Riverine Modeling November 13-15, 2013 Decision Support System - - PowerPoint PPT Presentation

▶
riverine modeling
SMART_READER_LITE
LIVE PREVIEW

Riverine Modeling November 13-15, 2013 Decision Support System - - PowerPoint PPT Presentation

Feb 01, 2023 486 likes •697 views

Instream Flow Technical Team Meeting - Riverine Modeling November 13-15, 2013 Decision Support System Options Prepared by R2 Resource Consultants 11/13/2013 DRAFT SUBJECT TO REVISION 2 Decision Support System - Define Decision:

slide-1
SLIDE 1

Instream Flow Technical Team Meeting - Riverine Modeling November 13-15, 2013 Decision Support System Options

Prepared by R2 Resource Consultants

DRAFT – SUBJECT TO REVISION 11/13/2013

slide-2
SLIDE 2

Decision Support System - Define

  • Decision: Choice among options on dam
  • perations (Existing Conditions, Optimum

Energy Output, OptA, OptB,....,)

  • Support System: Framework for evaluating
  • ptions based on “EVALUATION METRICS”

2 DRAFT – SUBJECT TO REVISION 11/13/13

slide-3
SLIDE 3

Decision Support System - Goal

  • “The goal of a decision support system is not

to make a decision, but rather to reduce the complexity of information and focus attention on tradeoffs involved in the decision.” (USGS: Auble, et al 2009, DSS for Gunnison River)

3

DRAFT – SUBJECT TO REVISION 11/13/13

slide-4
SLIDE 4

Decision Support System - RSP

  • Primary tool for “Instream Flow Study

Integration” (RSP Section 8.5.4.8)

  • “Evaluate the benefit and potential impacts
  • f alternative operational scenarios”
  • Focus attention on attributes of highest

priority for evaluation of operational scenarios

4 DRAFT – SUBJECT TO REVISION 11/13/13

slide-5
SLIDE 5

DSS: Important Considerations

  • Integration
  • Focus
  • Simplify
  • Transparency
  • Complexity of System/Time/Budget

5 DRAFT – SUBJECT TO REVISION 11/13/13

slide-6
SLIDE 6

DSS: Structured Decision Making

  • Requirements:
  • 1. Explicit management alternatives
  • Actions: Operational scenarios
  • 2. Explicit, quantifiable objectives
  • Values: Maximize energy output, Maximize chum spawning

habitat, etc.

  • 3. Models that predict #2 from #1
  • Beliefs: Assumptions, Data

6 DRAFT – SUBJECT TO REVISION 11/13/13

slide-7
SLIDE 7

DSS: Evaluation Metrics

  • A set of joint values
  • Must be quantifiable and spatially and temporally

explicit

  • Could be combined using “multicriteria methods”

7 DRAFT – SUBJECT TO REVISION 11/13/13

slide-8
SLIDE 8

DSS: Potential Approaches

  • Manual Matrix Method
  • USGS DSS for water management

– Gunnison, Upper Yakima, Delaware Rivers

  • Decision Analysis/Bayesian Belief Networks

8

DRAFT – SUBJECT TO REVISION 11/13/13

slide-9
SLIDE 9

Matrix Methods

  • Operational and Flow Scenarios

Evaluation Metrics

  • Some spatial and/or temporal variability included

– Future 50 years is weighted average of dry, average, wet years responses – Averaged over Focus Areas in MR

  • Uncertainties/assumptions are dealt with ahead of time

– Choice of “average” flow year; choice of models; HSC methods

  • Result = decision matrix comparing all operational

scenarios for all EMs

9

DRAFT – SUBJECT TO REVISION 11/13/13

slide-10
SLIDE 10

10

EXAMPLE SUBSET of a Matrix

Resource Area Temporal Scale Spatial Scale Evaluation Metrics (EXAMPLE) Existing Conditions OS1 OS2 OS3 Power Nov-March average

  • ver expected 50 year

flow n/a Power Generation (MWh) Hydrologic Nov-March minimum

  • ver expected 50 year

flow n/a 2Day Low Flow (cfs) Riparian Years 10-20 Geomorphic Reach Floodplain Plant Community Colonization Area (acres) Resident Fish Averaged over expected 50 year flow Geomorphic Reach Grayling weighted usable spawning habitat (ft2) Ice processes Median date at year 50 n/a Timing of ice breakup Anadromous Fish Averaged over expected 50 year flow Focus Area Coho effective spawning/incubation habitat area in FA-104 (Whiskers Slough), averaged over expected 50 year flow. Anadromous Fish Averaged over expected 50 year flow Focus Area Chinook effective spawning/incubation habitat area in FA-104 (Whiskers Slough), averaged over expected 50 year flow. Anadromous Fish Averaged over expected 50 year flow Focus Area Chinook juvenile rearing habitat area in FA-104 (Whiskers Slough), averaged over expected 50 year flow. Anadromous Fish Averaged over expected 50 year flow Focus Area Coho juvenile outmigration habitat area in FA-104 (Whiskers Slough), averaged over expected 50 year flow. Anadromous Fish Averaged over expected 50 year flow Focus Area Chinook adult migration habitat area in FA-104 (Whiskers Slough), averaged over expected 50 year flow.

DRAFT – SUBJECT TO REVISION 11/13/13

slide-11
SLIDE 11

Making a Decision From a Matrix

  • Potentially hundreds of metrics in multiple

resource categories

  • No single optimal choice
  • Three options:

– Multiple Criteria Decision Analysis – Focus on KEY metrics – Leave a huge matrix

11

DRAFT – SUBJECT TO REVISION 11/13/13

slide-12
SLIDE 12

Multiple Criteria Decision Analysis

  • Provide weighting of metrics (could be all

equal).

– Example 1: Number of metrics representing significant decline over existing conditions – Example 2: Weighted average of change from existing conditions, with anadromous fish habitat parameters double weighted.

  • Decision rule – based methods

– Example: Maximize power generation such that flow never drops below a minimum cfs criteria during juvenile rearing season.

12

DRAFT – SUBJECT TO REVISION 11/13/13

slide-13
SLIDE 13

Making a Decision From a Matrix

  • Three options:

– Multiple Criteria Decision Analysis – Focus on KEY metrics – Leave a huge matrix

13

DRAFT – SUBJECT TO REVISION 11/13/13

slide-14
SLIDE 14

USGS DSS Approach

  • USGS uses VBA macros in MSExcel to compute

similar decision matrix and show graphics automatically

  • User inputs flow scenarios and can change

some model parameters

  • Outputs many tables and graphs, as well as

decision matrix (red, yellow, green)

  • Can accommodate multiple criteria analysis

14

DRAFT – SUBJECT TO REVISION 11/13/13

slide-15
SLIDE 15

Summary of Matrix/DSS Methods

  • Has been successfully used for FERC

licensing

  • All decision metrics in one matrix
  • Uncertainty is not explicit

– Assumptions are vetted ahead of time or through re-running models under changes to assumptions (after the fact).

15

DRAFT – SUBJECT TO REVISION 11/13/13

slide-16
SLIDE 16

DSS: Other Options

  • Decision analysis methods for including

uncertainty

  • Bayesian Belief Networks (BBNs)

16

DRAFT – SUBJECT TO REVISION 11/13/13

slide-17
SLIDE 17

Decision Analysis/Inclusion of Uncertainty

17

Resource Area Temporal Scale Spatial Scale Evaluation Metrics (EXAMPLE) Assumptions Probability Existing Conditions OS1 OS2 HSC1 0.2 HSC2 0.8 HSC3 0.2 Weighted Average Anadromous Fish Averaged over expected 50 year flow Focus Area Coho effective spawning/incubation habitat area in FA-104 (Whiskers Slough), averaged over expected 50 year flow.

DRAFT – SUBJECT TO REVISION 11/13/13

slide-18
SLIDE 18

Example BBN for Spawning Habitat

18

EffSpawnHab 0 to 100 100 to 1000 1000 to 5000 5000 to 10000 36.9 6.40 53.3 3.35 1900 ± 2000 HSC HSC1 HSC2 HSC3 10.0 80.0 10.0 Operations OS1 OS2 OS3 33.3 33.3 33.3 Flow1 Wet Year Somewhat Wet Moderate Year Somewhat Dry Dry Year 5.00 20.0 45.0 25.0 5.00 18200 ± 9700 DamOutflow 0 to 3000 3000 to 6000 6000 to 12000 12000 to 24000 24000 to 40000 10.0 30.0 40.0 15.0 5.00 9400 ± 7400 UpwellingLocations UP1 UP2 UP3 10.0 80.0 10.0

DRAFT – SUBJECT TO REVISION 11/13/13

slide-19
SLIDE 19

DSS: conclusions

  • Matrix method will be done
  • Amount of automation/software not yet

determined

  • May include some explicit uncertainties if

possible

  • BBN interesting, but not planned due to

complexities/timing

19

DRAFT – SUBJECT TO REVISION 11/13/13