Reduced Tropical Cyclone Genesis in the Future as Predicted by a Machine Learning Model

Due to a lack of observations and limited understanding of the complex mechanisms of tropical cyclone (TC) genesis, the possible TC activity response to future climate change remains controversial. In this work, a machine learning model, called the maximum entropy (MaxEnt) model, is established using various environmental variables. The model performs slightly better than the genesis potential index (GPI) for historical TC activities based on the spatial correlation coefficient. Using Coupled Model Intercomparison Project Phase 6 (CMIP6) model projections, the MaxEnt model predicts a statistically significant decreasing trend of TC genesis probability under all shared socioeconomic pathway (SSP) scenarios. In addition, our analysis reveals that TC genesis might have a complex nonlinear relationship with potential intensity, which is different from the positive relationship reported in previous studies and might be the key factor leading to the model predicting reduced TC genesis in the future.

This work is conducted by Dr. Qian QiFeng and the corresponding author is Professor XiaoJing Jia from Earth Sciences department, ZheJiang University. The coauthor is Professor Li YanLuan from department of Earth System Science, Tsinghua University.

Fig. 1.The predicted TC activity under different scenarios. The spatial distribution of the (a)TC numbers obtained from the observational IBTrACS dataset and (b) ensemble mean of the predicted TC genesis probability by the MaxEnt model using historical datasets of five CMIP6 models during 2000–2009. The spatial correlation coefficient between Fig. 1a and Fig. 1b is 0.61. (c) The area-weighted average (45°S-45°N) of the MaxEnt-predicted TC genesis probability using five CMIP6 model outputs under all SSP scenarios from 1860 to 2100.

This research is funded by the National Natural Science Foundation of China (grant 42075050).

Details of the work:

Qian, Q.F., Jia, X.J.* and Lin, Y.L., 2022. Reduced Tropical Cyclone Genesis in the Future as Predicted by a Machine Learning Model. Earth Future, 10(2), Doi: 10.1029/2021EF002455. 

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2021EF002455