Drought driving mechanism and risk situation prediction based on machine learning models in the Yellow River Basin, China

Abstract

Under global warming, the acceleration of the water cycle has increased the risk of drought in the Yellow River Basin. Revealing the drought driving mechanisms in the basin and understanding the risk situation of drought have become particularly important. This paper uses wavelet analysis and transfer entropy to analyze the drought driving mechanisms. In addition, an Improved Particle Swarm Optimization (IPSO) coupled with Long Short-Term Memory (LSTM) is used for drought risk prediction. The results are as follows: (1) Hydrological drought lags behind meteorological drought by 2–3 months, and they show two main periods on different time scales, which are 5–6 months and 8–14 months, respectively. (2) Rainfall, runoff, temperature, humidity, and vapor pressure are the main drought driving factors, with rainfall and humidity having the most significant impact. (3) The IPSO-LSTM model has improved the process of selecting model parameters based on empirical experiences in the LSTM model, improving the prediction accuracy by an average of 3.1%. This paper provides a scientific basis for water resource management and drought risk assessment in the basin, to better cope with future climate challenges.

Keywords:

• Driving factors
• Drought risk prediction
• Transfer entropy
• Machine learning
• The yellow river basin

Authors’ contributions

Ling Kang: Supervision, Validation, Resources, Writing-review & editing, Funding acquisition. Yunliang Wen: Investigation, Conceptualization, Methodology, Data curation, Software, Formal analysis, Writing-original draft, Visualization. Liwei Zhou: Conceptualization, Methodology, Data curation, Software, Formal analysis. Hao Chen: Data curation, Software, Formal analysis, Validation. Jinwang Ye: Data curation, Software, Visualization.

Disclosure statement

No conflict of interest exits in the submission of this manuscript, and manuscript is approved by all authors for publication.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author, Yunliang Wen.

Additional information

Funding

This research was funded by the National Key R&D Program of China (Grant No. 2022YFC3002704).