Bias Corrected FGOALS-g3 Model Products for Detecting Changes in Extreme Precipitation in the Tienshan Mountains, Central Asia

Extreme precipitation (EP) is a small-probability event characterized by a strong sudden occurrence, low frequency, and wide influence. Such events often have a serious impact on economic and social development, as well as on people's lives. According to the Emergency Events Database, floods and landslides accounted for 56.55% of all natural disaster events in Central Asia between 1992 and 2022. Over 1.28 million people were affected and the damage exceeded US $1.31 billion. The latest research suggests the variation of heavy precipitation is projected to become even greater in direct relation to increasing global warming. Therefore, to better cope with and mitigate the effects caused by EP, it is necessary to project its potential changes as accurately as possible.

To address this practical question, researchers from Xinjiang Institute of Ecology and Geography, Chinese Academy of Sciences, quantitatively assessed the correction performance and its application to the study of extreme precipitation in the Tienshan Mountains, Central Asia (TMCA), based on the bias correction of the FGOALS-g3 model developed by the State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, based on a multi-source precipitation dataset and several statistical indicators. The results show that (1) Raw FGOALS-g3 can reflect the spatial distribution of multi-year mean precipitation in the TMCA, but there is an obvious enlargement in the magnitude. The error is most obvious in the eastern West Tienshan Mountains and the western Middle Tienshan Mountains. After downscaling and bias correction, the errors are significantly reduced, and the point-by-grid error is between -7.64% and 10.95%. (2) In terms of the reliability of capturing spatial changes for EP against the observed data, the spatial distribution pattern of EP can be reasonably reproduced by the corrected FGOALS-g3 results. The overall relative error value is between -34.93% and 29.70%, except for severe extreme indices such as R99.9 and R99.95. (3) Limiting warming to 1.5°C can better control the intensity and frequency changes in EP. (4) The main driving factors in EP are detected to be 30hPa zonal wind, relative number of sunspots, south Asian summer monsoon, and mean surface temperature.

The results were published as " Application of bias corrected FGOALS-g3 model products for detecting changes in extreme precipitation in the Tienshan Mountains, Central Asia " in Atmospheric Research .

Article links: https://authors.elsevier.com/c/1fqK8cd3SFH6w 

Fig. 1. (a) Intra-annual distribution and (b) empirical cumulative distribution of the observed (obs), FGOALS-g3 simulated (raw), and corrected (cor) precipitation.

Fig. 2. Spatial changes in EP of observation data against corrected FGOALS-g3 data: (a, b) CDD, (c, d) CWD, (e, f) PRCP, (f, h) Rx1day, (i, j) Rx5day, and (k, l) SDII.

Contact: LIU Jie, Xinjiang Institute of Ecology and Geography 

E-mail: liujie@ms.xjb.ac.cn