Measurements and Regional Climate Modeling of Soil Temperature With and Without Bias Correction Method Under Arid Environment: Can Soil Temperature Outperform Air Temperature as a Climate Change Indicator?

Soil temperature plays a critical role in many soil functions, particularly in arid ecosystems, but is scarcely reported as potential indicator of climate change. Although studies have been conducted worldwide to investigate (both measurements and modeling) temperature changes in the soil profile in response to ambient temperature, no information on soil temperature is available in the state of Kuwait. Hydrological and many bio-geochemical processes are more sensitive to soil temperature than air temperature. In this study, we used observed soil temperature data (2007–2016) from three sites with three soil depth increments (5, 50, 100 cm) and compared to regional climate and regression models' output. The most salient finding of this study is the tight association between observed and simulated soil temperature at shallower soil depths from both the regional climate model (RegCM4) and linear regression model. The RegCM4 model poorly predicted (underestimated) soil temperature at deeper soil layers relative to shallower soil layers. Application of the linear scaling (LS) method has significantly improved the RegCM4 model performance with respect to measured soil temperature, which allows an accurate evaluation of the impact of climate change on the soil temperature under different soil management systems. These findings indicate that RegCM4 can be applied as a reliable predictor of soil temperature under arid ecosystems for which there is a gap in site data availability. [ABSTRACT FROM AUTHOR]