Assessment and improvement of RegCM 4.6 coupled with CLM4.5 in simulation of land surface temperature in mainland China

Abstract Land surface temperature(LST) is an important indicator to study climate change and test the performance of regional model simulation. RegCM4.6 representative version RegCM, which coupled with advanced third-generation land NCAR CLM4.5. Currently, has become tool in China. However, its ability simulate temperature mainland China reasons for deviation have not been systematically studied, targeted improvement work lacking. The present first employ LST data collected from 809 Chinese meteorological stations last 30 years comprehensively assess CLM4.5 LST. Sensitivity tests soil thermal conductivity (STC) were carried out improve model. Although could accurately temporal spatial variation LST, a cold bias 2~8 °C existed all China, was larger seasons more precipitation greater moisture than other seasons. Deviation increased southeast northwest. caused by incoming long-wave radiation, sensible heat, latent heat simulated. There significant linear relationship between observed simulated LSTs, correlation coefficients ranged 0.75 0.9 ( P < 0.001). at rate 0.58 °C/decade, but lower rate. Assessment three different STC schemes showed that Lu-Ren scheme most suitable simulation Developing new considers role water vapor can effectively when used