Temporal analysis of Soil and Water Assessment Tool (SWAT) performance based on remotely sensed precipitation products

Abstract

No study has systematically evaluated streamflow modelling between monthly and daily time scales. This study examines streamflow from seven watersheds across the USA where five different precipitation products were used as primary input into the Soil and Water Assessment Tool (SWAT) to generate simulated streamflow. Time scales examined include monthly, dekad (10 days), pentad (5 days), triad (3 days), and daily. The seven basins studied are the San Pedro (Arizona), Cimarron (north‐central Oklahoma), mid‐Nueces (south Texas), mid‐Rio Grande (south Texas and northern Mexico), Yocano (northern Mississippi), Alapaha (south Georgia), and mid‐St. Francis (eastern Arkansas). The precipitation products used to drive simulations include rain gauge, NWS Multisensor Precipitation Estimator, Tropical Rainfall Measurement Mission (TRMM), Multi‐Satellite Precipitation Analysis, TRMM 3B42‐V6, and Climate Prediction Center Morphing Method (CMORPH). Understanding how streamflow varies at sub‐monthly time scales is important because there are a host of hydrological applications such a flood forecast guidance and reservoir inflow forecasts that reside in a temporal domain between monthly and daily time scales. The major finding of this study is the quantification of a strong positive correlation between performance metrics and time step at which model performance deteriorates. Better performing simulations, with higher Nash–Sutcliffe values of 0.80 and above can support modeling at finer time scales to at least daily and perhaps beyond into the sub‐daily realm. These findings are significant in that they clearly document the ability of SWAT to support modeling at sub‐monthly time steps, which is beyond the capability for which SWAT was initially designed. Copyright © 2012 John Wiley & Sons, Ltd.

Comparison and Evaluation of Gridded Precipitation Datasets in a Kansas Agricultural Watershed Using SWAT

“…Automated calibration ensures consistency of the process for all models and minimizes the modeler bias in calibration exercises conducted for different precipitation sources. Similar procedures were followed in other recent studies (Bitew et al 2012;Tobin and Bennett 2013;Yang et al 2014;Radcliffe and Mukundan 2017;Ren et al 2018). Initial parameter ranges were selected based on professional judgment and literature.…”

“…Studies ( Moon et al 2004 ; Kalin and Hantush 2006 ; Sexton et al 2010 ; Tuppad et al 2010 ; Gali et al 2012 ; Tobin and Bennett 2013 ; Gao et al 2017 ; Radcliffe and Mukundan 2017 ) have evaluated the SWAT model parametrization to precipitation data sources, along with how data spatial and temporal resolutions impact simulated streamflow, model calibration, and associated uncertainties. Prior studies ( Moon et al 2004 ; Kalin and Hantush 2006 ; Sexton et al 2010 ; Tuppad et al 2010 ; Gali et al 2012 ; Tobin and Bennett 2013 ; Price et al 2014 ; Gao et al 2017 ) concluded that there are spatial-scale dependencies for accuracy of model simulation, however, the studies compared only one or two gridded sources to gauged data. Many of these studies calibrated SWAT with monitored precipitation (from National Climatic Data Center [NCDC]) and then ran simulations using the parameters of that SWAT model with gridded precipitation with no further calibration.…”

Applications of TRMM- and GPM-Era Multiple-Satellite Precipitation Products for Flood Simulations at Sub-Daily Scales in a Sparsely Gauged Watershed in Myanmar

“…The representative SPPs in these two eras include Precipitation Estimation from Remotely Sensed Information using Artificial Neural Network (PERSIANN) [3], Climate Precipitation Center morphing method (CMORPH) [4], Climate Hazards Group Infrared Precipitation with Stations [5], TRMM Multi-satellite Precipitation Analysis (TMPA) [1], Global Satellite Mapping of Precipitation (GSMaP) [6], and Integrated Multi-satellite Retrievals for GPM (IMERG) [2]. These SPPs generally provide quasi-global precipitation maps on high spatiotemporal resolutions (finer than 0.25° spatial resolution and shorter than the daily time interval); TRMM-era SPPs, in particular, have been widely adopted in hydrological applications in many parts of the world [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23]. techniques.…”

Temporal Patterns of Shrub Vegetation and Variation with Precipitation in Gurbantunggut Desert, Central Asia

“…Shrub vegetation links the soil and atmosphere, displays seasonal and annual changes obviously, and acts as a sensitive indicator of climate changes in shrubland, which is vastly distributed in the desert [8,9]. Precipitation is the primary supplier for soil water in the desert, where soil water is a strong limiting factor for vegetation growth within a year [10][11][12][13][14][15]; thus studying how the shrub vegetation responds to precipitation variation is one important component of desert vegetation responses to climate change.…”