Abstract

The USDA Natural Resources Conservation Service (NRCS) is currently re-evaluating its need for erosion prediction technology from the USDA Agricultural Research Service (ARS). While the exact technology required has not yet been determined, the NRCS has proposed in the past to collaborate with the ARS to build a single process-based model to simulate both rainfall induced rill and interrill erosion as well as perform computations for wind erosion prediction. An additional priority within the NRCS is to incorporate any natural resource modeling technology (e.g., science module code and databases) for use by the agency within the Object Modeling System (OMS) being developed by the ARS Agricultural Systems Research Unit (ASRU) in Fort Collins, Colorado. The OMS is a modeling framework that uses an open-source software component-based approach to enable members of the scientific community to collaboratively address the many complex issues associated with the design, development, and application of natural resource models. The ARS has the lead in OMS design and development; however, strong collaborative support in terms of financial resources and personnel involvement has been provided by the NRCS and other federal agencies. OMS 3.0 represents a major milestone towards an easier to use, more transparent and scalable implementation of an environmental modeling framework. The main goal of OMS3 development is an easier integration of component-based model source code while being accommodating and flexible towards adopting existing legacy models. In OMS3, the internal complexity of the framework itself is reduced, while allowing models to implicitly scale from multi-core desktops to clusters to clouds, without burdening the model developer with complex technical details. This paper describes the development and application of a component-based, standalone, combined water and wind erosion model (WWEM) running under the OMS3 framework. Because OMS3 fully embraces multi-threading as the default execution model of simulation components, WWEM component execution is controlled by synchronization on objects passed from and to components, i.e., the components run in parallel without explicit programming by the component developer. The model (Figure 1) consists of both FORTRAN and Java-based components for hydrology (infiltration, water balance/redistribution and kinematic wave overland flow routing), plant cover (evapotranspiration), water erosion (sediment detachment, deposition, enrichment, and routing), and wind erosion (wind detachment). The components were adapted and modified primarily from the WEPP, WEPS, and RZWQM2 models and have been engineered for modularity and substitutability (e.g., the FORTRAN simulation components do not contain any common blocks). WWEM hydrology (surface runoff) and erosion (sediment loading) predictions were evaluated using data obtained from rainfall simulator studies on experimental plots in Colorado. Results show that the model was able to accurately capture runoff and erosion dynamics for discrete rainfall events at the plot-scale. Additional evaluation of the model is planned at the field scale. This study demonstrates the feasibility of using OMS3 as a modeling platform for enhancing open source component design through re-use of legacy code and improved model interoperability. Work is currently underway to add additional WWEM modeling components (e.g., the WEPS-based Unified Plant Growth Model being developed by ASRU scientists) and to improve existing model components (e.g., the WEPP-based hillslope water erosion modules). In conclusion, the component-oriented and modular approach of OMS3 and the erosion modules/models implemented in it will provide the basis for more efficient and internationally collaborative erosion model development in the future.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call