Flash Drought: Review of Concept, Prediction and the Potential for Machine Learning, Deep Learning Methods

Abstract

This paper reviews the Flash Drought concept, the uncertainties associated with FD prediction, and the potential of Machine Learning (ML) and Deep learning (DL) for future applications. For this, 121 relevant articles covering different aspects of FD ‐ definitions, key indicators, distinguishing characteristics, and the current methods for FD assessment (i.e., ‐ monitoring, prediction, and impact assessment) are examined. FD is typically a short‐term drought event ‐ characterized by the rapid progression of heat waves and precipitation deficits, causing cascading impacts on the land and surface hydrology. FD prediction is constrained by the lack of consistent FD definitions, key indicators, the limited predictability of FD at the subseasonal‐ to‐seasonal (S2S) timescale, and uncertainties associated with the current prediction methods. Some of the uncertainties in the current methods are associated with a lack of our understanding of the physical processes. They are also related to the error in the input datasets (imperfect representation of indicators), parameter uncertainty (parameterization scheme adopted by the prediction model), multicollinearity, nonlinear, and non‐stationary interactions among different indicators. Combining traditional methods and multisource fusion data with ML and DL methods shows promise to better understand FD evolution and improves prediction.

Increasing Risks of Future Compound Climate Extremes With Warming Over Global Land Masses

“…Here BDHI and BPHI were mainly utilized to monitor compound dry-hot events and compound pluvial-hot events at the seasonal scales, respectively, how does the monitoring ability of these two indexes in the shorter time scales (e.g., semimonthly)? In recent decades, many regions around the globe have been attacked by compound climate extremes with rapid onset and short duration (Pendergrass et al, 2020;Tyagi et al, 2022;Yuan et al, 2023), which poses a serious threat to agricultural production, water security, and ecological health. Therefore, the risk evaluation and quantification of compound climate extremes in a shorter time scale should also receive attention.…”

High Temperature Accelerates Onset Speed of the 2022 Unprecedented Flash Drought Over the Yangtze River Basin

“…To characterize the rapid evolution of flash drought at the sub‐seasonal time scale (Chen et al., 2019; Mohammadi et al., 2022; Tyagi et al., 2022), the 2022 YRB flash drought was analyzed at pentad (5‐day) time scales during June–August, and the percentiles of precipitation, temperature, ET, and SM at each pentad were calculated separately based on the same period during 1961–2021. The SM percentile is an important indicator for flash drought analysis (Ford & Labosier, 2017; Qing et al., 2022; Yuan et al., 2019; Zhu & Wang, 2021), and was used in this study to identify flash drought and analyze its onset speed and intensity (Figure S1; Text S2 in Supporting Information ).…”

Flash drought: A state of the science review

“…To improve flash drought forecast skills and complement dynamical forecasting systems, statistical and hybrid statistical-dynamical prediction models have been developed. In these models, linear regression and more advanced nonlinear machine learning (ML) and deep learning (DL) methods were applied to account for the dependence of flash drought predictands on predictors drawing from observed current and past atmospheric and land surface states, S2S dynamical forecasts, and potential sources of flash drought predictability (e.g., ENSO, MJO; Lorenz et al, 2017Lorenz et al, , 2018Lorenz et al, , 2021Tyagi et al, 2022). Results show considerable skill improvement for the lead times of up to 4 weeks, where land initial state is the dominant contributor with dynamical forecasts playing a secondary contributing role, suggesting the importance of improving land initialization in dynamical forecasting systems.…”