资源环境科技发展态势分析平台

As one of the most important indicators of nonlinear time series, temporal asymmetric (TA) or temporal irreversible behaviors of daily fluctuations from four temperature variables, including mean temperature (T-mean), maximum temperature (T-max), minimum temperature (T-min) and diurnal temperature range (DTR, DTR=T-max-T-min), have been quantified through both observations and reanalysis over China by two TA measures. One is L-2 from directed horizontal visibility graph and the other is L-1 from consecutive increasing and decreasing steps. The results show that there are differential TA features among daily temperature fluctuations. Firstly, there are marked differences in TA strengths among different temperature variables. It is found that dominated uniformly significant TA (larger than the threshold from corresponding surrogates) emerges in almost all observed T-mean fluctuations. However, this kind of uniformly significant TA can't be found in T-max, T-min and DTR for both observed and reanalysis data sets. Since both TA measures L-1 and L-2 quantify the temporal structures in the given series, this distinguishable TA strengths found in different temperature variables indicates that there are distinct temporal structures in the different temperature variables' variations. Secondly, the TA in each temperature variable is region dependent. The TA strength for each temperature variable is spatially non-uniform with some strong and weak TA regional patterns and these strong and weak TA regional patterns may depend on local weather or climate conditions. Moreover, comparison studies of the same temperature variable reveal that time irreversible features are distinguishable between observations and reanalysis, and this differential feature can be taken as an index to evaluate the quality of reanalysis.