Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #350319

Research Project: Integrating Remote Sensing, Measurements and Modeling for Multi-Scale Assessment of Water Availability, Use, and Quality in Agroecosystems

Location: Hydrology and Remote Sensing Laboratory

Title: Impacts of Global Circulation Model (GCM) bias and WXGEN on Modeling Hydrologic Variables

item Lee, S. - University Of Maryland
item Wallace, C. - Collaborator
item Sadeghi, Ali
item Mccarty, Gregory
item Zhong, H. - University Of Maryland
item Yeo, I-y - University Of Newcastle

Submitted to: Water
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/8/2018
Publication Date: 6/12/2018
Citation: Lee, S., Wallace, C., Sadeghi, A.M., Mccarty, G.W., Zhong, H., Yeo, I. 2018. Impacts of Global Circulation Model (GCM) bias and WXGEN on Modeling Hydrologic Variables. Water. 10(6):764.

Interpretive Summary: Use of a weather generator in watershed hydrologic models with Global Climate Model (GCM) data is a common practice. However, the impacts of the use of GCM data with a weather generator on watershed model predictions have rarely been examined. Raw GCM data often include an excessive number of wet days and therefore those data should be carefully used with a weather generator because the number of wet days per a month has a great impact on the estimation of solar radiation. These are critical data for projecting future hydrology and crop growth using hydrologic models. This study examined how climate variables from a weather generator using precipitation data with excessive wet days affected the Soil and Water Assessment Tool (or SWAT) model predictability on hydrologic variables and crop growth. We compared simulated outputs from two GCMs with assumed excessive and reasonable numbers of wet days to demonstrate the importance of input data preparation for a weather generator. Our simulation results showed the use of a weather generator with GCM data, including excessive wet days, led to unreasonable model predictions. Thus, GCM data should be thoroughly processed and prepared to be used for a weather generator included in hydrologic models.

Technical Abstract: A weather generator (WXGEN) is commonly used to simulate daily climate data for the Soil and Water Assessment Tool (SWAT) model, when input climate data are not fully available. Of all input data for WXGEN, precipitation is the critical one due to the sensitivity of WXGEN algorithm for the number of wet days per a given month. Since Global Climate Model (GCM) data tend to have excessive wet days, the use of GCM precipitation data for WXGEN may cause errors in the estimation of climate variables generated, thus affect the accuracy of SWAT output predictions. To quantify such impacts of GCM data, we prepared two climate databases for SWAT using WXGEN with both the original GCM data with the excessive number of wet days (OGCM) and the processed GCM data with the reasonable number of wet days (DGCM). We then compared the SWAT simulations (e.g., streamflow, nitrate loads, and crop biomass) using OGCM and DGCM databases. Results showed that because of the excessive wet days in OGCM, solar radiation generated by WXGEN was underestimated, subsequently leading to 143 mm lower ET and 0.6 – 0.8 cubic meters per second greater streamflow, compared to the simulations from DGCM. Although the OGCM database led to a greater streamflow than DGCM database, nitrate loads under OGCM and DGCM were similar due to a substantial export of fertilizer-driven nitrate under DGCM. Simulated crop biomass under OGCM was smaller than DGCM due to less solar radiation. Reduced ET resulted in less crop water stress under OGCM relative to DGCM even though OGCM indicated 56 mm lower precipitation than DGCM during summer growing seasons. Our findings demonstrate that the excessive number of wet days in GCM data lead WXGEN to generate inaccurate climate data, resulting in unreasonable SWAT model predictions.