Bias Analysis and Correction of Ground Surface Temperature Observations across China

Based on the ground surface temperature (GST) and snow surface temperature (SST) measurements during the period of adjustment from manual to automatic observation systems in China, the influence of observation methods on GST and its relationship with snow cover is analyzed. GST is corrected by SST, and the correction effect is evaluated. The results show that, during the parallel observation period, the winter GSTs from automatic observations are generally higher than those from manual observations, with the automatically observed national daily GST 1.18°C higher. The adjustment has a greater impact on GSTs at 0200, 0800, and 2000 Beijing Time (BT) than at 1400 BT, and it has the greatest impact in Northeast and Northwest China, with deviations of 5.24 and 2.09°C, respectively. The GST deviation is closely related to the snow depth and annual snow totals. The average daily GST deviation increases at the rate of 0.66°C per 1-cm increase of snow depth when it is ＜ 15 cm, while it tends to be stable at around 10°C for snow depth over 15 cm. The GST deviation at a station is affected by its winter snow totals in Northeast and Northwest China, where the largest deviations are found where snow totals are all above 1000 cm. After the correction with SST, the mean deviation between the automatic and manual observations as well as the false trend can be effectively reduced. Following the correction, the mean deviation of daily GST decreases by 5.8°C, and its trend decreases from 1.87 to 0.65°C decade⁻¹.