Comparing two weather generator-based downscaling tools for simulating storm intensification and its impacts on soil erosion under climate change

Authors: Zhang, Xunchang; Busteed, Phillip; CHEN, JIE - Wuhan University; YUAN, LIFENG - Us Environmental Protection Agency (EPA)

Submitted to: International Journal of Climatology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/5/2022
Publication Date: 12/26/2022
Citation: Zhang, X.J., Busteed, P.R., Chen, J., Yuan, L. 2022. Comparing two weather generator-based downscaling tools for simulating storm intensification and its impacts on soil erosion under climate change. International Journal of Climatology. 43:2220-2237. https://doi.org/10.1002/joc.7971.
DOI: https://doi.org/10.1002/joc.7971

Interpretive Summary: Frequency and magnitude of heavy storms are projected to increase in the future under climate change. Such increases are often referred to as storm intensification for simplicity. It is of great importance to simulate properly future storm intensification because most damage to agricultural production is caused by extreme precipitation events. However, it is rather difficult to simulate extreme precipitation using climate models or to statistically represent it for climatic impact assessment. The objectives of this study are to compare two computer software tools (SYNTOR and GPCC) for simulating daily precipitation extremes by manipulating daily precipitation distributions, both with and without storm intensification options, and to further evaluate the responses of simulated surface runoff and soil loss to generated precipitation extremes under various cropping and tillage systems. The study was conducted for a site near Weatherford, Oklahoma. Results show that all data sources, including Global Climate Models (GCM), Regional Climate Models (RCM), and Statistically-downscaled GCM output are able to provide signals of future storm intensification for use in SYNTOR and GPCC, if proper models or particular ensemble members (simulations with different initial conditions) of each model are selected. Thus, it is recommended that all data sources including as many ensemble members as possible should be used in model screening or selection. Compared with SYNTOR, GPCC tool tends to generate a stronger storm intensification. As a result, simulated soil loss and surface runoff rates with GPCC were significantly greater than those with SYNTOR due to GPCC generating more frequent and heavier storms. Given the sizable differences in extreme precipitation generation, both tools should be used to generate an upper and lower bound of the potential impacts of storm intensification under climate change on soil erosion and surface hydrology. This work provides useful information on simulating storm intensification using the SYNTOR and GPCC tools for assessing the impact of climate change on agricultural production and natural resource conservation.

Technical Abstract: Storm intensification is projected to increase under climate change; however, it is quite challenging to statistically downscale storm intensification for climatic impact assessment. The objectives of this study are to compare two weather generator-based tools in simulating daily precipitation extremes, both with and without storm intensification options, and to further evaluate the responses of simulated surface runoff and soil loss to generated precipitation extremes under various cropping and tillage systems for a site in central Oklahoma, U.S.A. Results show that all data sources, including raw Global Climate Models (GCM), raw Regional Climate Models (RCM), and statistically downscaled GCM datasets are capable of providing signals of future storm intensification, depending on particular models and their ensemble members in each source. Thus, it is recommended that all data sources including as many members as possible should be used in model screening. Compared with the SYNthetic weather generaTOR (SYNTOR), the Generator for Point Climate Change (GPCC) tool tends to generate a stronger storm intensification, partially because GPCC takes into consideration the increased variance of monthly precipitation as projected by GCMs in adjusting daily precipitation variance for future climate generation. However, there are no discernable differences between storm intensification options with each tool because, weaker signals of storm intensification are projected for the study site coupled with a projected decline in future annual precipitation. Simulated soil loss and surface runoff rates with GPCC were significantly greater than those with SYNTOR due to GPCC generating more frequent and heavier storms. Given the sizable differences in extreme precipitation generation, both tools should be used to generate an upper and lower bound of the potential impacts of storm intensification under climate change on soil erosion and surface hydrology.