Quantification of Streamflow Uncertainty in SWAT Models Role of Hydroclimatic Data Sources
Abstract
Understanding the variations in the water balance is becoming increasingly important in watershed management. One of the major reasons of these variations is due to climate change resulting in the increasing rates of extreme precipitation events and surface runoff. Precipitation is one of the primary hydroclimatic variable for the water source component of the hydrological budget. The prediction of streamflow changes will significantly depend on precipitation and its uncertainty. One of the uncertainties with use of precipitation data in hydrologic models is the sources of data such as station observed data (sparse), satellite data (good spatial coverage), and reanalysis gridded data (long-term synthetic data). The efficacy of the hydrologic models greatly depends on the density and accuracy of the available spatial coverage of precipitation data. The detailed analysis will help in quantification of input uncertainty from precipitation sources and to understand the variations in the streamflows. In this study, the sources of precipitation data considered are Global Precipitation Climate Centre (GPCC) gauge datasets, Tropical Rainfall Measurement Mission (TRMM) satellite datasets, Next Generation Weather data (NEXRAD) ground-based radar datasets, and reanalysis data are used to quantify the model output based on interactions with other watershed characteristics such as topography and land use/land cover datasets. The observed rain gauge data obtained from Climate Forecast System Reanalysis (CFSR) using Global Weather Database as the standard dataset for comparing the results. The Soil and Water Assessment Tool (SWAT) model is used to predict the streamflow in Peachtree River near Atlanta city, USA. The efficacy of the models performance is carried out using the statistical indicators of NashSutcliffe efficiency (NSE), Root Mean Square Error to Standard Deviation of Observed Stream-Flow (RSR), and PercentBias Error (PBIAS). Preliminary results indicate the there is a significant effect on the SWAT model performance due to precipitation data sources. The results from the study will be helpful for end-user/stake-holders in understanding effect of precipitation source uncertainty, which can help them to effective planning and management of water resources.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFM.H13H..02C