Prediction of Drought/flood Intensities Based on a 500-Year Time Series in Three Different Climate Provinces of China

Peng Yu,Wang Zhaohua,Cui Yixin,Su Yali,Tang Yiyun,Luo Shiqian,Liu Lianlin,Zhou Yanqiu,Dong Jihui,Wang Wentao,Li Xiang,Shang Jiaxin,Huang Yihang,Lu Yitong,He Xiaoao,Bai Ting
DOI: https://doi.org/10.1007/s10113-022-01936-w
IF: 4.704
2022-01-01
Regional Environmental Change
Abstract:In this study, the intensities of droughts and floods in three provinces of China are projected using three approaches: wavelet analysis, auto-regressive integrated moving average (ARIMA) models, and random tree (RT) models based on a 500-year time series. The motivation for this work was the observation that drought/flood intensities are generally increasing, the prediction of drought/flood intensities for the next 50 years, along with their future spatiotemporal patterns. Based on the three approaches, with few exceptions, the results reveal a likely increasing trend for both droughts and floods, with higher intensity under the RCP8.5 scenario for droughts and the RCP4.5 scenario for floods by the year 2050. Notwithstanding this worsening trend, it was found that drought/flood intensities for most regions are projected to neither entirely increase nor entirely decrease in the near future. Drought and flood intensities may alternatively occur on a multi-year or a decadal timescale. Inner Mongolia, located in an arid temperate climate zone, is projected to experience a higher average disaster intensity, followed by the warm monsoon region of Henan, while the humid subtropical region of Guizhou is projected to be the least affected, with high spatial heterogeneity within each region. Climate variation, particularly temperature range and deviation across either the coldest or wettest months, was closely associated with drought/flood intensities. We conclude that the major challenge for drought/flood prediction is not the general change trends or their responses to climate change, but the spatiotemporal heterogeneity at multiple scales.
What problem does this paper attempt to address?