Exploring the interconnections between total cloud water content and water vapor mixing ratio with other cloud microphysical variables in northward-moving typhoon precipitation via information entropy: A hybrid causal analysis approach using wavelet coherence and Liang–Kleeman information flow

Abstract

Causal analysis of cloud microphysical variables constitutes an effective means for characterizing the microphysical attributes and causal mechanisms of precipitation clouds. Causal analysis methods primarily rely on Granger causality tests based on lagged variables and linear regression. However, most cloud physical precipitation processes are nonlinear. Herein, a novel hybrid approach involving information entropy, wavelet decomposition, and Liang–Kleeman information flow is introduced to enhance the dependability and effectiveness of causal analysis for the self-organizing process of precipitation clouds in this paper. Based on the Weather Research and Forecasting (WRF) model, a case study is conducted of the northward-moving process of Typhoon Maysak in 2020. Gridded data with 30-min intervals and a 6 km × 6 km resolution is extracted. Through empirical analysis, using the total cloud water content (TWC) and water vapor mixing ratio (QV) as the principal variable and atmospheric vertical velocity (OMG), precipitable water (PW) and outgoing longwave radiation (OLR) as covariates, the hybrid causal analysis methodology is assessed. TWC and QV are direct and potential influencing factors of precipitation, respectively. Results indicate that the probability distributions of TWC and QV are significantly different at different stages. In the typhoon stage, typical self-organizing characteristics of high mean and low information entropy values are presented; in the tropical storm stage, information entropies increase, TWC increases, and QV decreases, with self-organizing characteristics weakening; in the tropical depression stage, both the mean and information entropies of TWC and QV show a significant decrease. Wavelet coherence analysis indicates that IEOLR and IEPW can better explain IETWC, and IEPW and IEOMG can better explain IEQV. There is a significant causal relationship between IETWC and IEPW at different time scales. At larger periodic scales, IEQV has significant causal relationships with IEOMG, IEPW and IEOLR. Overall, this approach provides insights into the complex causal relationships of cloud microphysical variables in a precipitation cloud system, broadening our understanding of these complex phenomena.