Hydroclimate extremes driven by decadal atmosphere-oceanic modes in tropical and subtropical regions: a case study in the Greater Bay Area, South China

Abstract:

Decadal atmosphere-oceanic modes control persistent droughts and flooding. At the subtropical Greater Bay Area (GBA) in South China, a multiscale analysis approach is developed to investigate how the atmosphere-ocean modes control hydrological extremes. Extreme indices related to flooding and droughts are derived using historical datasets, the China Meteorological Administration (CMA) reanalysis data, and the simulations from the Noah-Multiparameterisation Land Surface (NoahMP) models that are driven by the Global Land Data Assimilation System (GLDAS) data. In empirical spectrums of hydroclimate data, interannual and decadal peaks are found to be significant for both precipitation and runoff time series in the GBA based on a red-noise null hypothesis. Persistent droughts and flooding in the GBA are shown to be related to interannual and decadal atmosphere-ocean oscillations in the Pacific. Although the atmosphere-ocean oscillations have varying influences on different precipitation extreme percentiles in various parts of the region, the cumulative precipitation extremes of 5 or more days in the GBA are largely related to the Central Pacific sea-surface temperature (SST) variations. Based on the NoahMP runoff simulations, the SST variations of the Pacific decadal oscillations are related to the runoff drought indices exemplified for the Donjiang and Xijiang rivers inside the GBA. The warm and cold phases of the Pacific SST variations affect the moisture fluxes and deep convection mechanisms by interacting with the Asian monsoon. Based on these results, a new framework for flood and drought risk management is proposed based on decadal atmosphere-oceanic modes. The proposed framework is expected to be valuable to characterise wet and dry regimes in tropical and subtropical regions, to improve seasonal forecasts based on the scale continuum of hydroclimate time series.