Venugopal Thallam: Quantifying and Reducing Uncertainties in Studying Atmospheric Rivers and Moisture Transport: From Data and Heuristics to Scaling and Impacts

Abstract

Reducing uncertainties in mapping atmospheric rivers (ARs) and their associated meteorological extremes is crucial for developing effective strategies to mitigate hazards and adapt to climate change. This study uses Bayesian AR detection from the Toolkit for Extreme Climate Analysis to assess uncertainties in AR detection across the pan-Atlantic and moist tropical regions with long-term, high-resolution reanalysis data. Significant biases exist among reanalysis products when estimating AR intensities, with winds and specific humidity in the lower atmosphere being key factors in determining total column water vapour and AR strength. Recent trends show an increase in the intensity of ARs in the North Atlantic, alongside notable decadal variability and a poleward shift. Moisture flux sources in the open ocean also exhibit strong latitudinal dependence, impacting AR formation and enhancement. 

A large spread in aggregated AR probabilities results in differences in attributes like frequency, intensity, and impacts on weather extremes over Europe. Model configurations and sensitivity to boundary conditions amplify these biases over Western and Northern Europe, especially in ocean regions where ARs form. These biases persist across climate modes, such as strong positive El Nino-Southern Oscillation. The new Uppsala University AR scale addresses limitations in assessing AR impacts tailored for the Pan-Atlantic region. Alternative metrics, such as the AR Severity Index and Risk Index, are proposed to comprehensively evaluate AR impacts, blending physical strength with contextual factors. These refined metrics and new AR scale enhance understanding of AR impacts, contributing to better forecasting, disaster preparedness, and water resource management. Alongside the moisture transport, multiple dynamic and thermodynamic processes influence extreme precipitation events in tropical moist environments. A multi-faceted approach that enhances financial and technological resources integrates AI and big data and prioritises community preparedness to improve forecasting and early warning.