CREAMS/GLEAMS: model use, calibration, and validation. (Special Issue: Model use, calibration, and validation.)

摘要

The Chemicals, Runoff, and Erosion from Agricultural Management Systems (CREAMS) model was developed by a multidisciplinary team of research scientists from the USDA Agricultural Research Service. The primary purpose of the model was to aid the USDA Natural Resources Conservation Service (NRCS) specialists in assessing nonpoint-source pollution from agricultural areas and to compare impacts of alternative management practices. Existing readily modifiable and new components were assembled into a field-scale model. Modelers also were charged with publication of model documentation, validation, and a user manual. Model validation with available data and sensitivity analyses of parameters in the hydrology, erosion, plant nutrient, and pesticide components were essential to demonstrate the model capabilities and effectiveness. Technology transfer to NRCS personnel was necessary to gain acceptance and proper use as well as to identify weaknesses and areas for needed improvement. Following publication and NRCS acceptance, improvements to CREAMS were made to better represent soil layering, crop rotations, irrigation, soil water routing, and chemical movement, which resulted in the Groundwater Loading Effects of Agricultural Management Systems (GLEAMS) model. Climatic records were increased from a maximum of 20 years to 50 years. The modifications were validated with available data and included in subsequent publications. The user manual gives more comprehensive description of the model parameters and their relative sensitivity. This article describes the historical development of the CREAMS and GLEAMS models with an emphasis on providing model users with an understanding of the degree of model verification and validation that was undertaken during development of these multidisciplinary models. Model calibration is discussed, including sensitive parameters and recommended procedures. Finally, two specific case studies are presented along with brief synopses of numerous case studies that have been used to validate model components. Future users are encouraged to use the model source code to further expand its utility for analyzing the nonpoint-source pollution impacts of agricultural management practices.