Comparing the capability of distributed and lumped hydrologic models for analyzing the effects of land use change Murari Paudel, E. James Nelson, Charles W. Downer and Rollin Hotchkiss ABSTRACT Empirically based lumped hydrologic models have an extensive track record of use for various engineering applications. Physically based, multi-dimensional distributed models have also been in development and use for many years. Despite the availability of high resolution data, better computational resources and robust, numerical methods implemented in such models, their usage is still limited, especially in the realm of surface water runoff simulation. Lumped models are often extended to solve complex hydrologic problems that may be beyond their capabilities. Here we attempt to differentiate the ability of lumped and distributed models to analyze a common watershed development issue such as land use change. For this, we employ two common US Army Corps of Engineers (USACE) models, well established in the literature and application, using the Hydrologic Engineering Center – Hydrologic Modeling System (HEC-HMS) model in a fully lumped mode and the fully distributed model Gridded Surface Subsurface Hydrologic Analysis (GSSHA). A synthetic watershed is used to establish that a distributed model like GSSHA more intuitively simulates land use change scenarios by distinguishing the spatial location of the change and its effects on the watershed response. An actual watershed at Tifton, Georgia is used to validate the observations made from the synthetic watershed. Key words 9 GSSHA, HEC-HMS, Landuse change modeling, Lumped and distributed models INTRODUCTION Hydrologic modeling is commonly used to simulate runoff and subsequent streamflow from watersheds. Streamflow estimations may be used for a variety of purposes, such as design of hydraulic structures, prediction of flood stages and ecological restoration design. While useful for understanding the overall system, a simulation model is a simplification of the actual watershed. The nature of the model and the simplification involved in model formulation partially deter- mines model applicability to both common and advanced hydrologic modeling tasks (Barnes 1995). In terms of spatial discretization and resolution, hydro- logic models can be organized on an ascending scale of sophistication beginning with empirically based lumped models and ending with physically based distributed models (Jones 1997). Originally, hydrologic simulation modeling systems were all of the lumped parameter types. Lumped parameter models typically have conceptual or empirical formulations for pre- cipitation, infiltration, surface runoff, etc., and represent the watershed as well as the hydrologic components as a single homogeneous unit. More recently, models that attempt to simulate both the spatial heterogeneity and the physical processes occurring within a watershed have been developed (Bobba et al. 2000). Murari Paudel (corresponding author) Brigham Young University, 300 CB, Brigham Young University, Provo UT 84602, USA E-mail: mpaudel@et.byu.edu E. James Nelson Civil and Environmental Engineering, 242 K CB, Brigham Young University, Provo UT 84602, USA Charles W. Downer Hydrologic Systems Branch, Coastal and Hydraulics Laboratory, US Army Engineer Research and Development Center, USACE c/o SFWMD, Bldg B2, IMC, 3301 Gun Club Road, West Palm Beach, FL 33406, USA Rollin Hotchkiss Civil and Environmental Engineering, 242 J CB, Brigham Young University, Provo UT 84602, USA doi: 10.2166/hydro.2010.100 & IWA Publishing 2011 461 Journal of Hydroinformatics 9 9 2011 13.3 Downloaded from http://iwaponline.com/jh/article-pdf/13/3/461/386567/461.pdf by guest