operational multisite forecast and reservoir … · operational multisite forecast and reservoir...
TRANSCRIPT
OPERATIONAL MULTISITE FORECAST AND RESERVOIR MANAGEMENT IN NORTHERN
CALIFORNIA
KONSTANTINE P. GEORGAKAKOS, THERESA M. CARPENTER,
AND NICHOLAS E. GRAHAM
Hydrologic Research Center, San Diego, CA 92130
in collaboration with
ARIS P. GEORGAKAKOS, HUAMING YAO AND MARTIN KISTENMACHER
Georgia Water Resources Institute, Georgia Tech, Atlanta, GA 30334
Sponsored by National Oceanic and Atmospheric Administration
(Award No. NA07OAR4310457)
HRC LIMITED DISTRIBUTION REPORT NO. 34
Hydrologic Research Center
12780 High Bluff Drive, Suite 250, San Diego, CA 92130, USA
26 September 2010
Page left blank
i
ACKNOWLEDGEMENTS
The work described in this report was sponsored by the National Oceanic and Atmospheric
Administration (NOAA), Climate Program Office (CPO) and Office of Hydrologic Development
(OHD) under Award No. NA07OAR4310457. It is conceived and executed as a joint project
between the Hydrologic Research Center with expertise in hydrometeorological modeling and
forecasting and the Georgia Water Resources Institute with expertise in reservoir decision support
systems. We are grateful to Dr. Chet Koblinsky, CPO, and Dr. Gary Carter, OHD, for their
support of this Transition to Operations Project. The present research project is a complement to
the INFORM (Integrated Forecast and Reservoir Management) Demonstration Project and benefits
from the advice and support of the INFORM Oversight and Implementation Committee, which
consists of representatives from operational forecast and management Agencies in Northern
California. The authors thank the members of the INFORM Oversight and Implementation
Committee for their feedback and guidance during this project period. Information about the
INFORM Project and the OIC may be found in Georgakakos et al., HRC Technical Report No. 5, August
2006 (available in print from the Hydrologic Research Center and on line at the California Energy
Commission site http://www.energy.ca.gov/pier/project_reports/CEC-500-2006-109.html.
The authors also wish to thank Robert Hartman and Pete Fickenscher of the California Nevada
River Forecast Center for their continuing advice on Northern California operational hydrologic
modeling, and associated data and software support. The ideas and opinions in this report are of the
authors and need not reflect those of the funding or collaborating Agencies.
.
This report should be cited as follows:
Georgakakos, et al. 2010: Operational Multisite Forecast and Reservoir Management in Northern
California. HRC Limited Distribution Report No. 34. Hydrologic Research Center, San Diego, CA, 26
September 2010(NA07OAR4310457), 104 pp.
ii
Page left blank
iii
EXECUTIVE SUMMARY
This report documents activities pertaining to the three-year research project entitled: Multiscale
Operational Forecasting and Reservoir Management for Northern California. The project activities
facilitate the transition to operations of the INFORM (Integrated Forecast and Reservoir
Management) system forecast and decision components for Northern California.
Project activities consisted of: (a) meetings and collaboration with operational forecast and
management Agencies in Northern California aiming to plan the transition to operations of the
INFORM systems, (b) streamlining of the INFORM operational forecast component for land
surface hydrology to emulate with fidelity the current operational hydrologic forecast system of the
California Nevada River Forecast Center (CNRFC), (c) analysis of multi-month water-supply
forecasts and water resources impacts, and (d) generation and assessment of ensemble forecasts and
risk-based water use trade-offs for 2008 through 2010 in support of operational water resources
management by the California Department of Water Resources and the Central Valley Operations of
the Bureau of Reclamation. All assessments were carried out with a start forecast date in the interval
March 1 to 15, which corresponds to the start of critical water use period (growing season). In each
year, the forecast and assessment horizons were nine months.
The meetings with the operational forecast and management Agencies of Northern California
resulted in plans for the transition process that includes alignment of the INFORM and CNRFC
operational hydrologic components. Alignment was necessary because the definition of the
hydrologic segments was updated by the CNRFC since the time of the first definition of INFORM
(2005) concerning the Folsom, Oroville, New Bullards Bar, and Shasta drainages (the Trinity
configuration has not changed). The INFORM multi-month ensemble reservoir-inflow forecasts,
driven by the National Centers of Environmental Prediction (NCEP) CFS (Climate Forecast
System) ensemble forecasts, were used by the INFORM decision component together with the
Sierra Nevada snowpack and large Northern California reservoir initial conditions for March 1 of
the years 2008, 2009 and 2010. INFORM results suggested that compared to the previous two years
(2006 and 2007), in 2008 the Northern California system of reservoirs would have significantly
reduced ability to accommodate the multiple objectives and to meet water demands in Northern
California. The analysis of INFORM forecasts and management results with a March 1 initial date
iv
using observed data indicates that the INFORM system provides reliable predictions and
management trade-offs for a range of climatic conditions for Northern California.
The transition to operations of the INFORM system is on-going (not complete) because: (a) the US
National Weather Service operational forecast system configuration is currently in transition (from
the current NWSRFS to the FEWS/CHPS system) and alignment of the INFORM system with the
CNRFC operations requires additional changes to conform to the FEWS/CHPS software
architecture; (b) NCEP CFS products are currently in transition (making available complete three-
dimensional fields for lead times out to 45 days rather than selected fields and surface variables only)
and the INFORM system forecast component should be changed to take advantage of the new
information for more reliable ensemble predictions in the first forecast month; and (c) upstream
regulation significantly biases operational ensemble reservoir inflow predictions, which at the present
state of INFORM development mirror the operational predictions pertaining to unimpaired
reservoir inflow predictions.
A newly initiated project with the California Energy Commission aims to address (a) and (b) above
(INFORM II project, 2009-2012), and a project with the NWS Office of Hydrologic Development
provides the theoretical basis and formulation for (c) (Project on Upstream Regulation Adjustment
to Ensemble Streamflow Prediction, 2009-2010). Transitioning the tested formulations for upstream
regulation effects into INFORM and River Forecast Center operations is a necessary future
development.
v
TABLE OF CONTENTS
List of Tables vii
List of Figures ix
Introduction 1
First Year Activities (2007-2008) 11
Second Year Activities (2008-2009) 27
Third Year Activities (2009-2010) 32
Discussion and Concluding Remarks 51
References 57
Appendix A: Preliminary Assessment of the INFORM Phase I Forecast-Decision System Results vs. Observed Data for the 2006, 2007, and 2008 Seasons 59
Appendix B: Operational Multiscale Forecast and Reservoir Management in Northern California, Assessments 2009 75
vi
Page left blank
vii
LIST OF TABLES
Table 2-1: New Snow and Soil Parameters for the Folsom Drainage 13
Table 2-2: Channel Routing Reaches and Model Parameters for Folsom Drainage 14
Table 3-1: Snow water equivalent estimated for each INFORM hydrology model
subcatchment. 29
Table B-1: Monthly Average Inflows for Selected Locations (TAF) 95
Table B-2: Reservoir Monthly Parameters 96
Table B-3: Monthly Minimum and Target River Flows 96
Table B-4: Monthly Demands 103
Table B-5: Initial Reservoir Storages on March 15, 2009 104
viii
Page left blank
ix
LIST OF FIGURES
Figure 1-1: Schematic diagram of the distributed INFORM system configuration with data links
indicated. Black arrows signify real-time data links while grey arrows signify off-line
data links. 4
Figure 1-2: A schematic of the INFORM Reservoir and River System 5
Figure 1-3: INFORM DSS Modeling Framework 7
Figure 2-1: Folsom Lake drainage and tributary basins as configured in 2005. NFDC1 – North
Fork American drainage; MFAC1 – Middle Fork American drainage; CBDC1 –
South Fork American drainage; FOLC1 – Local drainage of Folsom Lake. 12
Figure 2-2: Folsom Lake drainage and tributary basins configuration as updated in 2008 (red
outline). The additional segments provide information upstream of smaller
reservoirs in the Forks of the American River. 13
Figure 2-3: Schematic representation of INFORM processing pathways that utilize CFS and
GFS data for the generation of ensemble streamflow predictions with lead time up to
9 months. 15
Figure 2-4: Blended ensemble streamflow predictions of Folsom Reservoir inflow produced by
the INFORM forecast component with initial conditions on 1 March 2008 and
based on GFS (short term up to 16 days) and CFS (longer term up to 9 months).
The first 30 days are shown to highlight the blending period. 17
Figure 2-5: Mean inflow forecasts for 2006, 2007, and 2008 and historical mean inflows for all
major reservoirs. 17
Figure 2-6: Northern California system storage volumes on 1 March 2008 for all major Northern
California reservoirs. 18
Figure 2-7: Water Supply vs. Carryover Storage vs. Energy Tradeoffs for 2008. 20
x
Figure 2-8: Mean Carry Over Storage and Energy Generation Comparisons for 2006, 2007, 2008
for 50% water 21
Figure 2-9: Delta outflow and X2 location Forecast Ensembles. 22
Figure 4-1: Long Range Inflow Forecasts 35
Figure 4-2: Mid Range Inflow Forecasts 36
Figure 4-3: Forecasted Inflow Mean Comparison; Trinity 37
Figure 4-4: Forecasted Inflow Mean Comparison; Shasta 37
Figure 4-5: Forecasted Inflow Mean Comparison; Oroville 38
Figure 4-6: Forecasted Inflow Mean Comparison; Folsom 38
Figure 4-7: Basin average inflow comparisons 39
Figure 4-8: Reservoir Initial Storages 39
Figure 4-9: Sample Tradeoff Plot 1; 40
Figure 4-10: Sample Tradeoff Plot 2; 40
Figure 4-11: Reservoir Elevation Sequences 41
Figure 4-12: Reservoir Release Sequences 42
Figure 4-13: Reservoir Energy Generation Sequences 43
Figure 4-14: X2 Location Sequences 44
Figure 4-15: Delta Outflow Sequences 45
Figure 4-16: Mean Water Delivery Comparisons 46
Figure 4-17: System Energy Generation Comparisons 46
Figure 4-18: Mid Range Reservoir Elevation Sequences 47
Figure 4-19: Mid Range Reservoir Release Sequences 48
xi
Figure 4-20: Mid Range Energy Generation Sequences 49
Figure A-2.1: Historical Average Inflows (Solid Bars Correspond to Modeled Nodes) 61
Figure A-2.2: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow
Sequences, 2006 63
Figure A-2.3: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow
Sequences, 2007 64
Figure A-2.4: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow
Sequences; 2008 65
Figure A-2.5: System Monthly Water Diversion Sequences 67
Figure A-2.6: Observed Total Diversions 67
Figure A-2.7: Observed Total System Initial and Terminal Storage 68
Figure A-2.7: Observed System Energy Generation 69
Figure A-2.8: System Carry-over Storage Comparison 71
Figure A-2.9: System Energy Generation Comparison 71
Figure A-2.10: INFORM Tradeoff: Total Water Delivery vs. System Carryover Storage; 2006 72
Figure A-2.11: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2007 72
Figure A-2.12: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2008 73
Figure B-1: Long Range Inflow Forecasts 80
Figure B-2: Mid Range Inflow Forecasts 81
Figure B-3: Forecasted Inflow Mean Comparison; Trinity 82
Figure B-4: Forecasted Inflow Mean Comparison; Shasta 82
Figure B-5: Forecasted Inflow Mean Comparison; Oroville 83
xii
Figure B-6: Forecasted Inflow Mean Comparison; Folsom 83
Figure B-7: Basin average inflow comparisons 84
Figure B-8: Reservoir Initial Storages 84
Figure B-9: Sample Tradeoff Plot 1; 85
Figure B-10: Sample Tradeoff Plot 2; 85
Figure B-11: Reservoir Elevation Sequences 86
Figure B-12: Reservoir Release Sequences 87
Figure B-13: Reservoir Energy Generation Sequences 88
Figure B-14: X2 Location Sequences 89
Figure B-15: Delta Outflow Sequences 90
Figure B-16: Mean Water Delivery Comparisons 91
Figure B-17: System Energy Generation Comparisons 91
Figure B-18: Mid Range Reservoir Elevation Sequences 92
Figure B-19: Mid Range Reservoir Release Sequences 93
Figure B-20: Mid Range Energy Generation Sequences 94
1
CHAPTER 1: INTRODUCTION
1.1 OBJECTIVES
Current operations in the water sector typically do not use climate forecasts for real time hydrologic
predictions and do not use climate-driven hydrologic predictions for reservoir management. The
investigators have demonstrated the value of climate forecasts in operations through feasibility
studies and through the development of a demonstration project in close collaboration with
operational forecast and management agencies (e.g., Georgakakos et al. 2005). A near real time system
(INFORM) has been developed and implemented for the five largest Northern California reservoirs
(Folsom, New Bullards Bar, Oroville, Shasta and Trinity) producing ensemble inflow predictions
and risk-based planning and management reservoir regulation policies on the basis of NOAA NCEP
Global Forecast System (GFS) and Climate Forecast System (CFS) products (HRC-GWRI 2006).
The GFS products are used for producing ensemble reservoir inflow forecasts out to 16 days, while
the CFS forecasts are used to produce ensemble inflow forecasts out to nine months. To maintain
synchronization with NOAA hydrologic forecast models, the hydrologic components of the
INFORM system are linked with NWS operational databases at the California Nevada River
Forecast Center. A series of technology transfer seminars has paved the way for the transfer of the
decision component to the California Department of Water Resources (DWR) and the Bureau of
Reclamation (Central Valley Operations – CVO). The present project provides support for
collaboration with operational forecast and management agencies in Northern California to continue
validation of the system output in an operational environment and to develop protocols for using
system products in operational forecast and management of Northern California water supply.
The specific scientific objectives of the present project work are to:
a) Complete the validation of integrated system operational products (ensemble precipitation,
streamflow simulations and forecasts, and management scenario assessments and benefits)
through the three additional California wet seasons covered by the project activities.
b) Making necessary adjustments to the component formulations and implementations.
c) Collaborate with CNRFC, DWR and CVO staff pertaining to reciprocal technology transfer,
and training of their staff to ensure a successful transition of the system (both physical
2
hydrology and decision support components) to operations and appropriate use by
operational agencies.
d) Assess and report on the manner and the extent to which the INFORM products are used
by the operational forecast and management agencies.
To provide context for the present research, in the following sections we provide an overview of the
salient features of the INFORM system, followed by a summary of the report contents.
1.2 OVERVIEW OF INFORM SYSTEM
The INFORM software system consists of a number of diverse components for data handling,
model runs, and output archiving and presentation. At its current state of development and input
data availability, it is a distributed system with on-line and off-line components. The system routinely
captures real-time National Center for Environmental Predictions (NCEP) ensemble forecasts. It
uses both ensemble synoptic forecasts from NCEP’s Global Forecast System (GFS) and ensemble
climate forecasts from NCEP’s Climate Forecast System (CFS). The former are used for producing
real-time short-term forecasts, and the latter are used off-line for producing longer-term forecasts as
needed. The reason for the difference between the GFS and CFS processing is the data type
available in real time from each source (three dimensional 6-hourly fields from GFS, and monthly
average surface precipitation and temperature from CFS).
The INFORM ensemble forecast output feeds an off-line decision system for producing risk-based
short- and long-term management alternatives for a nine-month decision horizon. The INFORM
forecast component is implemented at the Hydrologic Research Center (HRC) for real time use and
with data links to the California Nevada River Forecast Center (CNRFC) databases. In addition, the
ensemble reservoir inflow forecasts and maps of the ensemble surface precipitation forecasts of
INFORM out to several days are posted on a secure internet site for INFORM developing
institutions and collaborating forecast and management agencies. The INFORM decision
component is implemented at the Georgia Water Resources Institute (GWRI), the U. S. Bureau of
Reclamation (USBR) and the California Department of Water Resources (DWR) for off-line use.
Figure 1-1 shows a schematic of the system distributed configuration, indicating the data links. The
3
arrows point to the site of the database from which the organization initiating the link receives and
deposits data.
GFS ensemble forecasts of three-dimensional atmospheric fields are captured, archived, ingested
and quality controlled in real time for further use. Downscaling components that use the ingested
ensemble fields produce corresponding ensemble gridded forecasts of surface precipitation and
temperature over the INFORM application area of Northern California. A Geographic Information
System (GIS) locates the gridded forecasts over the terrain of Northern California in geodetic
coordinates and estimates mean areal precipitation and surface air temperature for all ensembles and
forecast lead times, and for the hydrologic catchments that comprise the drainage areas of interest.
Hydrologic models use the downscaled ensemble forecast mean areal quantities as input to produce
ensemble forecasts of snow depth and snow melt during the cold season, and of surface and
subsurface runoff and streamflow, including reservoir-site inflow, throughout the year.
CFS ensemble forecasts of surface air temperature and precipitation with monthly resolution and
with a nine-month maximum forecast lead time are also captured in real time by the INFORM data
ingest system at HRC. At a user-specified time, a probabilistic downscaling component uses the
ensemble CFS forecasts and produces high spatial and temporal resolution surface precipitation and
temperature estimates for each hydrologic catchment in the INFORM region. The hydrologic
component of INFORM is then engaged to produce ensemble reservoir inflow estimates for the
primary reservoir sites of interest. Downscaling and hydrologic forecasting is done off-line (typically
once per month) in this case of CFS processing. The short-term (GFS-based) and long-term (CFS-
based) ensemble reservoir inflow forecasts of INFORM are blended to produce a consistent series
of input to the decision component.
The INFORM DSS is designed to support the decision making process, which is characterized by
multiple decision makers, multiple objectives, and multiple temporal scales. Toward this goal, the
INFORM DSS includes a suite of interlinked models that address reservoir planning and
management at hourly, daily, seasonal, and over-year time scales. The DSS includes models for each
major reservoir in the INFORM region, simulation components for downstream river reaches as
necessary to incorporate downstream decision objectives, optimization components suitable for use
with ensemble forecasts, and a versatile user interface. The decision software runs off-line, as
forecasts become available, to derive and assess planning and management strategies for all key
4
system reservoirs. The DSS is embedded within a user-friendly, graphical interface that links the
models with the database and helps visualize and manage results. A policy assessment model has
also been developed and is part of the DSS.
Training and collaboration with staff of CNRFC, USBR and DWR has produced an efficient
distributed INFORM system for risk-based management and planning.
Figure 1-1: Schematic diagram of the distributed INFORM system configuration with data links indicated. Black
arrows signify real-time data links while grey arrows signify off-line data links.
1.2.1 RESERVOIR SYSTEM
The scope of the originally proposed INFORM Decision Support System (INFORM DSS) included
four reservoirs: Trinity, Shasta, Oroville, and Folsom. However, the operational planning and
5
management of these reservoirs is depended upon the downstream facilities and water uses
including the Sacramento-San Joaquin Delta, and the export system to Southern California. Thus,
based on extensive discussions with the INFORM Oversight Committee, the California Department
of Water Resources, the US Bureau of Reclamation, and the US Corps of Engineers, it was decided
to expand the scope of the original four reservoir system to include most downstream elements that
have a bearing on planning decisions. More specifically, the original project scope was expanded to
include the elements shown on Figure 1-2.
Figure 1-2: A schematic of the INFORM Reservoir and River System
This system encompasses the Trinity River system, the Sacramento River system, the Feather River
system, the American River system, the San Joaquin River system, and the Sacramento-San Joaquin
Delta. Major regulation and hydropower projects on this system include the Clair Eagle Lake
San Joaquin River
San Luis
Clair Engle Lake
Trinity Power Plant
Lewiston
Lewiston
JF Carr
Whiskeytown
Shasta
Keswick
ShastaSpring Cr
Keswick
Oroville
Thermalito
Folsom
Natoma
New Melones
Tulloch
Goodwin
Oroville
Folsom
Nimbus
Melones
Tracy Pumping
Banks Pumping
San Joaquin River
Amer
ican
Riv
er
Feat
her R
iver
Sacramento River
Trinity River Clear Creek
Yuba River
Bear River
Delta-Mendota Canal
California Aqueduct
O’Neill Forebay
To Dos Amigos PP
To Mendota Pool
Sacramento San Joaquin River DeltaReservoir/
Lake
Power Plant
Pumping Plant
River Node
Reservoir/Lake
Power Plant
Pumping Plant
Reservoir/Lake
Power Plant
Pumping Plant
River Node
ISV
IFT
IES,IMC,IYB,ITI
DDLT,DBS,DCCWD,DNBA
DDM
DFDM
DDA
DSF
DSB
Black Butte
New Bullards Bar
6
(Trinity Dam) and the Whiskeytown Lake on the Trinity River, the Shasta-Keswick Lake complex
on the upper Sacramento River, the Oroville-Thermalito complex on the Feather River, the Folsom-
Nimbus complex on the American River, and several storage projects along the tributaries of the
San Joaquin River including New Melones. The Sacramento River and the San Joaquin River join to
form an extensive Delta region and eventually flow out into the Pacific Ocean. The Oroville-
Thermalito complex comprises the State Water Project (SWP), while the rest of the system facilities
are federal and comprise the Central Valley Project (CVP).
The Northern California river and reservoir system serves many vital water uses, including providing
two-thirds of the state’s drinking water, irrigating 7 million acres of the world’s most productive
farmland, and being home to hundreds of species of fish, birds, and plants. In addition, the system
protects Sacramento and other major cities from flood disasters and contributes significantly to the
production of hydroelectric energy. The Sacramento-San Joaquin Delta provides a unique
environment and is California’s most important fishery habitat. Water from the Delta is pumped
and transported through canals and aqueducts south and west serving the water needs of many more
urban, agricultural, and industrial users.
An agreement between the US Department of the Interior, Bureau of Reclamation, and the
California Department of Water Resources (1986) provides for the coordinated operation of the
SWP and CVP facilities (Agreement of Coordinated Operation-COA). The agreement aims to
ensure that each project obtains its share of water from the Delta and protects other beneficial uses
in the Delta and the Sacramento Valley. The coordination is structured around the necessity to meet
the in-basin use requirements in the Sacramento Valley and the Delta, including Delta outflow and
water quality requirements.
The expanded INFORM system is intended to “drive” the decision making process at the long
range (planning) level. An overview of the INFORM DSS is provided next to better clarify the role
of the expanded system.
At present, a number of tools are being used by the federal and state agencies responsible for the
management of the northern California water resources system. Such tools include spreadsheet
models (USBR), hydropower scheduling models (USBR), simulation models (DWR, USACE), and
forecasting models (CNRFC). However, these tools are neither vertically (planning to management
to operations) nor horizontally (agency wise) fully integrated. Perhaps, the most significant
7
contribution of the INFORM project is that it provides an integration framework and a common set
of tools for all operational agencies involved.
1.2.2 INFORM DSS OVERVIEW
The INFORM DSS modeling framework is illustrated in Figure 1-3. The DSS includes multiple
modeling layers designed to support decisions pertaining to various temporal scales and objectives.
The three modeling layers shown in the figure include (1) short range and near real time operations
decision support (which has hourly resolution and a horizon of one day), (2) mid range reservoir
management (which has an daily resolution and a horizon of several months), and (3) long range
planning (which has a monthly resolution and a horizon of one or two years). The INFORM DSS
also includes an assessment model which replicates the system response under various inflow
scenarios, system configurations, and policy options.
Figure 1-3: INFORM DSS Modeling Framework
Response Functions• Energy• Flood Damage• Spillage
Water DistributionFlow RegulationHydro Plant OperationEmergency Response
Monthly Decisions• Releases/EnergyTarget Conditions• State Variables
Planning Tradeoffs
• Water Supply/Allocation• Energy Generation• Carry-over Storage• Env.-Ecosystem Management
Development Tradeoffs
• Urban/Industrial • Agriculture• Power System• Socio-economic & Ecological
Sustainability
Operational Tradeoffs
• Flood Management• Water Distribution• Energy Generation• Env.-Ecosystem Management
Response Functions• Energy• Flood Damage• Spillage
Scenario/Policy Assessment
Monthly / Several Decades
Actual Hydrologic Conditions
Actual Demands
Climate-Hydrologic Forecasts
Demand Forecasts• Water• Food • Energy• Env.-Ecosystem
Climate-Hydrologic Forecasts
Demand Forecasts• Water Supply• Power Load/Tariffs• Flood Damage• Env.-Ecosystem Targets
Inflow Scenarios
Development/Demand Scenarios• Water/Energy Projects• Water/Benefit Sharing Agreements
Daily Decisions• Releases/EnergyTarget Conditions• State Variables
Short Range / Near Real Time Decision Support
Hourly / 1 Day
Mid Range Decision Support
Daily / several Months
Long Range Decision Support
Monthly / 1-2 Years
Infrastructure Develpmnt.Water Sharing CompactsSustainability Targets
Management Policy
Man
agem
ent
Agen
cies
/Dec
isio
ns
Plan
nin
g A
gen
cies
/Dec
isio
ns
Oper
atio
nal
Pla
nnin
g a
nd M
anag
emen
tO
ff-lin
eA
sses
smen
ts
8
The INFORM DSS is designed to operate sequentially. In a typical application, the long range
planning model is activated first to consider long range issues such as water conservation strategies
for the upcoming year in view of the climate and hydrologic forecasts. As part of these
considerations, the DSS quantifies several tradeoffs of possible interest to the planning and
management agencies and system stakeholders. These include, among others, assessments regarding
relative water allocations to water users throughout the system (including ecosystem demands),
reservoir carry over storage, reservoir coordination strategies and target levels, water quality
constraints, and energy generation targets. This information is provided to the planning and
management agencies to use as part of their decision process together with other information. After
completing these deliberations, key decisions are made on monthly water supply contracts, reservoir
releases, energy generation, and reservoir coordination strategies. The INFORM DSS planning level
is linked to the INFORM forecast component through the use of the long range forecasts (9-month
forecast ensemble). The mid range management model is activated next to consider system
operation at finer time scales. The objectives addressed here are more operational than planning
and include flood management, water supply, and power plant scheduling. This model uses mid
range forecast ensembles with a daily resolution and is intended to quantify the relative importance
of, say, upstream versus downstream flooding risks, energy generation versus flood control, and
other applicable tradeoffs. Such information is again provided to the management agencies (the
operational departments) to use it within their decision processes to select the most preferable
operational policy. Such policies are revised as new information on reservoir levels and flow
forecasts is acquired. The model is constrained by the long range decisions, unless current
conditions indicate that a departure is warranted. Lastly, the short range and near real time
operations models are activated to determine the turbine and spillway operation that realize the
hourly release decisions made by the mid range decision process. The results of this model can be
used for near real time operations.
In developing the INFORM DSS, particular attention has been placed on ensuring consistency
across modeling layers, both with respect to physical system approximations as well as with respect
to the flow of decisions. For example, the mid range management model utilizes aggregate power
plant functions that determine power generation based on reservoir level and total plant discharge.
These functions are derived by the short range and turbine load dispatching models which
determine the optimal turbine loads for each plant corresponding to the particular reservoir level
9
and total discharge. Thus, the mid range model “knows” how much power generation will actually
result from a particular daily release decision. Furthermore, the mid range model generates similar
energy functions to be used by the long range planning model. In this manner, each model has a
consistent representation of the benefits and implications of its decisions.
The three modeling layers discussed earlier address planning and management decisions. The
scenario/policy assessment model addresses longer term planning issues such as the implications of
increasing demands, inflow changes, storage re-allocation, basin development options, and
mitigation measures. The approach taken here is to simulate and compare the system response
under various inflow, demand, development, and management conditions.
Altogether, the purpose of the INFORM DSS is to provide a modeling framework responsive to the
information needs of the decision making process at all relevant time scales and water uses. The
INFORM DSS has been provided to the INFORM participating agencies. Training and
demonstration workshops have been conducted to ensure that agency personnel have the necessary
knowledge and experience to correctly use and interpret the results of the software.
1.3 REPORT ORGANIZATION
The report is organized by year of activities performed. The next three chapters summarize the
activities for the first (2007-2008), second (2008-2009) and third (2009-2010) year of project tenure.
Chapter 5 presents a discussion of lessons learned and challenges remaining, as well as conclusions
drawn from the activities of this project. Chapter 6 contains the list of references of the report.
10
Page Left Blank
11
CHAPTER 2: FIRST YEAR ACTIVITIES (2007-2008)
2.1 INTRODUCTION
First year project activities focused on the alignment of the INFORM forecast component
hydrologic models with the operational hydrologic models used at CNRFC for real-time flow
prediction, on the development of INFORM assessments for the spring and summer of 2008 based
on initial conditions on 1 March 2008 and on CFS operational forecasts, and on meetings with
operational forecast and management agencies for strategic planning, coordination and information
exchange. This Chapter summarizes these first-year activities (and associated findings) of this
transition-to-operations project.
2.2 PROJECT ACTIVITIES
2.2.1 HYDROLOGIC MODEL UPDATES
The INFORM hydrologic model component is designed to emulate the hydrologic forecast models
used in real time operations by the CNRFC. The design of the models is based on the snow, soil
models used as part of the National Weather Service River Forecast System (NWSRFS) with
hydrologic segments defined for each large watershed of interest as defined for operational
forecasting. The alignment was first done in 2005. After discussions with staff of the CNRFC it
was found necessary to re align the segments defined to conform to the updates of the operational
hydrologic forecast system. These updates involved the configuration of hydrologic segments
(basins) within large watersheds that provide inflow to the INFORM main reservoirs, and the
estimation of snow and soil model parameters. They also involved changes in the unit hydrograph
estimates for several of the segments that necessitated changes in the channel routing models of
INFORM. It is noted that the INFORM routing models use more detail to route the water in the
stream network than do the operational models, but changes in the configuration of the hydrologic
segments in some cases necessities changes in the channel routing model configuration as well.
Work focused on the large watersheds that provide inflows to Folsom (American River), New
Bullards Bar (Yuba River), Oroville (Feather River), and Shasta (Pit-McCloud-Sacramento Rivers).
The first step was to produce, test and implement software that decodes the segment definition files
12
for the NWSRFS and generates parametric files for the INFORM snow and soil components for
user specified basins. The second step was to generate software to read the unit hydrograph
coordinates from the NWSRFS segment definition files, and to estimate the parameters of a cascade
of conceptual reservoirs (INFORM routing model) that provide the least-square fit to the unit
hydrograph coordinates. The third step was to change the channel segment topological
configuration in the INFORM system parametric files to correspond to the updates in NWSRFS
segment definitions.
As an example, we show the updates done to the INFORM parametric input for the American River
that provides inflow to the Folsom Reservoir. Figure 2-1 shows the original segment
(subcatchment) configuration for the Folsom drainage. It consisted of three tributary basins
(drainages of the North, Middle and South Forks of the American River) and the local drainage
basin of the Folsom Lake. The updated configuration involves several more segments as shown in
Figure 2-2. Updating was necessary to produce inflows to upstream reservoirs in the watershed and
to correct basin boundaries. The definition of the new hydrologic segments necessitated changes in
the NWSRFS segment definition files that contain parameters for the snow, soil and unit
hydrograph models. New generic software written during the first year of the project reads the new
parameters of the segment definition files and creates the parametric tables necessary for the
INFORM models.
Figure 2-1: Folsom Lake drainage and tributary basins as configured in 2005. NFDC1 – North Fork American
drainage; MFAC1 – Middle Fork American drainage; CBDC1 – South Fork American drainage;
FOLC1 – Local drainage of Folsom Lake.
CBDC1
MFAC1
FOLC1
NFDC1
13
Figure 2-2: Folsom Lake drainage and tributary basins configuration as updated in 2008 (red outline). The
additional segments provide information upstream of smaller reservoirs in the Forks of the American
River.
Table 2-1 shows the new snow and soil parameters for all the segments now defined for the Folsom
drainage. Note that in some cases an upper and a lower subcatchment is defined to capture the
snow line separation of upper and lower elevation areas for snow and soil water accounting. These
values are used by the INFORM hydrologic forecast component models to produce ensemble
inflow forecasts for the Folsom Reservoir.
Table 2-1: New Snow and Soil Parameters for the Folsom Drainage SNOW PARAMETERS American % NFDC1UP NFDC1LW MFAC1UP MFAC1LW CBDC1UP CBDC1LW FOLC1L FMDC1 HLLC1 RRGC1 UNVC1 AKYC1 1.2000 1.1000 1.2800 1.1000 1.2000 1.1000 1.1000 1.1500 1.0900 1.1600 1.2500 1.2500 0.9000 0.9500 0.9500 1.0000 0.9000 0.9500 0.9000 1.0500 0.9000 1.0000 1.1000 1.2000 0.0700 0.0800 0.0500 0.0500 0.0500 0.0500 0.0500 0.0300 0.0300 0.0300 0.0500 0.0600 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.0200 0.1200 0.0800 0.0400 0.0800 0.0400 0.0400 0.0800 0.0800 0.0300 0.0200 0.0100 0.2500 0.2500 0.2500 0.2500 0.2500 0.2500 0.2500 0.1000 0.1000 0.2500 0.2500 0.2500 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 2.0000 2.0000 0.1200 0.0800 0.0800 0.0800 0.0800 0.0800 0.0400 0.0400 0.0400 0.0800 0.0600 0.0200 0.3000 0.3000 0.3000 0.3000 0.3000 0.3000 0.1000 0.1500 0.1500 0.3000 0.3000 0.3000 1.9000 1.5000 1.9000 1.5000 1.8000 1.5000 1.5000 1.4000 0.5000 1.5000 1.5000 1.0000 1600.0000 409.0000 1600.0000 409.0000 1600.0000 409.0000 200.0000 1600.0000 1600.0000 1600.0000 1600.0000 1600.0000 19.0000 11.0000 19.0000 11.6000 19.5000 10.4000 6.2500 19.2000 20.7000 19.0000 19.4000 21.7500 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 19.8100 10.6700 19.8100 10.6700 19.8100 10.6700 4.8800 19.8100 19.8100 19.8100 19.8100 19.8100 1.2500 1.0200 0.9500 0.9500 1.2000 0.9500 0.9600 1.2000 1.4500 1.3000 1.1400 1.0300 SAC PARAMETERS American 145.0000 155.0000 155.0000 155.0000 155.0000 155.0000 155.0000 115.0000 95.0000 115.0000 120.0000 115.0000 50.0000 50.0000 50.0000 40.0000 40.0000 40.0000 30.0000 40.0000 30.0000 40.0000 40.0000 30.0000 250.0000 310.0000 300.0000 310.0000 350.0000 380.0000 140.0000 150.0000 150.0000 90.0000 170.0000 150.0000 210.0000 70.0000 110.0000 180.0000 60.0000 140.0000 180.0000 50.0000 70.0000 20.0000 70.0000 60.0000 130.0000 90.0000 60.0000 120.0000 110.0000 120.0000 45.0000 80.0000 70.0000 80.0000 100.0000 170.0000 0.2500 0.1500 0.2000 0.2500 0.2000 0.2500 0.4500 0.2000 0.4000 0.2500 0.2500 0.5000 0.0020 0.0040 0.0020 0.0020 0.0020 0.0050 0.0030 0.0020 0.0020 0.0020 0.0020 0.0020 0.0800 0.1000 0.0600 0.1000 0.0600 0.1000 0.0900 0.0600 0.0600 0.0600 0.0600 0.0600 12.0000 9.0000 14.0000 16.0000 10.0000 16.0000 12.0000 15.0000 14.0000 12.0000 12.0000 15.0000 1.0500 1.1000 1.2000 1.1000 1.2000 1.1000 1.3000 1.3000 1.2000 1.2000 1.2000 1.2000 0.3000 0.2000 0.2000 0.2000 0.2000 0.2000 0.3000 0.2000 0.1500 0.0600 0.3000 0.3000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.2500 0.0000 0.0000 0.0000 0.0000 0.0000 0.1000 0.1000 0.0400 0.1500 0.1300 0.1200 0.0050 0.0080 0.0050 0.0080 0.0050 0.0080 0.0400 0.0050 0.0050 0.0200 0.0050 0.0050 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
FOLC1
NFDC1
MFAC1
CBDC1
AKYC1
HLLC1
UNVC1
RRGC1
FMDC1
14
Changes in the hydrologic segment definition may necessitate changes in the routing configuration
of the channel routing model of the INFORM forecast component. Kinematic routing is used for
these mountainous watersheds with the routing equations written in the form of a cascade of linear
reservoirs for each routing channel reach. This model can emulate the unit hydrograph model used
in operations by CNRFC by forcing the parameters of the cascade to produce unit pulse response
equivalent to the unit hydrograph response. In this case “equivalent” is in terms of least square
error (see Sperfslage and Georgakakos 1996 for details). During this first year of the project software
was written to produce the parameters of the cascade of linear reservoirs for each segment for which
there is a unit hydrograph defined by CNRFC. In addition, changes in the association between
routing channel reaches and newly defined hydrologic segments were made to allocate correctly the
lateral local inflow to the channel reaches. Table 2-2 shows the channel routing parameters for the
Folsom drainage as obtained after the updating process.
Table 2-2: Channel Routing Reaches and Model Parameters for Folsom Drainage KINEMATIC CHANNEL ROUTING MODEL PARAMETERS Reach n alpha Area Segment Description 1 2 5.69 327.6 1 %NF-UP 2 2 3.36 557.8 2 %NF-LO 3 2 6.66 277.6 3 %MF-UP 4 2 3.43 538.9 4 %MF-LO 5 2 8.01 191.1 5 %SF-UP 6 2 2.39 639.9 6 %SF-LO 7 2 5.63 293.7 7 %J1-LO 8 2 5.63 146.8 7 %LOCAL-NF+MF 9 2 3.38 587.3 7 %LOCAL-SF 10 3 5.73 148.9 8 %FMDC1 (ALL UP) 11 2 3.46 295.1 9 %HLLC1 (ALL UP) 12 3 3.99 123.3 10 %RRGC1 (ALL UP) 13 2 2.11 217.5 11 %UNVC1 (ALL UP) 14 2 2.11 499.7 12 %AKYC1 (ALL UP)
Analogous procedures were followed for all the watersheds in the INFORM Northern California
area and new parametric files were obtained for INFORM. This process is a vitally important
process to permit transition to operations of the INFORM system in a way that will be easily
ingested and sustainable (without the need for extensive changes in model structure and parameters
for use by field personnel). The first year work also produced a set of utility codes that may be used
in the future when new updates are made to the hydrologic segment definitions by the CNRFC or
other operational forecast agencies.
15
2.2.2 INFORM ASSESSMENTS FOR 2008 OPERATIONS
Prior assessments made using INFORM system ensemble forecast results produced for a date of 1
March in each year and for a 9-month period may be found in HRC-GWRI (2006, 2007). This
section documents the activities pertaining to the assessments made for 2008 using the INFORM
system and initial conditions on 1 March 2008. The ensemble forecasts produced by the INFORM
system component that served as input to the INFORM system decision component were driven by
Climate Forecast System (CFS) ensemble predictions. Figure 2-3 shows the processing pathway
followed for the CFS-driven ensemble predictions used herein. It is noted that, as opposed to the
GFS-driven real time INFORM ensemble predictions that utilize the 3-D ensemble predictions of
Global Forecast System (GFS - right pathway in Figure 3), the CFS-driven longer-term predictions
utilize the predicted monthly surface variables of precipitation and temperature.
Figure 2-3: Schematic representation of INFORM processing pathways that utilize CFS and GFS data for the
generation of ensemble streamflow predictions with lead time up to 9 months.
16
These ensemble CFS predictions of the surface variables on a monthly scale are used to condition
the operational Ensemble Streamflow Prediction (ESP) procedure that employs historical
precipitation and temperature time series information. A description of the probabilistic approach
for conditioning and downscaling is given in Carpenter and Georgakakos (2001). Discussion of the
blending of the shorter-term (0 – 16 days) GFS-driven ensemble inflow predictions with the longer-
term (1 – 9 months) CFS-driven ensemble inflow predictions is given in HRC-GWRI (2006). We
only note here that blending is accomplished through the persistence of the snow pack and soil
moisture states in INFORM. We also note that the on-going second phase of the INFORM project
will produce CFS-driven ensemble predictions using available 3-D information from CFS output,
much like is done for GFS output.
Comparison of the 1 March 2008 CFS ensemble predictions of surface precipitation with the
climatology of such predictions for the last 20 years indicated that the nine month period is expected
to be near the average. Therefore conditioning on CFS this year produced the same results as the
unconditioned ESP procedure (Carpenter and Georgakakos 2001). As an example of the results, Figure
2-4 shows the Folsom Reservoir ensemble inflow predictions, conditioned on operational CFS
ensemble predictions of surface precipitation, for the blending period of the first month. Analogous
results were produced for all the other reservoir inflows for the same prediction horizon of 9
months. Comparison of the 2008 ensemble forecast predictions to those of the last two years and
to the historical averages may be made by reference to the results of Figure 2-5. This Figure displays
the inflow means for each case of year and major reservoir. Oroville and Shasta show significantly
lower mean forecast inflows in 2008 than in the previous two years.
These ensemble inflow predictions were used by the INFORM decision component to assess the
impacts on reservoir operations for the prediction horizon. The components of the INFORM
decision system used for the assessments are shown in Figure 1-2. The components are described in
detail in HRC-GWRI (2006). They include: Trinity River System (Clair Engle Lake, Trinity Power
Plant, Lewiston Lake, Lewiston Plant, JF Carr Plant, Whiskeytown, Clear Creek, and Spring Creek
Plant); Shasta Lake System (Shasta Lake, Shasta Power Plant, Keswick Lake, Keswick Plant, and the
river reach from Keswick to Wilkins); Feather River System (Oroville Lake, Oroville Power Plants,
Thermalito Diversion Pond, Yuba River, and Bear River); American River System (Folsom Lake,
Folsom Plant, Natoma Lake, Nimbus Plant, Natoma Plant, and Natoma Diversions); San Joaquin
River System (New Melones Lake, New Melones Power Plant, Tulloch Lake, Demands from
17
Goodwin, and Inflows from the main San Joaquin River); and Bay Delta (Delta Inflows, Delta
Exports, Coordinated Operation Agreement--COA, and Delta Environmental Requirements).
Figure 2-4: Blended ensemble streamflow predictions of Folsom Reservoir inflow produced by the INFORM
forecast component with initial conditions on 1 March 2008 and based on GFS (short term up to 16
days) and CFS (longer term up to 9 months). The first 30 days are shown to highlight the blending
period.
Figure 2-5: Mean inflow forecasts for 2006, 2007, and 2008 and historical mean inflows for all major reservoirs.
18
Forecasted inflows were provided to the decision component with start date 1 March (112 traces, 9 -
month horizon, and five locations: Clair Engle Lake (Trinity), Shasta, Oroville, Folsom, and Yuba).
Historical monthly average values were used for locations where forecasted inflows were not
available. The decision model was run using monthly reservoir parameters and constraints (max,
min, and target storage levels; evaporation rates); minimum river flow and Bay Delta requirements;
and base monthly demands at all locations. The decision objective was to develop the tradeoff
between water supply deliveries and carry-over storage that meets all other stated system
requirements.
The initial water volumes in the major reservoirs in Northern California on the 1st of March play a
significant role in management operations. Figure 2-6 shows that in 2008, reservoir storage at the
beginning of March was lower than that of the years 2006 and 2007, especially for the Oroville and
Shasta reservoirs. This together with the lower mean inflow volumes predicted (as mentioned earlier
for the results of Figure 2-5), is expected to substantially influence system operations in 2008.
Figure 2-6: Northern California system storage volumes on 1 March 2008 for all major Northern California
reservoirs.
Reservoir Initial Storages On March 1st
2013
3872
2992
463
2020 20191902
3786
2997
594
20021895
1490
2660
1456
375
1532
1774
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Trinity Shasta Oroville Folsom New Melones San Lius
Stor
age
(100
0 AF
)
Y2006Y2007Y2008
19
More specifically, the cumulative initial system storage on March 1st, 2008, was approximately 50%
of the total storage capacity, in contrast to the initial storage situation on March 1st, 2007, which was
approximately 80% of the total storage capacity, and to the total storage on March 1st, 2006, which
was 81% of the total storage capacity. The low initial storage and the below normal inflow forecasts
combine to create stressful conditions for the Northern California river system in 2008.
These stresses are depicted on the tradeoffs of Figure 2-7 which quantify the expected relationship
between water deliveries, carry over storage, and energy generation for 2008 (red line) and the
average hydrology (black line). The figure shows that for the same water delivery fraction, carry
over storage and energy generation are expected to be below average in 2008. For water deliveries
corresponding to 50% of the base line demands, carry over storage and energy generation are
expected to be 17.6% and 9.5% less than those that would have materialized under normal
hydrologic conditions.
Figure 2-8 compares the end of November storages (carry over storages) and energy generation for
2006, 2007, and 2008 for water deliveries equal to 50% of the baseline demand. The figure illustrates
that meeting water deliveries at the 50% baseline level in 2008 has the potential to deplete carry over
storage level down to 20% of the system capacity. Furthermore, hydro-generation in 2008 is
expected to be 32% less than in 2006 and 24% less than in 2007.
Alternatively, the previous results imply that to maintain the same system carry over storage, water
deliveries in 2008 must be reduced almost by 50% of those in 2006 or 45% of those in 2007, a very
significant reduction.
20
Figure 2-7: Water Supply vs. Carryover Storage vs. Energy Tradeoffs for 2008.
Total Demand Fraction vs. Terminal Storage Tradeoff
8871
8040
7210
6373
5533
88808372
8078
7478
6718
4000
5000
6000
7000
8000
9000
10000
0 0.1 0.2 0.3 0.4 0.5 0.6
Demand Fraction
Term
inal
Sto
rage
(100
0 A
F)
Y2008
Historical Mean
Total Demand Fraction vs. System Energy Tradeoff
3736
4164
4556
4923
5254
47554932
5233
5515
5802
3000
3500
4000
4500
5000
5500
6000
0 0.1 0.2 0.3 0.4 0.5 0.6
Demand Fraction
Ener
gy (G
WH
)
Y2008
Historical Mean
21
Figure 2-8: Mean Carry Over Storage and Energy Generation Comparisons for 2006, 2007, 2008 for 50% water
delivery fraction.
The previous discussion is focused on mean quantities. The INFORM DSS provides the means to
assess the entire frequency distrbution as illustrated on Figure 2-9. This figure shows the delta
outflow and X2 location forecast ensembles for the 10% water deliveries case. All model runs
ensure that the delta X2 constraint (maximum 80% distance from the Golden Gate Bridge) is met
for a user-specified reliability, set to 90% in the above described experiments.
Simulated Total Terminal Mean Storage Comparison
95378995
5533
0
2000
4000
6000
8000
10000
12000
Y2006 Y2007 Y2008
1000
AF
Simulated System Mean Energy Comparison
7841
6946
5254
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
Y2006 Y2007 Y2008
GW
H
22
Figure 2-9: Delta outflow and X2 location Forecast Ensembles.
2.2.3 MEETINGS WITH OPERATIONAL FORECAST AND MANAGEMENT AGENCIES
During the first year of the project two meetings were held in Sacramento at the CNRFC offices.
The first on 20 March involved the INFORM Oversight and Implementation Committee (OIC) and
the second on 3 April involved representatives from the Department of Water Resources (DWR)
and CNRFC. The first meeting was to provide strategic advice to the PIs of the project by
representatives of operational forecast and management Agencies that are interested in the use of
INFORM forecasts and decision support and to present to such representatives the assessments for
2008 (see previous section). The second meeting was to discuss the INFORM management focus
for the California Department of Water Resources, the updating of the INFORM hydrologic
modeling components (see section 2.1 above), the possible expansion of the OIC to include
representatives from additional Agencies and expertise, and to develop a plan for treating upstream
regulation in an ESP framework. The Summaries of meeting proceedings for both meetings follow.
23
2.2.3.1 Meeting of 20 March 2008
PARTICIPANTS
Agency Representatives
Michael Anderson California Department of Water Resources
Paul Fujitani Central Valley Operations, U.S. Bureau of Reclamation
Robert Hartman California Nevada River Forecast Center, National Weather Service, NOAA
Joe O’Hagan California Energy Commission (through a conference call)
Tom Morstein-Marx Central Valley Operations, U.S. Bureau of Reclamation
Eric Strem California Nevada River Forecast Center, National Weather Service, NOAA
Dingchen Hou National Centers of Environmental Prediction, NOAA (through a
(for Zoltan Toth) conference call)
INFORM Co-PIs and INFORM Project Scientists
Aris Georgakakos (Co-PI) Georgia Water Resources Institute
Kosta Georgakakos (PI) Hydrologic Research Center
Nick Graham (Co-PI) Hydrologic Research Center
Martin Kistenmacher Georgia Water Resources Institute
Huaming Yao Georgia Water Resources Institute
LOCATION AND TIME
The meeting was held at the Joint Operations Center (3310 El Camino Ave.) in Sacramento on the
20th of March 2008. The meeting started at 1:15PM and ended at 3:15PM.
PURPOSE AND INFORMATIONAL MATERIAL
The meeting served as the first critical review meeting for INFORM (Integrated Forecast and
Reservoir Management), Phase II. The meeting presentations in PDF form were made available to
all the OIC members and participants through the following link at the HRC web site:
http://www.hrc-lab.org/INFORM/. In addition, the published reports of the previous phases of
INFORM were also made available via links to the INFORM Core Program Office web site and
through transmission via email. The INFORM Core Program Office web site is:
24
http://www.hrc-lab.org/projects/dsp_projectSubPage.php?subpage=inform.
INFORM STATUS PRESENTATION AND DEMONSTRATION RESULTS FOR LAST
THREE YEARS
The Co-PIs summarized the technical activities of (a) the three-year first phase of INFORM (funded
by CALFED, the California Energy Commission, and the Climate Program Office of NOAA), (b)
the Interim Phase of INFORM (funded by the Climate Program Office and the Office of
Hydrologic Development of NOAA), and (c) the activities of Phase II of INFORM so far (funded
by NOAA Climate Program Office and with contracting in progress with the California Energy
Commission and the California Department of Water Resources). The Co-PIs discussed the design
and implementation of the forecast and decision support components of INFORM and the focus of
the demonstration effort so far, including a presentation of the principle results for the
demonstration winter seasons: ’05 - ’06, ’06 – ’07, and ’07 – ’08. It was noted that with the current
long term (up to 9 months) and large scale forecasts by the National Centers of Environmental
Prediction (NCEP) of the National Oceanic and Atmospheric Administration (NOAA), the current
conditions of snow pack and soil moisture, and the current levels of reservoir storage in Northern
California, the INFORM forecast and decision components indicate that this year (’07 – ’08) has a
good chance of being significantly below average in terms of possible reservoir releases available to
meet demands. These assessments were made with information at the beginning of the month of
March 2008.
DISCUSSION
The participants discussed the role of INFORM for real time operations and the possible expansion
of the Oversight and Implementation Committee to include other key participants for the second
phase of INFORM (e.g., possibly with knowledge and experience in forecasting and management
associated with flood control, fisheries, Bay Delta issues, energy industry concerns). The OIC
members also emphasized the importance of generating INFORM products that are easily ingested
within routine operations of the forecast and management agencies at the end of Phase II. The
INFORM developers emphasized their desire to produce INFORM products and software that can
be used by management agencies to determine the range of the most profitable release policies at
various risk levels and with due consideration of short and long term forecasts prior to the
25
engagement of their existing detailed simulation models that are used to finalize the most
appropriate release policy. The suggestion was made to link some of the INFORM demonstration
activities with Folsom reoperation, as a possible demonstration effort. It was also mentioned that a
significant component of the Phase II project will be to continue the effort to make forecast and
management agencies familiar with the basis and use of the INFORM methodology, which has been
designed and tested over a period of two decades to address decision problems in operations given
uncertain short and long term forecast information.
As a result of the discussions held, it was decided to have a follow-on visit of the INFORM PI in
two weeks time to coordinate with Northern California forecast and management agencies the
details of the expansion of the OIC, and the development of a phased plan for agency science
cooperation and technology transfer in the context of INFORM Phase II.
2.2.3.2 Meetings of 3 April 2008
The meeting was with the Department of Water Resources representatives at first and then with
CNRFC representatives in Sacramento at the Operations Center. The participants were Gary
Bardini and Mike Anderson from DWR and Robert Hartman from CNRFC. A number of ideas
were discussed with respect to the phase II of INFORM and items of particular interest to these
forecast and management Agencies were identified. Among other things, the real time production
of ensemble forecasts driven by GFS and CFS operational ensemble predictions for the Northern
California domain with resolution of 10 km was identified as a useful complement to operations at
CNRFC, and the uncertainty management ability of the decision component of INFORM was
identified as particularly useful for DWR operations. Several potential additional members of the
OIC committee were identified and the PI of the project was tasked to approach such individuals to
discuss their potential participation once the second phase of INFORM is fully contracted (in
progress). With respect to more technical matters, the issue of upstream regulation was the focus of
much of the technical discussion with CNRFC.
Current Ensemble Streamflow Prediction (ESP) procedures provide risk based information through
the production of ensemble members for full natural flows for regulated points or actual flows for
unregulated points. However, users of this information are typically concerned with actual flows
even at points with upstream regulation. It is desired to design and test a procedure to incorporate
26
upstream regulation to the production of an ESP. Instead of a site specific implementation, which
appears to be infeasible for general application as part of the NWS ESP procedures, it appears more
appropriate to develop a structured parameterized approach to simulate upstream regulation so that
it may be used for various sites with the requirement that historical data is available to estimate the
parameters and possibly the structure functions of the methodology.
The American River was discussed as a potential case study for this analysis, with emphasis on the
Middle and South Forks with significant upstream regulation. To obtain data and information for
the structure function approach to be followed, participation of Agencies responsible for upstream
regulation is desirable. A workshop for bringing such agencies together was discussed for May 2008.
2.3 CONCLUDING REMARKS
The first year of the NOAA transition to Operations Project titled: Operational Multiscale Forecast
and Reservoir Management in Northern California was devoted to close collaboration with
operational forecast and management agencies, alignment of the forecast and management systems
developed by the PIs with those of the operational Agencies, and production of assessments for the
ability of the system to meet requirements and satisfy objectives of operation for 2008. Based on
Agency recommendations a number of new utility software components were developed to facilitate
the transition of the INFORM forecast and management components to operations. The 2008
assessment for Northern California water supply concluded that 2008 will be a low year for Oroville
and Shasta both in terms of expected inflows but also in terms of initial storage volume, and thus
will lead to lower terminal storage volumes, lower energy production and less flexible trade-offs. It
appears likely that 2008 water deliveries might be 40 to 50% less than those in 2007 and 2006.
27
CHAPTER 3: SECOND YEAR OF ACTIVITIES (2008-2009)
3.1 INTRODUCTION
The second year activities focused on establishing the validity of the system quantitative results for
the Northern California water management agencies through an independent evaluation of the
forecasts and assessments for the first three years using the INFORM system against observations
of hydrologic conditions and against actual reservoir management outcomes. The results were
presented in a meeting of the agencies in Northern California. These results together with the
continuing collaboration with the Northern California forecast and management agencies
contributed to the funding of the second 3-year phase of INFORM (INFORM II) that will
complement as matching funding the transition to operations that is partially funded by the present
project. In addition during the second year, the project team worked on the development of
INFORM assessments for the spring, summer and fall of 2009 based on initial conditions on 15
March 2009 and on GFS/CFS operational forecasts. Chapter 3 summarizes these second-year
activities (and associated findings) of this transition-to-operations project.
3.2 PROJECT ACTIVITIES
3.2.1 EVALUATION OF INFORM REAL TIME FORECAST AND MANAGEMENT
ASSESSMENTS (2006, 2007 AND 2008)
The California Energy Commission undertook the independent evaluation of the first three years of
the INFORM project real time assessments (spring-fall of 2006, 2007, and 2008) to assess the utility
of the project from the perspective of the management agencies against actual observations of
hydrologic conditions and actual reservoir release decisions. As part of this process, during the
period September – December 2008, HRC and GWRI cooperated with the Energy Commission and
with their consultants to develop an initial evaluation of the project results and effectiveness. The
HRC-GWRI resultant report is included herein in its entirety as Appendix A.
The main conclusions are:
28
(a) This initial comparison indicates that the information provided by the INFORM system in
operational planning and management of the northern California river and reservoir system is
relevant, reliable, and decision worthy. This is corroborated by the favorable comparison of the
predicted and actually observed system outputs (water deliveries, carry-over storages, and energy
generation amounts) over the March to November 2006, 2007, and 2008 seasons.
(b) The value of the INFORM forecast-decision system ultimately depends on whether the
information it generates provides management agencies with good appreciation of the benefits and
risks associated with different decisions and helps them adopt those that serve well the interests of
the system stakeholders. The Northern California management Agencies were very supportive of
the system in interviews associated with this evaluation.
The results of the evaluation were discussed in several conference calls and a presentation for the
management agencies of Northern California. This evaluation and the continuing collaboration of
the HRC-GWRI team with operational forecast and management agencies in Northern California
contributed to the decision of the California Energy Commission to continue the support of the
INFORM project for another 3 years (phase II, 2009-2012) as matching funding for needed
INFORM upgrades (requested by the management agencies), the complete transition to operations
and continuing evaluation of the real time forecasts and decision assessments.
3.2.2 INFORM ASSESSMENTS FOR 2009 OPERATIONS
Prior assessments made using INFORM system ensemble forecast results produced for a date of 1
March in each year and for a 9-month period may be found in HRC-GWRI (2006, 2007) and
Georgakakos et al. (2008). This section documents the activities pertaining to the assessments made
for 2009 using the INFORM system and initial conditions on 15 March 2009. The ensemble
forecasts produced by the INFORM system component that served as input to the INFORM
system decision component were driven by Climate Forecast System (CFS) ensemble predictions.
Figure 2-1 shows the processing pathway followed for the CFS-driven ensemble predictions used
herein. It is noted that, as opposed to the GFS-driven real time INFORM ensemble predictions that
utilize the 3-D ensemble predictions of Global Forecast System (GFS - right pathway in Figure 2-1),
the CFS-driven longer-term predictions utilize the predicted monthly surface variables of
precipitation and temperature.
29
These ensemble CFS predictions of the surface variables on a monthly scale are used to condition
the operational Ensemble Streamflow Prediction (ESP) procedure that employs historical
precipitation and temperature time series information. A description of the probabilistic approach
for conditioning and downscaling is given in Carpenter and Georgakakos (2001). Discussion of the
blending of the shorter-term (0 – 16 days) GFS-driven ensemble inflow predictions with the longer-
term (1 – 9 months) CFS-driven ensemble inflow predictions is given in HRC-GWRI (2006). We
only note here that blending is accomplished through the persistence of the snow pack and soil
moisture states in INFORM.
To generate the ensemble reservoir inflow forecasts for this year, we introduced uncertainty in the
initial snow water equivalent (uncertainty range of 15% of initial value) for all the subcatchments
modeled in INFORM to account for repeated failures of the transmission of the full ensemble of
GFS forcing from NCEP. These failures introduced uncertainty to the estimation of the snow water
equivalent of the snow pack by the INFORM hydrology component in March 2009. Table 3-1
shows the snow water equivalent of all the model subcatchments for 2009 and compares it to that of
earlier years.
Table 3-1: Snow water equivalent estimated for each INFORM hydrology model subcatchment.
(The imposed uncertainty range is shown in parenthesis)
FOLSOM DRAINAGE (Number of subcatchments=7) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 455 565 982 856 (730-980) 2 2 48 90 37 ( 32- 43) 3 225 358 630 521 (445-597) 4 35 168 179 141 (121-162) 5 340 355 676 757 (647-868) 6 0 77 42 49 ( 42- 56) 7 0 3 0 0 ( 0- 0) OROVILLE DRAINAGE (Number of subcatchments=12) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 210 245 516 451 (391-504) 2 75 102 327 159 (138-178) 3 142 186 289 188 (164-211) 4 3 91 47 0 ( 0- 0) 5 353 500 909 605 (525-677) 6 0 47 65 0 ( 0- 0)
30
7 10 130 71 92 ( 79-102) 8 0 55 6 4 ( 3- 5) 9 350 488 753 613 (532-685) 10 6 121 276 97 ( 84-108) 11 335 575 1133 775 (673-867) 12 0 27 0 0 ( 0- 0) SHASTA DRAINAGE (Number of subcatchments=10) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 m 65 67 67 ( 59- 77) 2 i 33 37 0 ( 0- 0) 3 s 128 185 113 ( 98-129) 4 s 30 40 2 ( 2- 3) 5 i 707 1044 946 (826-1080) 6 n 32 96 66 ( 58- 76) 7 g 422 798 697 (608-795) 8 ! 65 108 77 ( 67- 88) 9 ! 37 0 0 ( 0- 0) 10 ! 34 72 55 ( 48- 63) TRINITY DRAINAGE (Number of subcatchments=2) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 680 440 705 735 (642-839) 2 5 73 111 35 ( 30- 40) YUBA DRAINAGE (Number of subcatchments=4) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 407 506 507 395 (337-452) 2 0 66 73 0 ( 0- 0) 3 457 450 787 650 (556-745) 4 0 21 0 0 ( 0- 0)
The results show that the snow water equivalent in 2009 for all the reservoir drainages but Shasta
and Trinity is estimated to be lower than in 2008, especially for the southernmost drainages of the
American and Yuba Rivers included in the INFORM system. The ensemble reservoir inflow
forecasts were generated and were used as input to the INFORM decision support (DSS)
component.
A detailed analysis of the decision model assessment runs is included in Appendix B. A summary of
the findings is provided below.
The 2009 assessment portends a dry year similar to 2008. In fact, system performance is expected to
be somewhat worse than 2008 due to the already depleted reservoir storages. Depending on the
31
amount of carry-over storage target (at the end of 2009), average water deliveries are estimated to be
in the range of 2 to 2.4 million acre feet (MAF), which represents only 35% of the average 2007
amount an 30% of the average 2006. Beyond this water deliveries commitment, the likelihood that
Shasta and Oroville may fully deplete becomes appreciable. Average energy generation over the
March 15 to November 15 period ranges from 4,152 to 4,397 GWH, or approximately 55% of the
2007 generation and 45% of the 2006 generation.
In light of the above assessment results, we recommend that 2009 water management policies and
commitments be conservative and revised adaptively as inflow information becomes available.
3.3 CONCLUDING REMARKS
The second year of the NOAA transition to operations Project titled: Operational Multiscale
Forecast and Reservoir Management in Northern California was devoted to evaluating the
effectiveness of the INFORM real time forecasts and management assessments in collaboration with
operational forecast and management agencies, and production of assessments for the ability of the
system to meet requirements and satisfy objectives of operation for 2009. The system evaluation
indicated that INFORM is a useful tool for integrated forecast and management in Northern
California. The 2009 assessment for Northern California water supply management concluded that
2009 will be a difficult year in terms of meeting demand, with higher demand deficits possible than
in 2008. Adaptive management is recommended to maximize benefits for the system under these
conditions.
32
CHAPTER 4: THIRD YEAR ACTIVITIES (2009-2010)
4.1 INTRODUCTION
The third year of the NOAA TRACS project was devoted to the assessments of the 2010 period and
collaboration with the Northern California operational forecast and management agencies for the
profitable incorporation of the INFORM ensemble forecasts and management planning assessments
in operational decision making. Several meetings took place in Sacramento with the California
Nevada River Forecast Center (CNRFC) of the US National Weather Service (NWS), and with the
California Energy Commission, the California Department of Water Resources (DWR) and the U.S.
Bureau of Reclamation, Central Valley Operations (CVO), to coordinate the interpretation and use
of the INFORM outlooks and assessments in a beneficial manner for the agencies. Also, an external
evaluation of the INFORM system was commissioned by the Energy Commission with input by
end-user agencies and developers with positive results for the system. As a result of the activities of
this project and of the positive external review, additional funding by the agencies was generated to
produce enhancements to the existing INFORM system to accommodate agency requirements.
The present chapter focuses on the INFORM assessments for 2010, which were communicated to
the agencies in April 2010 and which provide one of the input pieces of information for decision
making regarding water allocation in California. The next section describes these assessments in
detail.
4.2 ASSESSMENTS FOR YEAR 2010 – DRY SEASON IN CALIFORNIA
The application described here utilizes the following input data:
· Forecasted inflows start from March 15, 2010 (88 traces, 9 month horizon, and five
locations: Clair Engle Lake, Shasta, Oroville, Folsom, and Yuba);
· Historical monthly average values are used for locations where forecasted inflows are not
available;
33
· Monthly reservoir parameters and constraints (max, min, and target storage, evaporation
rates);
· Minimum river flow requirements;
· Base monthly demands at all locations;
· Reservoir initial storages are set to their actual values on March 15, 2010.
The management system configuration is shown in Figure 1-3. As in past years, assessments used
the INFORM ensemble reservoir inflow forecasts generated using Climate Forecast System (CFS)
forcing from the National Centers of Environmental Prediction (NCEP) of NOAA.
Inflows: The forecasted monthly inflow ensembles are shown in Figures 4-1 and 4-2. The
comparisons between the forecasted inflow mean and the corresponding historical means for four
major reservoirs are plotted in Figures 4-3 through 4-6. As shown, the forecasted inflow means at
both Folsom and Trinity are higher than the historical values during high flow months and lower
during low flow months. The forecasted means at both Shasta and Oroville are lower than their
historical means. Figure 4-7 shows the forecasted basin total inflow means for 2006 to 2010. The
result for 2010 is close to that of 2007, slightly higher than those of 2008 and 2009. Overall, the
forecasts indicate that 2010 will be approximately 5% drier than the average year. Figure 4-8 shows
the initial reservoir storages on March 1 for years from 2006 and 2010, which indicates a slight
recovery from the previous year at the same time for all major reservoirs.
Water Deliveries and Energy Generation: Using the forecasted inflows, tradeoffs are generated
by changing the base demands for all locations with fractions 60% to 100%. The tradeoffs between
the total reservoir carryover storages and the system energy versus the system water deliveries are
depicted in Figures 4-9 and 4-10. As demands increase, the reservoir carryover storages decrease.
Energy generation increases as downstream demands increase because of higher reservoir releases.
The results show that the system can meet water deliveries up to 4,194 TAF, at the 92% tradeoff
point. Meeting demands beyond this level would result in significant reservoir drawdown (especially
at Shasta and Oroville) and diminished carryover storage.
Figure 4-16 compares water deliveries over the 2006-2010 five-year period assuming the actual initial
storages and approximately the same carryover storage of 7,500 TAF. As shown, the expected water
34
deliveries are reduced by 40% compared to 2006 and 33% compared to 2007, but increased by 80%
compared to 2008, 100% to 2009. Similarly, average system energy generation for 2010 is shown in
Figure 4-17, showing a decrease of 42% relative to 2006 and 35% relative to 2007, but an increase of
12% to 2008, 20% to 2009. In summary, the forecast for 2010 shows a wetter year than the two
previous years 2008 and 2009, but still drier than the historical average.
Selected reservoir elevation, release, and energy generation sequences corresponding to 3,728
thousand acre feet (TAF), 84% of base demands, are shown in Figures 4-11 through 4-13.
X2 and Delta Outflow: The X2 location sequences are shown in Figure 4-14, indicating all traces
below 80 km, the maximum constraint set in the study. The X2 location stays within this constraint
for all tradeoff points. The Delta outflow sequences are plotted in Figure 4-15.
The first month releases determined in the long range planning model are passed on to the mid
range model for generation of daily operation decisions. The daily sequences of elevation, release,
and energy generation of four major reservoirs Trinity, Shasta, Oroville, and Folsom are generated
and plotted in Figures 4-18 to 4-20. The average daily release values can be used to guide day-to-day
operations.
4.2.1 SUMMARY
The 2010 assessment portends a wetter year than the last two years 2008 and 2009, but still drier
than the historical average year. System performance is expected to improve comparing to 2009.
Depending on the amount of carry-over storage target (at the end of 2010), average water deliveries
are estimated to be in the range of 3 to 4.5 million acre feet (MAF), which represents an 80%
increase over 2008, and 100% increase over 2009. Beyond this water deliveries commitment, there
is a small likelihood that Shasta and Oroville may experience significant drawdown (below 50% of
their total capacity). Average energy generation over the March 15 to November 15 period ranges
from 4,437 to 5,520 GWH, approximately 12% higher than 2008 and 20% higher than 2009.
In light of the above assessment results, we recommend that 2010 water management policies and
commitments follow closely those of an average water year, but be adaptively revised as future
inflow information becomes available.
35
Figure 4-1: Long Range Inflow Forecasts
36
Figure 4-2: Mid Range Inflow Forecasts
37
Figure 4-3: Forecasted Inflow Mean Comparison; Trinity
Figure 4-4: Forecasted Inflow Mean Comparison; Shasta
Forecasted Inflow Means - Trinity
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10
cfs
F2010
His.
Forecasted Inflow Means - Shasta
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10
cfs
F2010
His.
38
Figure 4-5: Forecasted Inflow Mean Comparison; Oroville
Figure 4-6: Forecasted Inflow Mean Comparison; Folsom
Forecasted Inflow Means - Oroville
0
2000
4000
6000
8000
10000
12000
14000
Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10
cfs
F2010
His.
Forecasted Inflow Means - Folsom
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10
cfs
F2010
His.
39
Figure 4-7: Basin average inflow comparisons
Figure 4-8: Reservoir Initial Storages
Forecasted Basin Total Average Inflows
16962
20448
16754
13528 13981
16159
0
5000
10000
15000
20000
25000
Historical F2006 F2007 F2008 F2009 F2010
cfs
Reservoir Initial Storages On March 1st
2013
3872
2992
463
2020 20191902
3786
2997
594
20021895
1490
2660
1456
375
1532
1774
1045
2014
1385
432
1211
831
1178
2014.5
1397
426
1278
1449
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Trinity Shasta Oroville Folsom New Melones San Lius
Stor
age
(100
0 AF
)
Y2006Y2007Y2008Y2009F2010
40
Figure 4-9: Sample Tradeoff Plot 1;
Figure 4-10: Sample Tradeoff Plot 2;
Total Water Delivery vs. Carryover Storage Tradeoff
85638059
7493
6881
6263
89918460
7968
73416786
4000
5000
6000
7000
8000
9000
10000
2500 3000 3500 4000 4500 5000
Water Delivery (TAF)
Term
inal
Sto
rage
(100
0 A
F)
Y2010
Historical Mean
Total Water Delivery vs. System Energy Generation Tradeoff
4438
4689
4957
5262
5521
4555
4809
5079
5349
5620
3000
3500
4000
4500
5000
5500
6000
2500 3000 3500 4000 4500 5000
Water Delivery (TAF)
Ener
gy (G
WH
)
Y2010
Historical Mean
41
Figure 4-11: Reservoir Elevation Sequences
42
Figure 4-12: Reservoir Release Sequences
43
Figure 4-13: Reservoir Energy Generation Sequences
44
Figure 4-14: X2 Location Sequences
45
Figure 4-15: Delta Outflow Sequences
46
Figure 4-16: Mean Water Delivery Comparisons
Figure 4-17: System Energy Generation Comparisons
Simulated System Mean Water Delivery
6196
5541
20781864
3728
0
1000
2000
3000
4000
5000
6000
7000
Y2006 Y2007 Y2008 Y2009 Y2010
TAF
Simulated System Mean Energy Generation
8670
7573
4400 4152
4956
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Y2006 Y2007 Y2008 Y2009 Y2010
GW
H
47
Figure 4-18: Mid Range Reservoir Elevation Sequences
48
Figure 4-19: Mid Range Reservoir Release Sequences
49
Figure 4-20: Mid Range Energy Generation Sequences
50
Page Left Blank
51
CHAPTER 5: DISCUSSION AND CONCLUDING REMARKS
5.1 INTRODUCTION
This chapter describes the lessons learned from this research effort that aims to transition the end-
to-end INFORM (Integrated Forecast and Reservoir Management) system from the demonstration
phase to operations. It also presents recommendations that appear beneficial for the near future
along the lines of this work. The assessments are based solely on the authors experience during the
activities of the project documented in this report and they do not necessarily reflect the beliefs of
the participant agencies.
It is important to mention at the outset that INFORM provides science-based support for reservoir
managers of large multi-objective reservoir systems. As such, it targets skilled technical (engineering
and science) as well as management (planning) personnel, rather than the general public.
Furthermore, it is designed to support decisions that involve several stakeholders and decision
makers (some with conflicting objectives). The following agencies participate in the INFORM
Oversight and Implementation Committee: California Department of Water resources, California
Energy Commission, NOAA National Weather Service, NOAA Climate Program Office, U.S.
Bureau of Reclamation Central Valley Operations, and U.S. Army Corps of Engineers. The
reservoir system targeted by INFORM provides the majority of water to the California Bay Delta
region and as such it provides two-thirds of the drinking water and 7 million acres worth of
cropland irrigation for California, contributes to the hydroelectric generation of 15% of the State’s
electricity, and regulates the flow, temperature and salinity of aquatic environments to safeguard the
habitats of numerous fish, birds, mammals, and plant species.
5.2 SIX PROJECT LESSONS
Transition to operations of the end-to-end INFORM (Integrated Forecast and Reservoir
Management) system involved steps that begin with the design of the system to conform to
operational norms in close collaboration with relevant agencies, include workshops with operational
forecast and management agencies for reciprocal technology transfer with the INFORM system
52
developers, and provide relevant operational system assessments on key dates in the decision cycle
associated with water management in Northern California.
The INFORM design uses modular forms for data ingest and models so that operational models
and procedures may be used and easily compared to more advanced models for the benefit of
operational forecast and management agencies. The project experience is that this is a fundamental
requirement of any system that aims to contribute information for forecast and management
operations in a sustainable manner. This approach to design allows an incremental transition to
more advanced models starting from models with which the operational agency personnel is familiar
and with products that they have been trained to use. In addition, it helps validate synthesis
products that use operational models as components together with other more advanced component
models. Lastly, after the transition to operations is complete, it generates systems that are
sustainable by the operational agency personnel without significant continuing involvement of the
developers. For the INFORM system, operational data ingested are the ensemble predictions of the
National Center of Environmental Predictions (NCEP) Global Forecast System (GFS) and Climate
Forecast System (CFS), the operational hydrologic models and update procedures of the California
Nevada River Forecast Center (CNRFC) for snow and soil water continuous accounting, and the
operational spreadsheets for reservoir management of the Bureau of Reclamation and the California
Department of Water Resources.
A second important lesson is that, prior to beginning the transition process, operational agencies
wish to see non-trivial comparisons of the current operational system output to corresponding
output from the new system to consider supporting transitions to operations. For the INFORM
system Carpenter and Georgakakos 2001, Yao and Georgakakos 2001, HRC-GWRI 2007 provide such
comparisons with existing procedures and show in a conclusive manner that it has significant skill in
most cases. As a result California Agencies financially supported the transition with a ratio of
funding of approximately 3 (California Agencies) to 1 (NOAA).
A third lesson relates to the role that the output of models, systems and procedures has in decision
making pertaining to reservoir management. That is, this output provides advisory information for
decision makers, while decisions may also rely on additional information and subjective judgment, as
well as non-quantifiable objectives. Thus, the more the systems to be transitioned facilitate this
advisory role for their products and are amenable to use with additional information, the more
53
suited they are perceived to be for transition to operations. The INFORM system produces risk-
based non-inferior trade-offs for use by decision makers with additional information to arrive at
decisions regarding reservoir releases.
A fourth lesson is that decision makers assign importance to the existence of uncertainty measures
associated with the products of the system that is to be transitioned to operations. Both operational
forecast and management agencies are concerned with uncertainty in recent years when gradual
changes are seen in climatic variables (e.g., earlier snowmelt in the Sierra Nevada and thus more
variable flows to be accommodated by the California reservoirs). Ensemble predictions of inflows
and other hydrologic variables are important as are commensurate risk-based decision component
outputs. Reliability of such uncertainty measures is important and validations of such reliability are
important prerequisites of a transition to operations. For instance, management agencies required
that prior to using them in operations the ensemble reservoir inflows were based and were
sanctioned by the operational forecast agencies. The INFORM system design was modified to allow
alignment of the INFORM hydrologic model states to those of the California Nevada River
Forecast Center operational model states. Also, California Agencies sought and obtained external
independent reviews of the INFORM system effectiveness and utility for California operations.
Such reviews were favorable for the INFORM project. Lastly, the effects of upstream regulation to
downstream ensemble streamflow predictions were identified as important by the operational
forecast agencies. Changes in the INFORM structure are necessary (not yet made) to enhance the
reliability of the ensemble reservoir inflow forecasts during the dry California summer and fall.
The fifth lesson has to do with the variable of the response of the operational forecast and
management agencies to this transition to operations effort. Although INFORM is an integrated
system with both forecast and reservoir management components, individual agencies were
interested in specific components of the system because they believe that the other components
have existing operational counterparts that are producing satisfactory results. It is not clear if this is
an individual’s preference or agency policy. Nevertheless, it suggests that large integrated systems
will meet favorable agency response if they are modular not only in design but also in the products
generated, and furthermore even in the demonstration phase if they make these products available to
the interested agencies in formats that are easy to digest and conform to the agency product types.
After substantial agency discussion that involved the Oversight and Implementation Committee, the
INFORM system was modified to generate both mean areal precipitation and reservoir inflow
54
ensemble products, and several products from the decision component pertaining to agency defined
management objectives.
The sixth lesson reinforces the value of frequent workshops to producing usable and sustainable
systems for operations. Throughout the project such meetings led to several changes in the initial
system and system-product configuration and lent credence to the initial feasibility studies and
system skill and utility. They also led to significant science cooperation and technology transfer
pertaining to several topics of interest both for the agencies participating and for the developers.
For example, this cooperative effort led to developing a common basis for system assessments,
identifying the current basis of multi-agency decision making in California, isolating a significant part
of a very large and complex California water management system which could be modeled by
INFORM for operational support, considering the process from ensemble reservoir inflow
predictions to risk-based trade-offs, identifying uncertainty pathways within INFORM, defining
system retrospective studies for validation, contributing to real time product validation, and others.
5.3 TRANSITION TO OPERATIONS PROJECT CHALLENGES
There were several challenges that the authors faced during the efforts to transition INFORM to
operations. Accommodating the suggestions (presented in the previous section in the form of
lessons) was a significant but necessary challenge for successful progress toward the operational use
of INFORM products. However, there was a significant challenge that the project faced which was
only partly accommodated, mainly with external resources.
The main challenge pertains to on-going changes to operational systems and products by both
operational forecast and management agencies, which made the incorporation of operational
components within the INFORM system to be transitioned (see previous section for the need) a
moving target. This is particularly so for the changes in the National Centers of Environmental
Prediction CFS products and the changes of the California Nevada River Forecast Center model
configuration (number and configuration of modeled basins and upstream regulation effects to
downstream ensemble streamflow predictions) and, most importantly, forecast system architecture
(from the National Weather Service River Forecast System to the Common Hydrologic Prediction
System (CHPS)). Such changes, although necessary for improving the products produced by these
55
agencies, were not foreseen at the time of the development of the project and required substantial
changes in INFORM system design and software architecture (as well as training of the developers
in the new systems), accruing significant additional cost. In addition, they prolonged the transition
process substantially beyond the developers control as external agency timetables are not aligned
with the transition timetable set for this project. The option of freezing the system to a particular
input product or software architecture effectively stops the transition to operations process, and it is
thus undesirable. INFORM developers sought additional funding to accommodate these changes
and were able to cover within the project timetable a part of the needed changes as documented in
this report (Chapter 2). However, additional changes are necessary and these are presented in the
recommendations section below.
5.4 RECOMMENDATIONS
Perhaps the most important recommendation to NOAA Climate Program Office is to continue
supporting activities of agency communication and training pertaining to the profitable use of the
INFORM ensemble forecasts and water management assessments for Northern California. The
time and resources invested are bearing fruit (this is based on third party independent reviews
solicited by the California Energy Commission) and they will help improve management in a
changing climate and demand environment in California. The authors believe that the lessons
learned in the INFORM Project transition-to-operations process highlight the fact that some of
impediments to the application of probabilistic weather and climate forecast information to end-user
systems are quite general. Thus the larger community dealing with these issues would benefit from
INFORM's near-unique experience of having produced an end-to-end system suitable for
operations and the many layers of detail (too often ignored) underlying the idea of using weather and
climate forecast information in real, large and managed end-user systems. These benefits could be
realized by having NOAA recognize and publicize the achievements and lessons learned in the
INFORM project and its transition-to-operations.
The second recommendation regards the needed changes to INFORM to complete the transition to
operations phase. This pertains to the ingest of the new CFS ensemble predictions (complete
variable fields of three-dimensional ensemble prediction suitable for dynamic downscaling) so that
56
the INFORM forecast component can take advantage of these fields for more reliable predictions in
the 0 to 30 day range for Northern California. It also pertains to the changes required to
accommodate the new CHPS architecture of the National Weather Service operational forecast
system. This work is now being supported by Phase II of the INFORM project sponsored by the
California Energy Commission (2009-2012), while requisite training is supported by the Technology
Transfer Program of the Hydrologic Research Center.
A third recommendation concerns the transition to operations of a sustainable and useful procedure
for accounting for upstream regulation effects on ensemble reservoir inflow predictions. The Office
of Hydrologic Development of the National Weather Service has supported the necessary research
and demonstration phase, and the transition to operations of the resultant methodologies remains to
be implemented as a new transition to operations project. It is recommended that this be supported
as needed to enhance credibility of NWS predictions mainly during the dry periods of the year
(summer and fall in Northern California).
57
CHAPTER 6: REFERENCES
Carpenter, T.M., and K.P. Georgakakos, 2001: Assessment of Folsom Lake Response to Historical
and Potential Future Climate Scenarios, 1, Forecasting. Journal of Hydrology, 249,148-175.
Georgakakos, K.P., Graham, N.E., Carpenter, T.M., Georgakakos, A.P., and Yao, H., 2005:
Forecasts and multiobjective reservoir management in Northern California. EOS 86(12), 122, 127.
Georgakakos, K.P., Graham, N.E., Carpenter, T.M., Georgakakos, A.P., Yao, H., and Kistenmacher,
M., 2008. Operational Multiscale Forecast and Reservoir Management in Northern California, First Year Progress
Report. HRC TN 34, Hydrologic Research Center, San Diego, CA, 18pp.
HRC-GWRI, July 2006. Integrated Forecast and Reservoir Management for Northern California:
System Development and Initial Demonstration. California Energy Commission, PIER Energy-
Related Environmental Research (Aquatic Resources). CEC-500-02-008, 244pp. (Available on
line: http://www.energy.ca.gov/pier/final_project_reports/CEC-500-2006-109.html)
HRC-GWRI, 2007. Integrated Forecast and Reservoir Management (INFORM) Demonstration Project, Winter
’06 – ’07 Operations. HRC TN 29, Hydrologic Research Center, San Diego, CA, 37 pp.
Sperfslage, J. A. and K. P. Georgakakos, 1996: Operational Implementation of the Hydrologic Forecast System
(HFS) Operation as part of the National Weather Service River Forecast System (NWSRFS). HRC Technical
Report No. 12 Hydrologic Research Center, San Diego, California 213pp.
Yao, H, and A. Georgakakos, "Assessment of Folsom Lake Response to Historical and Potential
Future Climate Scenarios," Journal of Hydrology, 249, 176-196, 2001
58
Page Left Blank
59
APPENDIX A: PRELIMINARY ASSESSMENT OF THE INFORM
PHASE I FORECAST-DECISION SYSTEM RESULTS VS. OBSERVED DATA FOR THE 2006, 2007, AND 2008 SEASONS
A.1 INTRODUCTION AND OVERVIEW
The purpose of this analysis is to assess the value of the INFORM DSS forecast-decision system
(http://www.energy.ca.gov/pier/project_reports/CEC-500-2006-109.html) in the planning and
management of the northern California reservoir system (Figure 1-2).
This should only be viewed as a preliminary and tentative assessment because (a) the INFORM
project is at its first of two planned development and demonstration phases and not all system
components have been completed, (b) a good part of the necessary actual data are unavailable at the
time of this writing, and (c) the time and resources for this effort are inadequate for a
comprehensive evaluation. It is stressed that a major objective of the second INFORM phase
(scheduled to begin in January, 2009) is to perform such a comparison and assess the INFORM
system benefits in pragmatic detail. To be meaningful, this comparison should be done after careful
planning of realistic INFORM applications in support of existing decision processes, with evaluation
of results jointly performed by stakeholder agency experts and INFORM developers.
The actual system data used in this assessment has been obtained from the USBR and DWD
websites. USBR data relate to water deliveries, reservoir storages, and energy generation for projects
managed by USBR. DWR data pertain to inflow data for all reservoirs. Energy generation data for
Oroville (state project) is unavailable at the time of this writing. The comparison is based on data
from March to November of 2006 and 2007, and March to October of 2008.
Based on this data availability, the comparison is carried out with respect to inflows at main system
nodes, water deliveries, carry-over storages of the major reservoirs, and energy generation.
60
A.2 OBSERVED DATA
A.2.1 INFLOW DATA AND FORECAST ASSESSMENT
The INFORM DSS requires inflows at the following nodes (Figure 1-2):
• Trinity
• Whiskeytown
• Shasta
• Keswick-Wilkins
• Oroville
• Folsom
• Yuba
• Sacramento Miscellaneous
• Eastside Streams
• Delta Miscellaneous Creeks
• New Melones
• San Joaquin River (SJR)
However, inflow forecasting models have been developed during the first INFORM phase for the
following locations only:
• Trinity
• Shasta
• Oroville
61
• Folsom
• Yuba
Forecasting models for all other nodes are scheduled for development during the second project
phase. Thus, in the INFORM forecast-decision runs to be presented, at the locations where models
are not yet available, the forecasts comprise historical seasonal means. This is expected to introduce
discrepancies which can be significant as the modeled tributary inflows on average comprise about
58% of the total. Figure A-2.1 displays the average historical natural inflows at different system
nodes, highlighting the modeled nodes with solid bars.
Figure A-2.1: Historical Average Inflows (Solid Bars Correspond to Modeled Nodes)
The observed inflows for the assessment seasons are posted on the California Data Exchange
Center website at http://www.usbr.gov/mp/cvo/deliv.html. The original data are in daily time
step. Monthly average sequences and other statistics are derived using daily data.
Historical Average Inflows (1968-1996)
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Trin
ity
Whi
skey
tow
n
Sha
sta
Kes
wic
k-W
ilkin
s
Oro
ville
Fols
om
Yub
a
Sac
ram
ent M
isc
Eas
tsid
eS
tream
s
Del
ta M
isc
Cre
eks
New
Mel
ones
SJR
cfs
62
As part of this assessment, nine month forecasts are issued for the five above-mentioned nodes
(Trinity, Shasta, Oroville, Folsom, and Yuba) with a starting date of March 1, 2006, 2007, and 2008.
These forecasts are not updated in the subsequent months as would be done in practice, highlighting
the preliminary nature of this assessment. Furthermore, at the present stage of INFORM system
development (end of Phase I), probabilistic rather than dynamic downscaling is performed to
produce basin precipitation and temperature fields from the large scale operational Climate Forecast
Model (CFS) predictions to drive the hydrologic components of INFORM for the basins modeled.
Thus, the generated precipitation and temperature fields reflect the characteristics of the historical
climatology of above and below average events more than the actual large scale forecast field
characteristics from the CFS.
Figures A-2.2., A-2.3, and A-2.4 present the monthly sequences of historical means, forecasted
mean/maximum/minimum, and observed sequences at the four nodes corresponding to the major
system reservoirs (Trinity, Shasta, Oroville, and Folsom).
The March 1, 2006, forecasts indicated an above average water year for Trinity, Shasta, and Folsom,
and an average water year for Oroville (with the exception of the March mean forecast which was
predicted above average). The observed data showed that all locations registered above average
flows. The forecast results are overall reliable in that they contain the actual sequences throughout
the forecast horizon. Forecast reliability is expected to increase in a real time application where
forecasts are routinely updated as the year evolves.
The March 1, 2007, forecasts indicated an above average water year for Trinity and Folsom, average
for Shasta, and below average for Oroville. The observed data showed that all locations received
below average flows. However, again, all observed sequences fell within the forecast ensemble.
Lastly, the March 1, 2008, forecasts indicated an average water year for Trinity, above average for
Folsom, and below average for Shasta and Oroville. The observed data showed that all locations
received below average flows.
System wide, the mean forecasts suggested a wet year for 2006, average year for 2007, and a dry year
for 2008. With the exception of 2007, the actual observations were consistent with model
predictions. Furthermore, all observed inflows were within the forecasted ranges.
63
Figure A-2.2: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow Sequences, 2006
2006 Forecasts; Trinity
0
2000
4000
6000
8000
10000
12000
Mar-06 Apr-06 May-06
Jun-06 Jul-06 Aug-06
Sep-06 Oct-06 Nov-06
cfs
Max.
Avg
Min
Obs.
His.
2006 Forecasts; Shasta
0
5000
10000
15000
20000
25000
30000
35000
40000
Mar-06 Apr-06 May-06
Jun-06 Jul-06 Aug-06
Sep-06 Oct-06 Nov-06
cfs
Max.
Avg
Min
Obs.
His.
2006 Forecasts; Oroville
0
5000
10000
15000
20000
25000
30000
35000
Mar-06 Apr-06 May-06 Jun-06 Jul-06 Aug-06 Sep-06 Oct-06 Nov-06
cfs
Max.
Avg
Min
Obs.
His.
2006 Forecasts; Folsom
0
5000
10000
15000
20000
25000
Mar-06 Apr-06 May-06 Jun-06 Jul-06 Aug-06 Sep-06 Oct-06 Nov-06
cfs
Max.
Avg
Min
Obs.
His.
64
Figure A-2.3: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow Sequences, 2007
2007 Forecasts; Trinity
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07
cfs
Max.
Avg
Min
Obs.
His.
2007 Forecasts; Shasta
0
5000
10000
15000
20000
25000
30000
35000
40000
Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07
cfs
Max.
Avg
Min
Obs.
His.
2007 Forecsats; Oroville
0
5000
10000
15000
20000
25000
Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07
cfs
Max.
Avg
Min
Obs.
His.
2007 Folsom
0
5000
10000
15000
20000
25000
Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07cf
s
Max.
Avg
Min
Obs.
His.
65
Figure A-2.4: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow Sequences; 2008
A.2.2 ACTUAL WATER DELIVERIES
System water diversions represent the total amount water taken out of the system. The observed
monthly data were taken from the USBR website ( http://www.usbr.gov/mp/cvo/deliv.html).
Table 21 of the website (Central Valley Project Diversions) contains the monthly diversions in acre-
feet for all projects operated by USBR. The following nodes are related to the INFORM DSS
system:
• Contra Costa
• Delta-Mendota
• Actual Fed Dos Amigos
66
• Madera
• Friant-Kern
• Corning
• Folsom-South
• Tehama-Colusa
• Thermalito Forebay
Historical water withdrawals for Thermalito Afterbay are provided by the Department of Water
Resources (DWR) of California.
The total system diversions are taken to be equal to the sum of diversions from all above nodes.
Thus, the actual monthly system water diversions (taken to represent deliveries) from March to
November for 2006 and 2007, and March to October for 2008 are shown in Figure A-2.5. The
corresponding total volumes are plotted in Figure A-2.6.
The water delivery shows a clear seasonal pattern. It starts to increase in May, peaks in July, and
gradually reduces to the lowest level in November. The total actual deliveries in 2006 were 5810
TAF (thousand acre-feet). In 2007, this amount was reduced to 4695 TAF, and in 2008 even further
to 3368 TAF. Thus, actual deliveries were reduced by almost 50% from 2006 to 2008.
67
Figure A-2.5: System Monthly Water Diversion Sequences
Figure A-2.6: Observed Total Diversions
Observed Monthly System Diversions
0
200
400
600
800
1000
1200
Mar Apr May Jun Jul Aug Sep Oct Nov
AF200620072008
Total Diversions from 2006 to 2008
0
1000
2000
3000
4000
5000
6000
7000
2006 2007 2008
AF
68
2.3 ACTUAL CARRY-OVER STORAGES
Actual reservoir storage sequences are obtained from the USBR website at
http://www.usbr.gov/mp/cvo/deliv.html. For comparison purposes, this analysis pertains to the
five major reservoirs: Trinity, Shasta, Oroville, Folsom, and New Melones. The initial time refers to
the beginning of March, and the terminal time refers to the end of November for 2006 and 2007,
and the end of October for 2008. The total system storage is the sum of the storages of all major
reservoirs. The observed data are plotted in Figure A-2.7. A clear downward trend is observed for
the terminal storages with a 50% reduction from 2006 to 2008.
Figure A-2.7: Observed Total System Initial and Terminal Storage
A.2.4 ACTUAL ENERGY GENERATION
Actual monthly energy generation are posted on the USBR website at
http://www.usbr.gov/mp/cvo/deliv.html For comparison purposes, this analysis includes
generation amounts only from five major reservoirs: Trinity, Oroville, Shasta, Folsom, and New
Melones. Data for Oroville is provided by DRW. The total generation values from March to
System Carryover Storages of Major Reservoirs
0
2000
4000
6000
8000
10000
12000
2006 2007 2008
TAF
Initial
Carryover
69
November for 2006 and 2007, and March to October for 2008 are shown in Figure A-2.8. This
figure shows a 33% reduction in actual energy generation from 2006 (6759 GWH) to 2007 (4479
GWH), and a more than 50% reduction from 2006 to 2008 (2975 GWH).
Figure A-2.7: Observed System Energy Generation
A.2.5 INFORM DSS RESULTS
The INFORM DSS uses the forecast ensembles from March to November to generate tradeoffs
among the system objectives such as water deliveries, energy generation, and carryover storage,
given applicable operational constraints. Three system runs are carried out for the years 2006, 2007,
and 2008. The tradeoffs illustrate the many feasible combinations of the above objectives and form
the basis for the comparison presented herein. More specifically, in each run, the tradeoff point that
matches the actual deliveries is identified and the carry-over storage and energy generation amounts
are compared with the actual corresponding quantities. This is a biased comparison against the INFORM
results, which in this analysis are generated by an initial nine-month run without subsequent monthly
Energy Generation from Major Plants
0
1000
2000
3000
4000
5000
6000
7000
8000
2006 2007 2008
GW
H
70
updating as would be its intended practical use. The comparison is depicted in Figures A-2.8 (carry-over
storage) and A-2.9 (energy generation) and is discussed below:
· An above average forecast was issued for 2006. Based on the forecasts and initial reservoir
storage, the INFORM DSS estimated that the system could support up to 7500 TAF of
water deliveries. The tradeoff curve was generated for a water demand range from 3850 to
7700 TAF (Figure A-2.10). The actually met demand of 5810 TAF was close to the first
tradeoff point 5887 TAF. Figures A-2.8 and A-2.9 show that the actual system carry-over
storage and energy generation are within the predicted ranges, with the later practically equal
to the forecasted mean.
· A relatively average to dry forecast ensemble was issued for 2007. Based on the forecasts
and high initial reservoir storage, the INFORM DSS estimated that the system can support
up to 5700 TAF water supply. The actual demand met was 4695 TAF which was close to
the first tradeoff point of 4694 TAF (Figure A-2.11). The actually observed inflows were in
the low forecast range, which resulted in a lower than the mean forecasted carryover storage
(Figure A-2.8). However, this value is still within the predicted range. The same comment
applies for energy generation (Figure A-2.9).
· A dry forecast was issued for 2008. With already reduced initial storage, the INFOM DSS
estimates that the system can only support up to 3000 TAF demand. The tradeoff curve
ranged from 700 to 3800 TAF. The actual water delivery (to October) of 3367 TAF is close
to the fifth tradeoff point (3315 TAF; Figure A-2.12). Figures A-2.8 and A-2.9 show very
good correspondence between average predicted and observed carry-over storage and energy
generation.
71
Figure A-2.8: System Carry-over Storage Comparison
Figure A-2.9: System Energy Generation Comparison
Carryover Storage Comparisons
0
2000
4000
6000
8000
10000
12000
2006 2007 2008
TAF
ObservedDSS AVGDSS MinDSS MAX
Energy Generation from Major Plants
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
2006 2007 2008
GW
H
ObservedDSS AVGDSS MINDSS MAX
72
Figure A-2.10: INFORM Tradeoff: Total Water Delivery vs. System Carryover Storage; 2006
Figure A-2.11: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2007
Tradeoff Curve for 2006
0
1000
2000
3000
4000
5000
6000
7000
8000
3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000
Water Delivery (TAF)
Syst
em C
arry
over
Sto
rage
(TA
F)
Tradeoff Curve for 2007
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000
Water Delivery (TAF)
Syst
em C
arry
over
Sto
rage
(TA
F)
73
Figure A-2.12: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2008
A.3 CONCLUSIONS
The previous preliminary comparison indicates that the information provided by the INFORM
system in operational planning and management of the northern California river and reservoir
system is relevant, reliable, and decision worthy. This is corroborated by the favorable comparison
of the predicted and actually observed system outputs (water deliveries, carry-over storages, and
energy generation amounts) over the March to November 2006, 2007, and 2008 seasons.
The value of the INFORM forecast-decision system ultimately depends on whether the information
it generates provides management agencies with good appreciation of the benefits and risks
associated with different decisions and helps them adopt those that serve well the interests of the
system stakeholders. In the second project phase, more detailed assessments are planned to quantify
the system response with and without this information.
It is expected that the value of the INFORM system will increase after the second project phase
which aims at completing its development and implementing it operationally. More specifically,
forecast improvements are anticipated as a result of the following planned activities:
Tradeoff Curve for 2008
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 500 1000 1500 2000 2500 3000 3500
Water Delivery (TAF)
Syst
em C
arry
over
Sto
rage
(TA
F)
74
· Enlarging the domain of hydro-meteorological modeling to include the entire application
area and Delta tributary flows;
· Enhancing the downscaling of the Climate Forecast System (CFS) model forecasts with a
downscaling procedure as used for the shorter term ensemble operational forecasts from the
Global Forecast System (GFS). Both GFS and CFS are run by NOAA at the National
Centers of Environmental Prediction (NCEP) and are the primary operational
meteorological forecast input to the INFORM forecast-management system; and
· Implementing the forecast component on a cluster computer so that more ensembles from
GFS and CFS may be processed in real time for more accurate evaluation of forecast
uncertainty for short and long lead times.
Likewise, decision system improvements are expected due to the following planned activities:
· Developing and incorporating a system-wide daily resolution model;
· Incorporating a river temperature model for assessing fishery needs;
· Managing system uncertainties in a spatially optimal manner; and
· Working with the stakeholder agencies to integrate the INFORM forecast-decision system
within the existing decision processes and to assess/demonstrate its value for planning and
management.
75
APPENDIX B: OPERATIONAL MULTISCALE FORECAST AND
RESERVOIR MANAGEMENT IN NORTHERN CALIFORNIA, ASSESSMENTS 2009
The contribution of Georgia Water Resources Institute to the NOAA TRACS project is to develop
and demonstrate the utility of Decision Support Systems (DSS) for reservoir management. The DSS
consists of databases, interfaces, and various application programs interlinked to provide meaningful
and comprehensive information to decision makers. During the current NOAA funding period,
GWRI efforts focused on assessing the system performance over the period from March 15, 2009,
to November 15, 2009. Inflow forecasts were provided by HRC starting from March 15, 2009. The
findings are discussed below.
B.1 INTEGRATED INFORM DSS AND FORECASTING MODELS - AN OVERVIEW
The INFORM DSS includes three modeling layers (Figure 1-3) designed to support decisions
pertaining to various temporal scales and objectives. The three modeling layers include (1) turbine
load dispatching (which models each turbine and hydraulic outlet and has hourly resolution over a
horizon of one day), (2) short/mid range reservoir control (which has a daily resolution and a
horizon of one month), and (3) long range reservoir control (which has a monthly resolution and a
horizon of up to one year).
Both the long range control model and the mid/short range control model use inflow forecasts as
inputs. The integration of the decision models and inflow forecasting models are done through data
exchange. The forecasted inflows are saved in a pre-formatted Excel file. The DSS provides easy
tools to read the data in the Excel file and save it into the database. The DSS also provides tools to
plot and validate the forecasting results.
The long range control model is designed to consider long range issues such as whether water
conservation strategies are appropriate for the upcoming year using the provided hydrologic
forecasts. As part of these considerations, the DSS would quantify several tradeoffs of possible
interest to the management agencies and system stakeholders. These include, among others, relative
76
water allocations to water users throughout the system (including ecosystem demands), reservoir
coordination strategies and target levels, water quality constraints, and energy generation targets.
This information would be provided to the forum of management agencies (the planning
departments) to use it as part of their decision process together with other information. After
completing these deliberations, key decisions would be made on monthly water supply contracts,
reservoir releases, energy generation, and reservoir coordination strategies.
The short/mid range control model considers the system operation at finer time scales. The
objectives addressed are more operational than planning and include flood management, water
supply, and power plant scheduling. This model uses hydrologic forecasts with a daily resolution and
can quantify the relative importance of, say, upstream versus downstream flooding risks, energy
generation versus flood control, and other applicable tradeoffs. Such information is again provided
to the forum of management agencies (the operational departments) to use it within their decision
processes to select the most preferable operational policy. Such policies are revised as new
information on reservoir levels and flow forecasts comes in. The model is constrained by the long
range decisions, unless current conditions indicate that a departure is warranted.
The three modeling layers address planning and management decisions. The scenario/policy
assessment model addresses longer term planning issues such as increasing demands, infrastructure
change (water transfers options), potential hydro-climatic changes, and mitigation measures. The
approach taken in this DSS layer is to simulate and inter-compare the system response under various
inflow, demand, development, and management conditions.
Altogether, the INFORM DSS provides a comprehensive modeling framework responsive to the
information needs of the decision making process at all relevant time scales.
B.2 INFORM DSS RESULTS FOR 2009 FORECASTS
The application described here utilizes the following input data:
· Forecasted inflows start from March 15, 2009 (88 traces, 9 month horizon, and five
locations: Clair Engle Lake, Shasta, Oroville, Folsom, and Yuba;
77
· Historical monthly average values are used for locations where forecasted inflows are not
available (Table B-1);
· Monthly reservoir parameters and constraints (max, min, and target storage, evaporation
rates; Table B-2);
· Minimum river flow requirements (Table B-3);
· Base monthly demands at all locations (Table B-4);
· Reservoir initial storages are set to their actual values on March 15, 2009.
The forecasted monthly inflow ensembles are shown in Figures B-1 and B-2. The comparisons
between the forecasted inflow mean and the corresponding historical means for four major
reservoirs are plotted in Figures B-3 through B-6. As shown, the forecasted inflow means at both
Folsom and Trinity are higher than the historical values during high flow months and lower during
low flow months. The forecasted means at both Shasta and Oroville are lower than their historical
means. Figure B-7 shows the forecasted basin total inflow means for 2006 to 2009. The results for
2009 are similar to those for 2008. Overall, the forecasts indicate that 2009 will be approximately
18% drier than the average year. Figure B-10 shows the initial reservoir storages on March 1 for
years from 2006 and 2009. The current year has the lowest initial storage values for all major
reservoirs, which adds to the system stress.
Using the forecasted inflows, tradeoffs are generated by changing the base demands for all locations
with fractions 10% to 50%. The tradeoffs between the total reservoir carryover storages and the
system energy versus the system water deliveries are depicted in Figures B-9 and B-10. As demands
increase, the reservoir carryover storages decrease. Energy generation increases as downstream
demands increase because of higher reservoir releases. The results show that the system can meet
water deliveries up to 2,330 TAF, at the 50% tradeoff point. Meeting demands beyond this level
would result in significant reservoir drawdown (especially at Shasta and Oroville) and diminished
carryover storage.
78
The reservoir sequences and other system outputs corresponding to all tradeoff points are saved in
the INFORM DSS database. Selected reservoir elevation, release, and energy generation sequences
corresponding to 1,860 thousand acre feet (TAF), 40% of base demands are shown in Figures B-11
through B-13. The X2 location sequences are shown in Figure B-14, indicating all traces below 80
km, the maximum constraint set in the study. The X2 location stays within this constraint for all
tradeoff points. The Delta outflow sequences are plotted in Figure B-15.
Figure B-16 compares water deliveries over the 2006-2009 four-year period, assuming the actual
initial storages and the same carryover storage of 7,500 TAF. As shown, the expected water
deliveries are reduced by 70% compared to 2006 and 65% compared to 2007. Similarly, average
system energy generation for 2009 is shown in Figure B-17, showing a decrease of 53% relative to
2006 and 45% relative to 2007. Generally, the impact of reduced inflows on water deliveries and
energy generation is not equally proportional because of the Bay Delta water requirements.
The first month releases determined in the long range planning model are passed on to the mid
range model for generation of daily operation decisions. The daily sequences of elevation, release,
and energy generation of four major reservoirs Trinity, Shasta, Oroville, and Folsom are generated
and plotted in Figures B-18 to B-20. The average daily release values can be used to guide day-to-
day operations.
B.3 SUMMARY
The 2009 assessment portends a dry year similar to 2008. In fact, system performance is expected to
be somewhat worse than 2008 due to the already depleted reservoir storages. Depending on the
amount of carry-over storage target (at the end of 2009), average water deliveries are estimated to be
in the range of 2 to 2.4 million acre feet (MAF), which represents only 35% of the average 2007
amount an 30% of the average 2006. Beyond this water deliveries commitment, the likelihood that
Shasta and Oroville may fully deplete becomes appreciable. Average energy generation over the
March 15 to November 15 period ranges from 4,152 to 4,397 GWH, or approximately 55% of the
2007 generation and 45% of the 2006 generation.
79
In light of the above assessment results, we recommend that 2009 water management policies and
commitments be conservative and revised adaptively as inflow information becomes available.
80
Figure B-1: Long Range Inflow Forecasts
81
Figure B-2: Mid Range Inflow Forecasts
82
Figure B-3: Forecasted Inflow Mean Comparison; Trinity
Figure B-4: Forecasted Inflow Mean Comparison; Shasta
Forecasted Inflow Means - Trinity
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09
cfs
F2009
His.
Forecasted Inflow Means - Shasta
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09
cfs
F2009
His.
83
Figure B-5: Forecasted Inflow Mean Comparison; Oroville
Figure B-6: Forecasted Inflow Mean Comparison; Folsom
Forecasted Inflow Means - Oroville
0
2000
4000
6000
8000
10000
12000
14000
Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09
cfs
F2009
His.
Forecasted Inflow Means - Folsom
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09
cfs
F2009
His.
84
Figure B-7: Basin average inflow comparisons
Figure B-8: Reservoir Initial Storages
Forecasted Basin Total Average Inflows
16962
20448
16754
13528 13981
0
5000
10000
15000
20000
25000
Historical F2006 F2007 F2008 F2009
cfs
Reservoir Initial Storages On March 1st
2013
3872
2992
463
2020 20191902
3786
2997
594
20021895
1490
2660
1456
375
1532
1774
1045
2014
1385
432
1211
831
0
500
1000
1500
2000
2500
3000
3500
4000
4500
Trinity Shasta Oroville Folsom New Melones San Lius
Stor
age
(100
0 AF
)
Y2006Y2007Y2008Y2009
85
Figure B-9: Sample Tradeoff Plot 1;
Figure B-10: Sample Tradeoff Plot 2;
Total Water Delivery vs. Carryover Storage Tradeoff
89938500
79947488
6982
101499643
91368629
8122
4000
5000
6000
7000
8000
9000
10000
11000
0 500 1000 1500 2000 2500
Water Delivery (TAF)
Term
inal
Sto
rage
(100
0 A
F)
Y2009
Historical Mean
Total Water Delivery vs. System Energy Generation Tradeoff
3439
3675
3917
4152
4397
3424
3681
3952
4218
4486
3000
3200
3400
3600
3800
4000
4200
4400
4600
0 500 1000 1500 2000 2500
Water Delivery (TAF)
Ener
gy (G
WH
)
Y2009
Historical Mean
86
Figure B-11: Reservoir Elevation Sequences
87
Figure B-12: Reservoir Release Sequences
88
Figure B-13: Reservoir Energy Generation Sequences
89
Figure B-14: X2 Location Sequences
90
Figure B-15: Delta Outflow Sequences
91
Figure B-16: Mean Water Delivery Comparisons
Figure B-17: System Energy Generation Comparisons
Simulated System Mean Water Delivery
6196
5541
20781864
0
1000
2000
3000
4000
5000
6000
7000
Y2006 Y2007 Y2008 Y2009
TAF
Simulated System Mean Energy Generation
8670
7573
4400 4152
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
Y2006 Y2007 Y2008 Y2009
GW
H
92
Figure B-18: Mid Range Reservoir Elevation Sequences
93
Figure B-19: Mid Range Reservoir Release Sequences
94
Figure B-20: Mid Range Energy Generation Sequences
95
Table B-1: Monthly Average Inflows for Selected Locations (TAF)
Month Whisktown Keswick-Wilkens
Sacrament Misc
Eastside Streams
Delta Misc Creeks New Melones SJR
Jan 8. -211.27 -100. 80.67 25.5 76. 133.Feb 4. -299.69 -220. 60.44 25.5 43. 31.Mar 2. -370.28 -330. 20.72 29. 34. 33.Apr 1. -267.47 -175. 21.89 19. 33. 28.May 1. -117.56 45. 28.71 11.1 31. 33.Jun 2. -125. -15. 33.2 0.8 30. 71.Jul 2. -31.24 121. 30.74 0.9 30. 62.Aug 4. 564.46 981. 21.52 1.2 30. 63.Sep 8. 841.7 1465. 21.52 1.8 30. 78.Oct 12. 1767.58 2482. 40.03 32.3 40. 94.Nov 45. 1021. 1763. 67.33 17.4 70. 103.Dec 16. 74.65 328. 146.34 15.4 110. 126.
96
Table B-2: Reservoir Monthly Parameters
Name Month Smax (TAF) Smin (TAF) Starget (TAF) Evap Rate (feet)
Clair Engle Jan 2287.00 312.63 2287.00 0.17
Clair Engle Feb 2287.00 312.63 2287.00 0.13
Clair Engle Mar 2287.00 312.63 2287.00 0.20
Clair Engle Apr 2287.00 312.63 2287.00 0.39
Clair Engle May 2287.00 312.63 2287.00 0.51
Clair Engle Jun 2287.00 312.63 2287.00 0.58
Clair Engle Jul 2287.00 312.63 2287.00 0.76
Clair Engle Aug 2287.00 312.63 2287.00 0.71
Clair Engle Sep 2287.00 312.63 2287.00 0.60
Clair Engle Oct 2287.00 312.63 2287.00 0.30
Clair Engle Nov 2287.00 312.63 2287.00 0.15
Clair Engle Dec 2287.00 312.63 2287.00 0.09
WhiskeyTown Jan 237.90 200.00 205.70 0.17
WhiskeyTown Feb 237.90 200.00 205.70 0.13
WhiskeyTown Mar 237.90 200.00 205.70 0.20
WhiskeyTown Apr 237.90 200.00 237.90 0.39
WhiskeyTown May 237.90 200.00 237.90 0.51
WhiskeyTown Jun 237.90 200.00 237.90 0.58
WhiskeyTown Jul 237.90 200.00 237.90 0.76
WhiskeyTown Aug 237.90 200.00 237.90 0.71
WhiskeyTown Sep 237.90 200.00 238.00 0.60
WhiskeyTown Oct 237.90 200.00 230.00 0.30
WhiskeyTown Nov 237.90 200.00 205.70 0.15
WhiskeyTown Dec 237.90 200.00 205.70 0.09
Shasta Jan 4552 1168 4552 0.17
Shasta Feb 4552 1168 4552 0.13
97
Shasta Mar 4552 1168 4552 0.20
Shasta Apr 4552 1168 4552 0.39
Shasta May 4552 1168 4552 0.51
Shasta Jun 4552 1168 4552 0.58
Shasta Jul 4552 1168 3882 0.76
Shasta Aug 4552 1168 3252 0.71
Shasta Sep 4552 1168 3252 0.60
Shasta Oct 4552 1168 3872 0.30
Shasta Nov 4552 1168 4252 0.15
Shasta Dec 4552 1168 4552 0.09
Oroville Jan 3538 855 3458 0.17
Oroville Feb 3538 855 3538 0.13
Oroville Mar 3538 855 3538 0.20
Oroville Apr 3538 855 3538 0.39
Oroville May 3538 855 3538 0.51
Oroville Jun 3538 855 3343 0.58
Oroville Jul 3538 855 3163 0.76
Oroville Aug 3538 855 3163 0.71
Oroville Sep 3538 855 3163 0.60
Oroville Oct 3538 855 3163 0.30
Oroville Nov 3538 855 3163 0.15
Oroville Dec 3538 855 3163 0.09
Folsom Jan 975 83 805 0.17
Folsom Feb 975 83 975 0.13
Folsom Mar 975 83 975 0.20
Folsom Apr 975 83 975 0.39
Folsom May 975 83 975 0.51
Folsom Jun 975 83 975 0.58
Folsom Jul 975 83 700 0.76
98
Folsom Aug 975 83 575 0.71
Folsom Sep 975 83 575 0.60
Folsom Oct 975 83 575 0.30
Folsom Nov 975 83 575 0.15
Folsom Dec 975 83 675 0.09
New Melones Jan 2420 273 2230 0.17
New Melones Feb 2420 273 2420 0.13
New Melones Mar 2420 273 2420 0.20
New Melones Apr 2420 273 2420 0.39
New Melones May 2420 273 2420 0.51
New Melones Jun 2420 273 2270 0.58
New Melones Jul 2420 273 1970 0.76
New Melones Aug 2420 273 1970 0.71
New Melones Sep 2420 273 1970 0.60
New Melones Oct 2420 273 1970 0.30
New Melones Nov 2420 273 1970 0.15
New Melones Dec 2420 273 2040 0.09
Tulloch Jan 67 57 57 0.00
Tulloch Feb 67 57 57 0.00
Tulloch Mar 67 57 58 0.00
Tulloch Apr 67 57 60 0.00
Tulloch May 67 57 67 0.00
Tulloch Jun 67 57 67 0.00
Tulloch Jul 67 57 67 0.00
Tulloch Aug 67 57 67 0.00
Tulloch Sep 67 57 62 0.00
Tulloch Oct 67 57 57 0.00
Tulloch Nov 67 57 57 0.00
Tulloch Dec 67 57 57 0.00
99
San Luis Jan 2027 450.00 1000.00 0.17
San Luis Feb 2027 631.60 1464.02 0.13
San Luis Mar 2027 748.10 1806.84 0.20
San Luis Apr 2027 835.60 1975.02 0.39
San Luis May 2027 879.92 1976.43 0.51
San Luis Jun 2027 694.72 1546.00 0.58
San Luis Jul 2027 442.12 1062.95 0.76
San Luis Aug 2027 181.12 642.62 0.71
San Luis Sep 2027 9.72 352.64 0.60
San Luis Oct 2027 8.32 312.90 0.30
San Luis Nov 2027 115.02 354.13 0.15
San Luis Dec 2027 286.72 514.21 0.09
Table B-3: Monthly Minimum and Target River Flows
Name Month Rmin (cfs) Rtarget (cfs)
Lewiston Jan 300 300
Lewiston Feb 300 300
Lewiston Mar 300 300
Lewiston Apr 300 300
Lewiston May 3939 300
Lewiston Jun 2507 783
Lewiston Jul 1102 450
Lewiston Aug 450 450
Lewiston Sep 450 450
Lewiston Oct 373 0
Lewiston Nov 300 300
Lewiston Dec 300 300
100
Clear Creek Jan 150 150
Clear Creek Feb 200 200
Clear Creek Mar 200 200
Clear Creek Apr 200 200
Clear Creek May 200 200
Clear Creek Jun 200 200
Clear Creek Jul 200 200
Clear Creek Aug 200 200
Clear Creek Sep 200 200
Clear Creek Oct 200 200
Clear Creek Nov 90 90
Clear Creek Dec 90 90
Spring Creek Jan 325 325
Spring Creek Feb 306 306
Spring Creek Mar 2749 2749
Spring Creek Apr 252 252
Spring Creek May 813 813
Spring Creek Jun 1681 1681
Spring Creek Jul 2602 2602
Spring Creek Aug 2114 2114
Spring Creek Sep 2017 2017
Spring Creek Oct 1138 1138
Spring Creek Nov 504 504
Spring Creek Dec 244 244
Keswick Jan 3250 3250
Keswick Feb 3250 3250
Keswick Mar 3250 3250
Keswick Apr 8000 8000
Keswick May 9600 9600
101
Keswick Jun 11000 11000
Keswick Jul 14500 14500
Keswick Aug 12000 12000
Keswick Sep 5500 5500
Keswick Oct 7200 7200
Keswick Nov 5700 5700
Keswick Dec 3250 3250
Wilkins Jan 0 0
Wilkins Feb 0 0
Wilkins Mar 0 0
Wilkins Apr 5000 5000
Wilkins May 5000 5000
Wilkins Jun 5000 5000
Wilkins Jul 5000 5000
Wilkins Aug 5000 5000
Wilkins Sep 5000 5000
Wilkins Oct 5000 5000
Wilkins Nov 0 0
Wilkins Dec 0 0
FeatherBelowThermalito Jan 1250 0
FeatherBelowThermalito Feb 1250 0
FeatherBelowThermalito Mar 1250 0
FeatherBelowThermalito Apr 1250 0
FeatherBelowThermalito May 2030 0
FeatherBelowThermalito Jun 0 2706
FeatherBelowThermalito Jul 0 5692
FeatherBelowThermalito Aug 5040 5156
FeatherBelowThermalito Sep 0 4386
FeatherBelowThermalito Oct 1980 2683
102
FeatherBelowThermalito Nov 1750 1815
FeatherBelowThermalito Dec 1250 0
AmericanRiverbelowNimbus Jan 800 0
AmericanRiverbelowNimbus Feb 800 0
AmericanRiverbelowNimbus Mar 1000 0
AmericanRiverbelowNimbus Apr 1500 0
AmericanRiverbelowNimbus May 2300 0
AmericanRiverbelowNimbus Jun 1800 0
AmericanRiverbelowNimbus Jul 0 0
AmericanRiverbelowNimbus Aug 0 0
AmericanRiverbelowNimbus Sep 0 0
AmericanRiverbelowNimbus Oct 0 0
AmericanRiverbelowNimbus Nov 1000 0
AmericanRiverbelowNimbus Dec 800 0
Goodwin Jan 175 175
Goodwin Feb 150 150
Goodwin Mar 268 268
Goodwin Apr 760 760
Goodwin May 800 800
Goodwin Jun 561 561
Goodwin Jul 396 396
Goodwin Aug 352 352
Goodwin Sep 240 240
Goodwin Oct 200 200
Goodwin Nov 200 200
Goodwin Dec 200 200
DeltaExit Jan 6001 6001
DeltaExit Feb 11398 11398
DeltaExit Mar 11401 11401
103
DeltaExit Apr 7848 7848
DeltaExit May 9319 9319
DeltaExit Jun 7092 7092
DeltaExit Jul 6505 6505
DeltaExit Aug 4261 4261
DeltaExit Sep 3008 3008
DeltaExit Oct 4001 4001
DeltaExit Nov 4655 4655
DeltaExit Dec 4505 4505
Table B-4: Monthly Demands
Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov DecThermolito 35 0 11 67 189 178 200 178 78 95 104 71
Folsom Pumping 4 4 4 7 8 12 13 12 10 7 5 4Folsom South Canal 1 1 1 1 2 3 4 4 3 2 1 1
OID/SSJID 0 0 14 60 90 90 95 95 74 14 0 0CVP Contractors 0 0 0 0 0 0 0 0 0 0 0 0
CCWD 14 17 18 18 14 14 13 13 13 10 11 13Barker Slough 2 2 1 2 4 5 7 7 6 5 3 3
Federal Tracy PP 258 233 258 250 135 169 270 268 260 258 250 258Federal Banks On-Peak 0 0 0 0 0 0 28 28 28 0 0 0
State Banks PP 390 355 241 68 108 125 271 278 238 175 193 390State Tracy PP 0 0 0 0 0 0 0 0 0 0 0 0
Delta Mendota Canal 30 60 100 120 190 220 270 240 180 110 40 30Federal Dos Amigos 40 50 60 70 110 180 238 178 68 30 30 30Federal O'Neil to Dos
Amigos 0 1 1 1 1 2 2 1 0 0 0 0San Felipe 6 6 10 15 19 20 21 20 13 11 8 8
South Bay/San Jose 2 2 2 5 5 7 7 8 7 12 8 6State Dos Amigos 105 127 158 105 348 348 423 388 269 229 196 61
Delta Consumptive Use -56 -37 -10 63 121 191 268 252 174 118 55 2
Freeport Treatment Plant 14 13 14 12 12 12 12 13 12 12 12 13
104
Table B-5: Initial Reservoir Storages on March 15, 2009
Reservoirs Max. Storage Min. Storage Initial Storage Ini Act. Sto. Fraction (%)Clair Engle Lake 2287 312.63 1123.2 41.05
Whiskeytown 237.9 200 212.07 31.85Shasta 4552 1168 2529 40.22Oroville 3538 855 1739.5 32.97Folsom 975 83 647.3 63.26
New Melones 2420 273 1278 46.81Tulloch 67 57 55 -20.00
San Luis 2027 0 931 45.93