Evaluation of re-analyses over China based on the temporal asymmetry of daily temperature variability

As an intrinsic feature of daily surface air temperature (SAT) variability found in station measurements, temporal asymmetry (TA) can be taken as an evaluation metric to access the quality of SAT re-analysis product. In this study, TA calculated from four SAT variables, i.e., daily mean SAT (T-mean), daily maximum SAT (T-max), daily minimum SAT (T-min) and diurnal temperature range (T-DTR = T-max - T-min), is applied to evaluate synoptic-scale performance of four reanalysis products (NCEP-2, JRA-55, ERA-I, and ERA-5) over China. The results show that four re-analyses overall overestimate the TA of daily T-max and T-min variability over China, but with a comparatively consistent estimated TA for T-mean. Moreover, the TA of T-mean variability for these four re-analyses shares high spatial consistency with those from the observation. However, four re-analyses own the similar region-dependent spatial patterns of overestimated TA for T-max and T-min variability, especially for T-max. Since high TA is an indicator for strong nonlinear feature, only T-mean reanalysis is the most suitable to explore synoptic-scale extreme events, such as heat waves and cold waves, which are highly related to the strong nonlinear processes.