Regional contrasting DTR’s predictability over China

Predictability is an important topic in weather and climate studies. Model-free and model-based quantification of diurnal temperature range (DTR) time series predictability is provided in this paper. Both intrinsic predictability quantified by the permutation entropy and realizable predictability defined by the model’s forecasting error are studied on DTR fluctuations. The intrinsic predictability tells us whether there are more predictive structures in DTR fluctuations and to what extent the maximal predictability can be reached. The realizable predictability measures the degree of prediction to which adopted model reaches. Results show that there is a well defined regime-dependent pattern between the intrinsic predictability and the realizable predictability in DTR fluctuations, and both the intrinsic predictability and the realizable predictability in DTR fluctuations over China are overall South–North (with the dividing line around the latitude of 37°N and almost along the Yellow River, and south and north to this line are called southern China and northern China in this study) asymmetric with higher predictability in southern China and weaker predictability in northern China. This South–North contrasting predictability behavior is closely related to the DTR’s South–North asymmetric multi-fractal behavior.