Quality Control for Surface Hourly Temperature Observations via An Improved SRT Method

The article aims to discuss the quality control methods for the surface hourly temperature observations. A new quality control method (RH-SRT) based on the spatial regression test (SRT) was adopted to identify potential outliers in the surface hourly temperature observations. The proposed method is an extension of SRT, which takes the relationship between relative humidity and temperature as the explaining variable to obtain the estimated value of surface temperature. In order to evaluate the proposed method, RH-SRT was applied to an annual temperature dataset with seeded errors for different regions in Jiangsu Province during 2007. Compared with the traditional SRT method, the results demonstrate that RH-SRT outperforms SRT for all cases. Moreover, comparison results demonstrate that the RH-SRT method is an effective version of SRT, which could be used to improve the quality of surface hourly temperature observations.