Assessment and Prediction of Extreme Temperature Indices in the North China Plain by CMIP6 Climate Model

• Published in: Applied sciences 2022-07, Vol.12 (14), p.7201
• Main Authors: Wang, Hui, Wang, Lu, Yan, Guoying, Bai, Huizi, Zhao, Yanxi, Ju, Minmin, Xu, Xiaoting, Yan, Jing, Xiao, Dengpan, Chen, Lirong
• Format: Article
• Language: English
• Subjects: Agricultural production; Artificial intelligence; Climate change; Climate models; CMIP6; extreme temperature index; Forecasting; global climate models; Global warming; Ground stations; Heat; High temperature; Low temperature; Machine learning; north China plain; Precipitation; Radiation; Regression analysis; Root-mean-square errors; Simulation; Spatial variations; Support vector machines; Temperature; the SVM
• ISSN: 2076-3417; 2076-3417
• DOI: 10.3390/app12147201

Extreme temperature events are becoming more frequent due to global warming, and have critical effects on natural ecosystems, social and economic spheres, human production and life. Predicting changes in temperature extremes and trends under future climate scenarios helps to assess the impact of climate change accurately. Based on climate observations from 54 meteorological stations in the North China Plain and the projection data from seven general circulation models (GCMs) from the Coupled Model Intercomparison Project phase 6 (CMIP6), this paper researches nine representative extreme temperature indices under four typical climate scenarios. The aim is to reveal the temporal and spatial variations in extreme temperature indices in the North China Plain during the past (1971–2010) and the future (2061–2100). The results show that: using a support vector machine (SVM) to perform regression analysis on the multi-GCMs prediction results, the root mean square error (RMSE) and relative standard deviation (RSD) of the multi-model ensemble simulations obtained by the SVM method are lower than those of the arithmetic mean method and can better match the trend of the historical extreme temperature index; the extreme high temperature index is predicted to show a significant upward trend in the future, while the extreme low temperature index will decrease significantly; and there are significant spatial differences in the extreme temperature index in both historical and future periods, with the extreme temperature index under the high radiation forcing scenario (SSP585) showing the most considerable variation and the most significant spatial differences.