An Integrated modeling environment within the CUAHSI Hydrologic Information System

Document Type

Article

Journal/Book Title/Conference

AGU Fall Meeting Abstracts

Publisher

Agu

Publication Date

1-1-2010

First Page

1237

Last Page

1237

Abstract

Modeling complicated hydrologic systems often requires the integration of disparate data and models. The CUAHSI Hydrologic Information System targets the problem of integrating data by using a service-oriented architecture and data exchange standards to make heterogeneous databases appear to an end user as a single data resource. Similar integration problems exist with hydrologic models. If one wishes to analyze a problem that requires logic within multiple models, the challenge becomes how to couple those models so that they are able to exchange data during model runtime. A solution to this problem has been proposed through the Open Modeling Interface (OpenMI) standard for integrating hydrologic models. Building from the OpenMI standard, we have created a modeling environment within the CUAHSI Hydrologic Information System HydroDesktop application for performing integrated modeling. The modeling environment, which we have named HydroModeler, allows for loose coupling of model, analysis, and data components. We provide components for reading and writing data to the HydroDesktop database, as well as a number of example model configurations for demonstration and education purposes. An advantage of adopting the OpenMI standard is that it enables one to include OpenMI compliant models written by other groups within HydroModeler and, likewise, components written specifically for HydroModeler — e.g. the HydroDesktop database writer and reader components - can be used within other OpenMI-compliant modeling environments. Through the process of building HydroModeler we have investigated topics including (1) creating process-level OpenMI components, (2) modeling component configurations with bi-directional links (feedback loops), and (3) the process for re-scaling data exchanges between spatially and temporally misaligned data components “on-the-fly” during model configuration runs. These are general challenges faced by many modeling systems that adopt a loose coupling paradigm, and we will present how such issues can be addressed within HydroModeler using OpenMI. Our future plans are to continue to grow the number of model components within the environment, focusing specifically on decomposing large watershed models into functional components that can then be linked together to simulate hydrologic systems and address specific science or management questions.

This document is currently not available here.

Share

COinS