The Modular Modeling System (MMS): A Toolbox for Water- and Environmental-Resources Modeling and Management

1
The Modular Modeling System (MMS) A Toolbox for
Water- and Environmental-Resources Modeling and
Management
G.H. Leavesley, S.L. Markstrom, R.J. Viger, and
L.E. Hay
U.S. Geological Survey, Denver, CO USA
2
STARTING POINTS

• There are no universal models
• Models for different purposes require different
  levels of detail and comprehensiveness
• Appropriate model process conceptualizations are
  a function of problem objectives, data
  constraints, and spatial and temporal scales of
  application

3
Modeling Conundrum

• Experimental science builds on hypothesis testing
  and interpretation based on earlier published
  hypotheses and results
• Modelers tend to build from the ground up because
  existing models are not well designed for
  incremental improvement by others

4
TOOL BOX MODELING VIEWS

• Research Model Developer Complex Detail
• Application Model Developer
• Model User
• Resource Manager
• Policy Maker Condensed
  Analysis

5
LEVELS OF MODULAR DESIGN
Single Purpose

• PROCESS
• MODEL
• FULLY COUPLED MODELS
• LOOSELY COUPLED MODELS
• RESOURCE MANAGEMENT DECISION SUPPORT SYSTEMS
• ANALYSIS AND SUPPORT TOOLS

Multi-objective, Complex
6
LEVELS OF MODULAR DESIGN

• PROCESS
• MODEL
• FULLY COUPLED MODELS
• LOOSELY COUPLED MODELS
• RESOURCE MANAGEMENT DECISION SUPPORT SYSTEMS
• ANALYSIS AND SUPPORT TOOLS

7
(No Transcript)
8
CRITERIA AND RULES FOR GOOD MODULE DESIGN
Modules should

• relate directly to real world components or
  processes
• have input and output variables that are
  measurable values
• communicate solely via these input and output
  variables

Reynolds J.F., and Acock, B., 1997, Modularity
and genericness in plant and ecosystem models
Ecological Modeling 94, p 7-16
9
(No Transcript)
10
Model Building Tool - XMBUILD
11
(No Transcript)
12
SELECTED MODELS AND MODULES IN MMS

• OTHER
• TOPMODEL
• Hydro-17 (NWS snowmelt)
• Sacramento Model (NWS)
• Snowmelt Runoff Model (SRM) (ARS)
• ENNS Model (modified HBV, Austria)

• USGS
• PRMS
• GSFLOW
• WEBMOD

13
LEVELS OF MODULAR DESIGN

• PROCESS
• MODEL
• FULLY COUPLED MODELS
• LOOSELY COUPLED MODELS
• RESOURCE MANAGEMENT DECISION SUPPORT SYSTEMS
• ANALYSIS AND SUPPORT TOOLS

14
Ground Water - Surface Water Coupling
DAFLOW
PRMS
MODFLOW
15
GSFLOW -- Coupled PRMS, MODFLOW, SFR, and
Unsaturated Zone Models
PRMS to SFR
Unsaturated Zone Model PRMS to UNSAT UNSAT
to MODFLOW
PRMS to MODFLOW
MODFLOW to SFR
Streamflow
16
LEVELS OF MODULAR DESIGN

• PROCESS
• MODEL
• FULLY COUPLED MODELS
• LOOSELY COUPLED MODELS
• RESOURCE MANAGEMENT DECISION SUPPORT SYSTEMS
• ANALYSIS AND SUPPORT TOOLS

17
LOOSELEY COUPLED MODELS
MMS Model
Off-the-shelf Model
Database
Data Management Interface (DMI)
18
LOOSELEY COUPLED MODELS
Hydraulics Model
19
LOOSELEY COUPLED MODELS
PRMS
Channel Geometry Analysis Program (CGAP)
Hydraulics Model
Habitat Suitability Index
20
Coupled Model Results
21
LOOSELEY COUPLED MODELS
MMS Model
Off-the-shelf Model
Database
Data Management Interface (DMI)
Model Management Interface (MMI) XML
22
OBJECT USER INTEFACE (OUI)
23
LEVELS OF MODULAR DESIGN

• PROCESS
• MODEL
• FULLY COUPLED MODELS
• LOOSELY COUPLED MODELS
• RESOURCE MANAGEMENT DECISION SUPPORT SYSTEMS
• ANALYSIS AND SUPPORT TOOLS

24
Recreation
Watershed and River Systems Management Program
Municipal Industrial
Irrigation
Hydropower
Research and development of decision support
systems and their application to achieve an
equitable balance among water resource issues.
Riparian Habitat
Endangered Species
25
(No Transcript)
26
(No Transcript)
27
RiverWare Lower Colorado River
28
Upper Gunnison River DSS
Hydromet Real-time climate data feed
Hydrologic Database
DMI
DMI
RiverWare Reservoir and River System Operations
Model
DMI
29
Generic DSS Framework for a Wide Range of
Management Issues
Any Resource Management Model
Climate Data Source
Any Database
DMI
DMI
DMI
MMI
Object User Interface Interface for data
visualization and modeling
Modular Modeling System Physical Process Models
30
WARSMP Basins
Currently Active
Gunnison, Truckee, Upper Rio Grande, Yakima
Under Development
San Juan, Umatilla, Upper Columbia
Future
Bitterroot, Carson, Central Platte, Lower Rio
Grande, Salmon
31
LEVELS OF MODULAR DESIGN

• PROCESS
• MODEL
• FULLY COUPLED MODELS
• LOOSELY COUPLED MODELS
• RESOURCE MANAGEMENT DECISION SUPPORT SYSTEMS
• ANALYSIS AND SUPPORT TOOLS

32
GIS WEASEL

• Delineation
• Only requires elevation Grid as input
• Interactively delineate
• Area of Interest
• Many kinds of features
• Streams
• Elevation bands
• Landuse
• Contributing areas
• Topographic index

33
DIGITAL DATABASES
STATSGO Soils (USDA)
Vegetation Type (USFS)
Satellite SW Radiation (U Md)
Vegetation Density (USFS)
Land Use-Land Cover (USGS)
Monthly PET
34
AUTOMATED PARAMETER ESTIMATION USING THE GIS
WEASEL
35
GIS WEASEL

• Parameterization
• 200 methods available
• Easily add custom methods

• Exploit many types of data
• Produce maps and ASCII files of parameters

• Configure recipes
• Apply to feature maps

36
MODEL RECIPE FILE
PARAMETER ESTIMATION METHOD
PARAMETER
37
Multi-step Calibration
38
Multi-step calibration visualization
39
ANALYSIS and SUPPORT TOOLS
Currently Available
Statistical and Graphical Analyses Parameter
Optimization Parameter Sensitivity Analysis
Beta Testing
Shuffle Complex Evolution Optimization Multi-Objec
tive COMplex Evolution Algorithm
(MOCOM) Generalized Likelihood Uncertainty
Estimation (GLUE) Visualization
40
Forecast Methodologies

• - Historic data as analog for the future
• Ensemble Streamflow Prediction (ESP)
• Synthetic time-series
• Weather Generator
• - Atmospheric model output
• Dynamical Downscaling
• Statistical Downscaling

41
MMS Development is Accomplished Through
Collaborative Multi-disciplinary Efforts
42
USGS
WATER, ENERGY, AND BIOGEOCHEMICAL BUDGETS (WEBB)
PROGRAM
43
WEBMOD
PRMS
National Weather Service - Hydro17
TOPMODEL
44
Use Chemical and Isotopic Tracers to Evaluate
Simulated Flow Paths and Residence Times
Next step Use tracers to constrain the
hydrologic solutions
45
Coupled WEBMOD and PHREEQC
46
Effects of Vegetation and Vegetation Change on
Water and Energy Partitioning
Natural
Managed
Satellite image of the Rodeo and Chediski fires
before they merged into one fire, eventually
burning some 470,000 acres
Fool Creek Watershed, Fraser Experimental Forest,
Colorado (714 acres)
Issues Process, Scale, Parameter Estimation
47
Vegetation Management Model
US Forest Service, Missoula, Mt

• Simulates vegetation patterns and processes
  emphasizing the dynamics of landscape level
  change.

• Produces ensemble output

48
(No Transcript)
49
(No Transcript)
50
Integrating Science with Resource Management
through Collaborative Approaches and Adaptive
Modeling Systems
Resource Management
Science Synthesis
Modeling Framework Development

51
Initial DSS Development Area
USGS, NPS, USFS, BLM, BIA, CO State University, N
AZ University
52
Integrating Water-Resource and Socio-Economic
Models

• Population and Demand Forecasts
• Water Markets
• (change in ownership in perpetuity)
• Water Banking
• (lease options over some period of time)

Issue The better the prediction of the spatial
and temporal distribution of water, the better
the markets can perform.
This requires the full integration of
physical-process and socio-economic models to
optimize economic and institutional analyses.
53
Adding Water Banking to a Water Resources
Decision-Support System
Hydromet Real-time climate data feed
Hydrologic Database
DMI
DMI
RiverWare Reservoir and River System Operations
Model
DMI
DMIs
DMI
Object User Interface Interface for data
visualization and modeling
Modular Modeling System Precipitation/Runoff
Model (PRMS)
54
MMS UNIVERSITY COLLABORATION

• University of Arizona
• NSF funded -- Sustainability of semi-Arid
  Hydrology and Riparian Areas (SAHRA)
• University of Colorado
• NOAA funded -- Coupling of hydrologic and
  atmospheric models to provide short- to long-term
  forecasts
• University of Nevada Reno
• Desert Research Institute funded Watershed
  modeling and coupled SW/GW modeling
• Colorado State University
• Interagency MOU funded GeoLEM and Parameter
  estimation, optimization, and sensitivity
  analysis tools

55
Working with the NRCS to incorporate the USGS
Modular Modeling System into the NRCS forecasting
toolbox
56
Integrated Framework Development

• Collaborative effort to integrate the Object
  Modeling System (OMS) and the Modular Modeling
  System (MMS)

- ARS - USGS - NRCS -Friedrich
Schiller University, Germany
Branding
57
(No Transcript)
58
Other Model Integration Approaches

• U.S.
• BASINS (EPA)
• FRAMES (DOE, NRC, DOD, EPA)
• DIAS (DOE)
• XMS (WMS, GMS, SMS, - EMS Inc.)
• OMS (USDA)
• International
• HarmonIT (Europe)
• Catchment Modeling Toolkit (Australia)

59
MMS SUMMARY

• Integrated toolbox approach to model development
  and application for water- and environmental-resou
  rces management
• Easily configured and/or enhanced for
  user-specific needs
• Supports multi-disciplinary model integration for
  decision support systems
• Flexible framework approach enables the
  incorporation of continuing advances in science,
  databases, and computer technology as well as
  changes in management objectives
• Open source software design allows many to share
  resources, expertise, knowledge, and costs

60
Chicken Soup for the Modeling Soul
POINT
A fool with a tool is still a fool.
System Development magazine
61
MORE INFORMATION

• http//wwwbrr.cr.usgs.gov/mms
• http//wwwbrr.cr.usgs.gov/weasel
• http//wwwbrr.cr.usgs.gov/warsmp
• http//oms.ars.usda.gov
• http//www.iscmem.org