Improved drought forecasting in Kazakhstan using machine and deep learning: a non-contiguous drought analysis approach
Kazakhstan is recently experiencing an increase in drought trends. However, low-capacity probabilistic drought forecasts and poor dissemination have led to a drought crisis in 2021 that resulted in the loss of thousands of livestock. To improve drought forecasting accuracy, this study applies Machin...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IWA Publishing
2024-02-01
|
| Series: | Hydrology Research |
| Subjects: | |
| Online Access: | http://hr.iwaponline.com/content/55/2/237 |
| Summary: | Kazakhstan is recently experiencing an increase in drought trends. However, low-capacity probabilistic drought forecasts and poor dissemination have led to a drought crisis in 2021 that resulted in the loss of thousands of livestock. To improve drought forecasting accuracy, this study applies Machine Learning and Deep Learning (ML and DL) algorithms to capture the sequences of drought events using a non-contiguous drought analysis (NCDA). Precipitation, 2-m temperature, runoff, solar radiation, relative humidity, and evaporation were collected from the ERA5 database as input variables. Combinations of inputs were used to build ML models, including seven classifiers (Logistic, K-NN, Kernel SVM, Decision Tree, Random Forest, XGBoost, and GRU). The output events were defined by standardized precipitation index (SPI) and SPEI indicators as binary classes. Weekly time series from 1991 to 2021 for each cell were used to forecast a lead time from 1 week to 6 months. GRU provided 97–99% accuracy in more volatile regions while Random Forest and XGBoost showed 94–99% accuracy at a lead time of 6 months. The accuracy evaluation was based on the confusion matrix and F1 score to analyze the stage change capture. This study demonstrates the effectiveness of using ML and DL algorithms for drought forecasting, with potential applications for other regions.
HIGHLIGHTS
Advanced Forecasting: ML and DL algorithms, including non-contiguous drought analysis, were implemented.;
Data Diversity: ERA5 data on precipitation, temperature, and more is used for model construction.;
High Accuracy: GRU achieves 97-99% accuracy, and Random Forest/XGBoost show 94-99% accuracy at a 6-month lead time.;
Global Relevance: Study highlights ML/DL effectiveness in drought forecasting, applicable to similar regions.; |
|---|---|
| ISSN: | 1998-9563 2224-7955 |