Probability links between meteorological drought and hydrological drought from a 3D perspective

Abstract

Study region: Lower Mekong River Basin (LMRB), Southeast Asia Study focus: The links between meteorological drought (MD) and hydrological drought (HD) from a 3D perspective have not been fully investigated. A 3D identification method was employed to identify MD and HD events in the LMRB. Response of HD to MD were evaluated using linear and nonlinear models. Furthermore, the occurrence probabilities and propagation thresholds (PTs) from MD to HD were examined by applying a copula joint distribution and conditional probability based on a Bayesian network (CPBN). New hydrological insights for the region: A total of 81 MDs and 87 HDs were identified during 1950–2021. HDs had longer durations but affected a smaller area and were less severe than MDs, with their centroids mainly located in eastern Thailand. 46 pairs of MD-HD events were matched spatiotemporally, revealing that a HD event in typical drought years could be driven by multiple MD events. The duration, area, and severity of these drought pairs were optimally fitted using linear, 3P-power, and polynomial function models, respectively. Additionally, the occurrence probabilities of moderate, severe, and extreme HDs successively decreased under the same MD conditions. The PT from MD to HD increased with increasing severity levels of HDs. The study of the links between MD and HD from a 3D perspective provides a more comprehensive understanding of drought in the LMRB.

Keywords

• Meteorological drought
• Hydrological drought
• 3D identification
• Response model
• Occurrence probability