&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

-------