.24
4.03 7.59
7.32 11.66
9.79 13.78
14.95 17.57
19.76 23.25
19.39 22.68
16.59 20.53
10.43 12.89
5.56 8.46
2.18 6.54
GFDL
1XC02 2XC02
-9.81 -2.78
-8.43 -2.26
-1.74 3.72
7.0 10.93
12.64 15.97
17.60 20.92
21.39 26.82
22.36 28.52
17.10 21.42
' 7.56 12.26
-1.77 3.92
-6.50 -3.33
76
-------
Table A- 2
Reported GCM Windspeed (m/s)
Month . °SU UKM° GISS
1XC02 2xC02 lxC02 2xCO2 lxC02 2xCO2
Jan. 1.8 1.2 5.5 5.2 3.9 4.0
Feb. 1.1 0.9 5.5 5.1 2.0 4.0
March 1.1 1.1 5.3 5>1 2>6 2> g
April 0.6 0.8 5.3 5.3 3.0 2.9
May 0.9 1.5 4.9 4.8 2.9 3.8
June 2.2 2.1 4. 3 3.6 3.4 2.8
July 2.4 2.7 3.6 3.2 3.6 3.2
Aug. 1.9 1.9 3. 6 3. 4 2.4 2.7
Sept . 1.8 0.9 4.1 4.1 2.9 3.0
Oct. 1.4 1.8 4.9 4.5 3. 7 3. 7
Nov. 1.1 2.1 5.2 4.7 3.4 4.3
Dec. 0.4 1.3 5.0 4.9 3.8 5.0
77
GFDL
lxCO2 2xCO2
4.0 4.0
4.7 4.7
3.7 3.5
2.6 3.1
2.2 1.8
2.0 .2.4
2.6 3.6
2.1 2.9
0.8 1 . Q
0.6 2.0
.3.1 3.1
3.9 4.2
.
-------
Table A-3
Reported GCM Atmospheric Pressure (kPa)
OSU
UKMO
GISS
Month
1XC02 2xC02 1XC02 2xC02 lxC02 2xC02
99.90 100.26 101.35 101.14 101.74 101.49
Feb. 100.23 100.16 101.19 101.18 101.51 101.23
.21 101.17 101.05 101.17 101.41
Jan,
March 100.11 100
April 100.37 100.16 101.20 101.11 101.27 101.62
May 100.73 100.61 101.26 101.20 101.47 101.42
June 100.66 100.63
101.22 100.83 101.27 101.30
July 100.70 100.82 100.86 100.54 100.61 100.95
Aug. 100.94 100.90 100.80 100.57 100.80 100.89
Sept. 100.84 100.76 101.42 101.27 101.10 100.74
Oct. 100.37 100.44 101.83 101.56 101.01 101.35
NOV. 100.54 100.30
101.53 101.53 101.33 101.24
Dec. 100.24 100.22 101.33 101.35 101.67 101.38
GFDL
1XCO2
99.28
99.27
99.03
99.25
99.19
99.15
99.26
99.37
99.55
99.44
99.39
99.34
2xC02
99.40
99.15
99.22
99.20
99.04
99.24
99.33
99.24
99.33
99.41
99.23
99.37
78
-------
Table A- 4
Reported GCM Total Cloud Cover
1
i^VOTT ,
\«/O U TTT^A/TO
1XC02 2XC02 lxC02 2xCO2 IxCO^Lm?
-^-"-^W^i &J\\*,\J ^
Month
Jan"
Feb- 80 77 _ _ 7Q
March 80 73
65 61
April 79 79
59 50
37 67 61 _ 58 5Q
June 55 47 - ' «
53 51
July 48 56
46 50
AUg" " 49 - - 42 40
8P 72 61 _ _ 43 34
73 ~ - 58 52
75 80 , , _ 6? ^
Dec.
1 UKMO total cloud cover not reported
79
(%)
GFDL
1XCO2 2xCO2
58 53
55 49
52 53
60 58
65 63
66 61
66 50
53 35
60 42
55 44
50 45
' 58 53 •
-------
Table A-5
Reported GCM Specific Humidity or Mixing
OSU
Month
Jan.
Feb.
32.9 43.6
33.5 43.0
March 37.9 48.8
April 48.4 56.1
May
54.4 65.0
June 69.1 79.9
July 78.6 91.9
Aug. 78.2 93.7
Sept. 69-4 82.0
Oct. 54.3 66.2
Nov. 44.4 51.4
Dec. 39.6 46.2
UKMO
16.7 40.6
21.1 49.4
37.3 50.4
54.1 63.3
89.8 85.2
139.9 145.3
130.1 185.7
90.3 173.6
49.5 122.5
29.7 77.6
19.9 52.5
16.7 40.6
1. Mixing Ratio reported
2. Specific Humidity reported
GISS
1XC02 2XC02 1XC02 2xCO2 lxCO2 2xCO2
Ratio (x 10 )
1
GFDL
lxCO2 2xCO2
18.0 23.9
21.5 29.2
27.0 29.4
28.7 39.7
35.7 45.1
52.6 59.2
66.0 81.3
56.1 71.4
47.2 62.9
37.2 40.2
27.8 30.3
23.2 28.6
18.3
21.1
34.4
63.7
91.0
124.3
157.0
159.6
114.1
64.0
33.3
24.6
31.4
32.0
49.5
81.4
113.7
153.5
202.8
195.2
137.7
83 .9
47.8
29.0
80
-------
i^
Table A- 6
Calculated GCM Relative Humidit]
1
OSU UKMO GISS
1XC02 2XC02 1XC02 2xCO2 lxCO2 2xCO
Month
Jan. 85.5 83.6 117.9112.9 48.2 44.
Feb. 83.7 82.9 116. 4 104.6 49.8 49.
March 83.0 81.8 158.1 90.1 53.8 45.
April 79.9 79.9 131.8 85.6 45.5 47.
May 73.9 71.3 141.6 82.4 48.0 46.
June 70.1 65.4 126.5 81.9 50.2 47.
July 67.2 65.2 76.6 81.5 46.1 46
Aug. 66.0 65.2 57.0 81,1 40.2 41
Sept. 71.0 68.1 46.2 83.2 40.5 42
Oct. 78.0 75.0 57.4 92.8 47.7 43
NOV. 80.9 81.2 73.0 105.3 49.8 44
Dec. 85.2 84.0 96.6 114.1 53.0 47
1. See appendix text for explanation of values
percent .
81
r (%)
-. GFDL
2 lxC02 2xC02
7 100.4 100 .3
9 103.8 98.1
8 101.3 98.2
2 100.4 98.0
5 97.7 97.9
8 96.5 96.5
0 95.9 88.8
8 91.9 77.5
.4 91.9 84.3
.9 97.2 92.7
.5 98.6 93.5
.9 104.2 96.5
greater than 100
-------
Table A-7
Reported GCM Precipitation (mm/day)
°SU UKMO GISS GFDL
1XC02 2XC02 1XC02 2xCO2 lxCO2 2xCO2 IxCO, -
Month
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
2.72
2.63
2.72
2.37
2.04
2.11
2.26
1.93
2.94
3.89
2.76
2.98
3.01
2.80
2.75
2.67
1.93
2.10
2.26
2.40
3.31
4.71
4.09
3.52
1.74
1.76
2.16
3.13
3.24
4.22
6.32
5.07
2.87
1.91
1-.73
1.49
2.23
2.32
2.89
3.12
3.74
5.83
7.11
5.26
3.25
2^.59
2.29
1.95
3.21
4.12
3.81
3.51
2.88
4.62
5.65
3.76
3 . 54
4.43
4.44
3.33
3.41
4.41
3.99
3.28
4.. 16
3.53
5.41
4.45
3.23
3.29
3.20
3.34
4.78
4.27
4.42
4.13
4.29
5.48
5.27
2.77
4.68
5.33
4.02
5.23
*-* ^x \~>\~s &
4.98
4.83
3.90
4.16
4.91
4.70
3.62
2.41
3.92
4 .95
4.04
3.03
82
-------
Month Hartford
Jan. -3.71
Feb. -2.54
March 2.48
April
May
June
July
9.30
15.09
20.23
22.91
Aug. 21.78
Sept. 17.24
11.34
5.33
Oct.
Nov.
Dec.
-1.56
Observed
Concord
-6.59
-5.55
-0.05
6.65
12.88
18.14
20.83
19.59
14.96
.9.05
2.99
-4.14
Table A-8
Air Temperature (°C)
Boston Mean
-1.42 -3.91
-0.99 -3.02
3.24 1-89
9.09 8.35
14.63 14.20
19.98 19.45
23.10 22.28
22.14 21.17
18.02 16.74
12.61 11.00
7.22 5.18
0.84 -1.62
83
-------
Table A-9
Observed Windspeecl (m/s)
Month
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Hartford
4.1
4.2
4.4
4.5
4.0
3.6
3.3
3.2
3.2
3.5
3.8
4.0
Concord Boston Mea
3-3 6.4 - 4.6
3-5 6.4 4>?
3'7 6.3 4.8
3.5 6.1 4.7
3.2 5.6 4.3
2-8 5.2 3.9
2-5 4.9 ^e
2-4 4.8 3.5
Q A i- -
2'4 5.1 3.6
2-7 5.4 3.9
2'9 5.9 4.2
3'2 6.2 4.4
84
-------
Month
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Table A- 10
Observed Possible Sunshine (%)
Hartford Concord Boston Mean
55.2 53.1 55.4 54.6
57.8 56.3 57.7 57.2
56.7 52.5 56.7 55.3
56.4 53.6 57.6 55.9
57.7 57.4 61.1 58.7
61.2 58.9 65.1 61.7
65.2 64.8 68.0 66.0
62.1 61.1 65.8 63.0
59.2 56.9 64.1 60.1
56.4 54.9 61.6 57.6
46.5 43.5 50.7 46.9
49.2 48.1 54.0 50.4
85
-------
Table A-ll
Observed Relative Humidity (%)
Month
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Hartford
66
64
63
60
64
68,
69.
72.
74.
71.
70.
70.
.5
.9
.7
.8
.5
.6
.4
2
7
4
5
0
Concord
68.7
66.0
66.6
63.8
65.3
70.5
70.6
73.5
75.2
72.3
73.4
71.9
Boston
63.9
62.6
63.4
62.6'
65.2
67.7
67.2
69.8
70.7
68.5
67.9
65.8
Mean
66.4
64.5
64.6
62.4
65.0
68.9
69.1 '
71.8
73.5
70.7
70.6
69.2
86
-------
Month
Jan.
Feb.
March
April
May
June
July
Aug.
Sept.
Oct.
Nov.
Dec.
Table A- 12
Observed Precipitation (mm/day
Hartford Concord Boston Mean
3.26 2.84 2.32 2.81
3.36 2.85 2.23 2.81
3.34 3.37 2.41 3.04
3.12 3.34 2.55 3.00
2.83 2.71 2.37 2.64
2.43 2.88 2.46 2.59
2.16- 2.51 2.35 2.34
' 3.00 3.27 2.73 3.00
2.82 3.32 2.58 2.91
2.72 2.82 2.50 2.68;
3.62 3.41 3.13 3.39,
3.64 3.41 2.81 3-. 29
87
-------
Figure A.I Temperature Comparison
^* Observed
CD OSU
UKMO
o
o
CD
O
L_
CD
CL
E
CD
C
O
CD
J F M A
Month
A .5 0 N D
•• Observed
CD GISS
GFDL
J F M A M J. j A SON D
Month
88
-------
Figure A.2 Windspeed Comparison
9
8
7
6
5
4
3
2
1
0
•• Observed
CZI OSU
KE2 UKMO
J F M A
M J J A
Month
S 0
N D
Observed
GISS
GFDL
M J J A
Month
89
-------
Figure A.3 Cloud Cover Comparison
CD
C
CO
C
13
00
_CD
.9
*CO
CO
O
CL
CK
v
CD
CO
C
D
CO
_Q
*co
CO
O
Q_
JFMAMJJASO'ND
Month
•• Observed
CD GISS
GFDL
F M A M J J A S 0 N D
Month
90
-------
Figure A.4 Relative Humidity Comparison
Observed
CZ3 OSU
UKMO
J F M A
Observed
M J J A
' Month
IZH GISS
S 0 N D
iGFDL
J F M A
M J J A
Month
0 iN D
91
-------
Figure A.5 Precipitation Comparison
>Y
a
~o
.9-
'o
CD
L_
CL
c
0
CD
a
•a
.9-
'o
CD
L_
Q_
c
O
CD
Observed
OSU
UKMO
JFMAMJ JASON D
Month
•i Observed
CD GISS
ESS GFDL
J F M A M J J A S 0 N D
Month
92
-------
APPENDIX B: NOTATION
CaM: low level drag coefficient (dimensionless)
Cp: specific heat of the air at constant pressure (kJ/kg-°K)
e°: saturated vapor pressure of air (kPa)
e: vapor pressure of air (kPa)
Ep: potential evaporation (mm/day)
Et: potential evapotranspiration (mm/day)
Et': actual evapotranspiration (mm/day)
G: soil heat flux (MJ/meter2-day)
G': reservoir heat flux (MJ/meter2-day)
Pa: air pressure (kg/rneter2)
P: air pressure (kg/meter2)
qh: specific humidity (dimensionless)
Rb: longwave backscatter radiation (MJ/meter2-day)
Rn: net radiation (MJ/Meter2-day)
ra: atmospheric vapor resistance (seconds/meter)
re: vegetation canopy vapor resistance (seconds/meter)
rs: stomatal vapor resistance (seconds/meter)
S: relative humidity (percent)
Ta: air temperature (°C)
U: windspeed (meters/second)
U*: friction velocity (meters/second)
93
-------
,1995-621-995/82126
------- |