A Modeling Framework for Improved Agricultural Water Supply Forecasting George Leavesley, Colorado State University, Olaf David, Colorado State University, David Garen, NWCC NRCS-USDA, Jolyne Lea, NWCC NRCS-USDA, Jim Marron, NWCC NRCS-USDA, Tom Pagano, NWCC NRCS-USDA, Tom Perkins, NWCC NRCS-USDA, and Michael Strobel, NWCC NRCS-USDA, ABSTRACT – The National Water and Climate Center (NWCC) of the USDA Natural Resources Conservation Service is moving to augment seasonal, regression-equation based water supply forecasts with distributed-parameter, physical process models enabling daily, weekly, and seasonal forecasting using an Ensemble Streamflow Prediction (ESP) methodology. This effort involves the development and implementation of a modeling framework, and associated models and tools, to provide timely forecasts for use by the agricultural community in the western United States where snowmelt is a major source of water supply. The framework selected to support this integration is the USDA Object Modeling System (OMS). OMS is a Java-based modular modeling framework for model development, testing, and deployment. It consists of a library of stand-alone science, control, and database components (modules), and a means to assemble selected components into a modeling package that is customized to the problem, data constraints, and scale of application. The framework is supported by utility modules that provide a variety of data management, land unit delineation and parameterization, sensitivity analysis, calibration, statistical analysis, and visualization capabilities. OMS uses an open source software approach to enable all members of the scientific community to collaboratively work on addressing the many complex issues associated with the design, development, and application of distributed hydrological and environmental models. A long-term goal in the development of these water-supply forecasting capabilities is the implementation of an ensemble modeling approach. This would provide forecasts using the results of multiple hydrologic models run on each basin. The mission of the NWCC is to lead the development and transfer of water and climate information and technology which support natural resource conservation. The NWCC develops water supply forecasts for hundreds of basins in the western United States which the agricultural community uses to optimize water use during the irrigation season. To address the agricultural communities’ requests for more information on the volume and timing of water availability, and to improve forecast accuracy, the NWCC is developing the capability to use distributed-parameter, physical process models to provide daily, weekly, and seasonal forecasts using an Ensemble Streamflow Prediction (ESP) methodology. The primary objective of this proposal is the development and implementation of a modeling framework, and associated models and tools, to provide timely and improved water supply forecasts for use by the agricultural community in the western US. Research Related Aspects Integrated Forecast System The Object Modeling System (OMS) Component-Based Modeling Model Building and Component Connectivity The Functions of the National Water and Climate Center include: Natural Resource Planning Support Provide water supply forecasts Provide water and climate analysis, information, and services for NRCS, partners and customers Data Acquisition and Management Operate the Snowpack Telemetry (SNOTEL) data collection system for the United States Design and manage an on-line, quality controlled, national database for NRCS and partners in support of farm, watershed and river basin scale planning Technology Innovation Assess and select technologies required to address resource concerns Adapt appropriate technologies Partnerships and Joint Ventures Technology Transfer Modeling Projects Component Library Assembled Model Parameter Editor Component Editor Output Analysis Component Repository Use Evaluation of de-biasing methods for ESP forecasts. Methods to be evaluated include quantile mapping, regression, adjustment by a constant, and time-series modeling. Evaluation of alternative climate scenarios in ESP. Use of climate generators for creating ensembles of climate scenarios will be compared with current ESP methodologies. Evaluation of alternative precipitation distribution methods. Alternative distribution methods will be evaluated for a range of climatic and physiographic regions, and data availability. Components for XYZ, detrended Kriging, and inverse distance methods are currently available. Development of an ensemble of models for forecasting. A long-term goal in the development and implementation of an ensemble modeling approach. The initial model being used in the system is the USGS model PRMS. OMS Framework Model Component Attribute Snow Snowmelt Soil Connect Explore Palette Use Model Support and Analysis Components Components typically represent a unique concept in a model like a physical process, a management practice, or a specific data input. Component attributes are fundamental data objects which a component reads input from and/or writes output to. Attributes are the global data elements in a model and provide communication among components. Components can be implemented in Java, Fortran, C, and C++. OMS is a Java-based modeling framework for model development, testing, and deployment. It consists of a library of stand-alone science, control, and database components, and a means to assemble selected components into a modeling package. The framework is supported by utility modules that provide a variety of data management, land unit delineation and parameterization, sensitivity analysis, calibration, statistical analysis, and visualization capabilities. OMS uses an open-source software approach to enable all members of the scientific community to collaboratively address the many complex issues associated with the design, development, and application of distributed hydrological and environmental models. A Model Builder tool provides a visual interface for assembling components into models. Components are managed regarding their data flow requirements, and attributes linkages among components are established. A set of control components are available to support a variety of component linking strategies. Components can be local, or explored and retrieved from a remote repository. Collaborative model development is supported by the USDA Collaborative Software Development Laboratory (CoLab). Hay and Umemoto, 2006, Multiple-Objective Step-Wise Calibration using Luca: U.S. Geological Survey OFR A variety of components are available to support the creation, visualization, and analysis of model input and model simulation results. These include graphical and statistical analysis tools, the USGS Luca multiple-objective step- wise parameter calibration tool, and Ensemble Streamflow Prediction (ESP) functionality. Visualization of spatially distributed basin characteristics, model parameters, and input variables, as well as the visualization and animation of spatially distributed model output are provided by the GeoWind tool. GeoWind is being developed using NASA World Wind and GeoTools software. LUCA Graphical Tools GeoWind The NWCC is developing the capability to use distributed-parameter, physical process models to assist in addressing a wide variety of water-user requests for more information on the volume and timing of water availability, and to improve forecast accuracy. This modeling effort will compliment the statistical water- supply forecasting approach currently being used to provide seasonal volume forecasts for water availability. Statistical water-supply forecasts will continue into the future using tools such as the Visually Interactive Prediction and Estimation Routines (VIPER). OMS is supporting the development and implementation of the forecast models and tools. A real-time data retrieval and update system is being developed to retrieve SNOTEL and COOP meteorological data from the NOAA Regional Climate Center Network to update model data files. This procedure will be coupled with the OMS implementation of the Ensemble Streamflow Prediction (ESP) methodology to provide forecasts of runoff peaks, low flow, volume, and timing. Watersheds identified as highly managed, with competing water uses, will be some of the first to be modeled. Six to eight basins will be used in the initial application. However, once the system is fully operational, all areas where there is a need for additional forecast products will be included in the system. Once fully implemented, the integrated model system will provide timely and improved water-supply forecasts for agriculture and for all water managers and interested user groups in the western US. VIPER The National Water and Climate Center (NWCC) OMS ESP C21A-0497 Associated with the development and implementation of this forecast system are a number of research issues that will also be addressed. These include: