Transactions of the ASABE Vol. 55(4): 1463-1477 2012 American Society of Agricultural and Biological Engineers ISSN 2151-0032 1463 WEPP: MODEL USE, CALIBRATION, AND VALIDATION D. C. Flanagan, J. R. Frankenberger, J. C. Ascough II ABSTRACT. The Water Erosion Prediction Project (WEPP) model is a process-based, continuous simulation, distributed parameter, hydrologic and soil erosion prediction system. It has been developed over the past 25 years to allow for easy application to a large number of land management scenarios. Most general or field agency users of WEPP rely upon ex- isting or special databases and/or interfaces that have been developed, tested, and verified by others. This article de- scribes WEPP model calibration and validation procedures, under ideal situations (where all necessary input data and runoff/sediment observations are available) as well as under more typical and less ideal conditions. Two case study appli- cations of the model from the literature are highlighted and discussed in detail as examples of single storm hillslope pro- file and continuous simulation watershed applications. Current and future development efforts on WEPP are also de- scribed. Keywords. Erosion by water, Hydrology, Modeling, Soil loss, Validation, Verification, WEPP. he Water Erosion Prediction Project (WEPP) was developed by a large team of federal and universi- ty scientists, beginning in 1985. The project was initiated by Dr. George R. Foster, at that time a hydraulic engineer with the USDA-ARS National Soil Ero- sion Research Laboratory (NSERL) in West Lafayette, In- diana. WEPP process-based technology was intended to ul- timately become a replacement for the mature, empirically based Universal Soil Loss Equation (USLE; Wischmeier and Smith, 1978), which was the standard tool used by the USDA Soil Conservation Service (SCS) for field conserva- tion planning on private lands, as well as for conducting na- tional soil resource inventories. Foster initiated the project, formed a core team of scientists, and worked with user agencies (SCS, Forest Service (FS), and Bureau of Land Management (BLM)) to develop a user requirements doc- ument to guide development of the model (Foster and Lane, 1987). The core team laid out the basic hydrologic and ero- sion process logic and designed extensive field experiments that were conducted in 1987-1988 to provide data for mod- el parameterization (Laflen et al., 1991; Elliot et al., 1989). WEPP model coding was conducted at ARS locations in Tucson, Arizona, and West Lafayette, Indiana, mainly from 1985 to 1995, and the documented and validated hillslope and watershed versions of WEPP were released at a special Soil and Water Conservation Society symposium in Des Moines, Iowa, in July 1995 (Flanagan and Nearing, 1995; Flanagan and Livingston, 1995). Since 1995, WEPP model, interface, and database development have continued at the USDA-ARS NSERL, although at a reduced level compared to the initial project efforts. Flanagan et al. (2001, 2007) pre- sent additional details on WEPP model development and sci- ence, field experimentation, and core scientific team compo- sition. Other journal articles related to WEPP development, validation, and application include Nearing et al. (1989a), Zhang et al. (1996), Ascough et al. (1997), Liu et al. (1997), Tiwari et al. (2000), Laflen et al. (2004), Clark et al. (2006), Cruse et al. (2006), Pieri et al. (2007), Moore et al. (2007), and Abaci and Papanicolaou (2009). The current WEPP model software, technical documentation, and user docu- mentation are publicly available from the NSERL website (www.ars.usda.gov/Research/docs.htm?docid=10621). Mod- el support or source code requests can be sent directly to the lead author of this article or to the e-mail address wepp@ecn.purdue.edu. WEPP DESCRIPTION WEPP is a distributed parameter model that can be run in either a hillslope profile or a small watershed configura- tion. The spatial scale for hillslope profile simulations is typically from 1 to ~100 m in length, although in some sit- uations longer hillslopes of several hundred meters may be adequately simulated. For small watersheds in which the application interest is catchment sediment yield, the rec- ommended maximum area of model application is about 260 ha, although again, in some situations, satisfactory per- formance may be achieved on somewhat larger areas. Al- ternately, for watersheds in which the main objective is to Submitted for review in October 2011 as manuscript number SW 9458; approved for publication by the Soil & Water Division of ASABE in July 2012. The USDA is an equal opportunity provider and employer. The authors are Dennis C. Flanagan, ASABE Fellow, Research Agri- cultural Engineer, and James R. Frankenberger, Information Technology Specialist, USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, Indiana; and James C. Ascough II, Research Hydraulic Engi- neer, USDA-ARS Agricultural Systems Research Unit, Fort Collins, Colo- rado. Corresponding author: Dennis C. Flanagan, USDA-ARS National Soil Erosion Research Laboratory, 275 S. Russell Street, West Lafayette, IN 47907; phone: 765-494-7748; e-mail: Dennis.Flanagan@ars.usda.gov. T