hide
Free keywords:
-
Abstract:
Diagnosing the physical mechanisms and moisture pathways of high-impact weather events remains a critical challenge in atmospheric science. Currently available models, including HYSPLIT, FLEXPART, and LAGRANTO, are optimised for global datasets such as ERA5. However, they necessitate considerable pre-processing to adequately utilise regional datasets. These models excel at calculating both forward and backward trajectories from a point source; however, they provide limited insights into the dynamics of air and moisture movement and changes in specific humidity around the location of the extreme event. To gain a more comprehensive understanding of extreme events, it is essential to examine the overall moisture dynamics both before and after the event. To address this, we present a novel, open-source Python framework—HPT, designed for comprehensive Lagrangian analysis of atmospheric phenomena. The framework is architected in two modular components: a main simulation engine and a post-processing analysis suite. The simulation engine initialises and advects air parcels in three-dimensional space using gridded meteorological data from reanalysis products, such as ERA5 and IMDAA, while calculating and archiving the evolution of key thermodynamic state variables (e.g., temperature, specific humidity) for each air parcel. The post-processing suite offers a powerful and flexible toolkit for analysing the trajectory dataset, allowing for the generation of a diverse range of diagnostics, including spatial maps of net thermodynamic change, aggregated time-series analyses, conditional vertical profiles, and detailed individual parcel histories. To demonstrate its utility, the new framework is applied to a forensic study of the catastrophic extreme precipitation event that occurred over Kayalpattinam, India, on December 18, 2023. The analysis effectively deconstructs the event, revealing it was driven by a clear moisture pathway originating from the Bay of Bengal. The framework's diagnostics quantify a canonical two-stage parcel lifecycle: a several-day low-level moisture buildup phase, followed by a quick release phase caused by intense, localised ascent. This ascent resulted in rapid adiabatic cooling, which led to intense precipitation. The framework's results offered notably more detailed insights into the event compared to the HYSPLIT and FLEXPART back trajectory analysis. These findings underscore the framework's capacity to link large-scale synoptic patterns with mesoscale processes that influence local impacts, rendering it a valuable tool for research and operational understanding of extreme precipitation events.
