Improving soil moisture accounting and streamflow prediction in SWAT by incorporating a modified time-dependent Curve Number method Mohammad Adnan Rajib and Venkatesh Merwade* Lyles School of Civil Engineering, Purdue University, West Lafayette, IN, USA Abstract: The objective of this study is to incorporate a time-dependent Soil Moisture Accounting (SMA) based Curve Number method (SMA_CN) in Soil and Water Assessment Tool (SWAT) and compare its performance with the existing CN method in SWAT by simulating the hydrology of two agricultural watersheds in Indiana, USA. Results show that fusion of the SMA_CN method causes decrease in runoff volume and increase in profile soil moisture content, associated with larger groundwater contribution to the streamflow. In addition, the higher amount of moisture in the soil profile slightly elevates the actual evapotranspiration. The SMA-based SWAT configuration consistently produces improved goodness-of-fit scores and less uncertain outputs with respect to streamflow during both calibration and validation. The SMA_CN method exhibits a better match with the observed data for all flow regimes, thereby addressing issues related to peak and low flow predictions by SWAT in many past studies. Comparison of the calibrated model outputs with field-scale soil moisture observations reveals that the SMA overhauling enables SWAT to represent soil moisture condition more accurately, with better response to the incident rainfall dynamics. While the results from the modification of the CN method in SWAT are promising, more studies including watersheds with various physical and climatic settings are needed to validate the proposed approach. Copyright © 2015 John Wiley & Sons, Ltd. KEY WORDS SWAT; SCS-CN; soil moisture; Cedar Creek; White River; Indiana Received 16 March 2015; Accepted 3 August 2015 INTRODUCTION Information of soil moisture and its spatio-temporal dynamics is needed for irrigation scheduling and yield forecasting, hydroclimatic prediction of flood and droughts, efficient water quality management and natural conservation practices in a watershed-scale (Walker et al., 2001; Western et al., 2002; Starks et al., 2006; Zucco et al., 2014). Accordingly, precise prediction of soil moisture has been a subject of long-standing research. Soil moisture information can be obtained from in situ sensor-based point measurements, satellite remote sensing, and physically based distributed hydrologic modelling of watersheds. Typically, at large scales, spatio-temporal variability of soil moisture is not well represented through point measurements from in situ sensor networks. In comparison, characterizations of surface soil moisture fields captured by the remote sensing techniques have been found sufficient enough to supplement the in situ point measurements (Njoku and Entekhabi, 1996). Yet, remote sensors have a shallow sensing depth, providing soil moisture estimates only for the top 5 cm of soil (Adams et al., 2013; Vereecken et al., 2014). For simulating major hydrologic processes, soil moisture condition across the whole root zone is more useful. Therefore, physically based hydrologic models are commonly used to simulate soil moisture content across the vertical profile of the root zone over a continuous period of time (e.g. Starks et al., 2003; Chen et al., 2011; Han et al., 2012a; Korres et al., 2013; Park et al., 2013). However, soil moisture-related conceptualization and parameterization within hydrologic models remain an outstanding concern due to non-physical representation of Soil Moisture Accounting (SMA) processes in the models. Simulation of soil moisture content through a physi- cally based continuous simulation distributed hydrologic model is often indirect and largely dependent on runoff generation simulation process (Neitsch et al., 2011; Han et al., 2012b). Numerous models of varying complexity use the Soil Conservation Service Curve Number (SCS- CN) method for computing rainfall excess [e.g. Erosion Productivity Impact Calculator (EPIC, USDA (1990)); Hydrologic Evaluation of Landfill Performance (HELP, Schroeder et al. (1994)); Long Term Hydrologic Impact Analysis (L-THIA, Harbor (1994)); Pesticide Root Zone *Correspondence to: Venkatesh Merwade, 550 Stadium Mall Dr., West Lafayette, IN 47907, USA. E-mail: vmerwade@purdue.edu HYDROLOGICAL PROCESSES Hydrol. Process. 30, 603–624 (2016) Published online 13 September 2015 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/hyp.10639 Copyright © 2015 John Wiley & Sons, Ltd.