Deep Learning-Based Analysis of Precipitation Blocking at the Northwestern Iran Border with Comparative Insights from Utah, USA

Abstract Precipitation blocking along the northwestern Iran–Turkey border significantly affects regional hydroclimate, leading to prolonged dry spells and episodic extreme events. This study employs deep learning models, including CNN-LSTM architectures, to analyze blocking dynamics and the interplay of vortices, Rossby waves, and high- and low-pressure systems. The rapid desiccation of Lake Urmia alters local surface energy and moisture fluxes, reinforcing mesoscale vortices that, together with stationary high-pressure ridges, inhibit the eastward progression of moisture-laden air. Downstream low-pressure troughs and eddy activity further modulate airflow, redirecting precipitation systems toward southern Iraq and the Arabian Peninsula. Jet stream positioning is also crucial, as meridional deviations amplify blocking persistence and influence the seasonal migration of precipitation systems. High-resolution meteorological datasets and satellite observations were integrated with deep learning models to capture subtle topography–vortex–pressure system interactions, which conventional numerical weather prediction often fails to resolve. Comparative analysis with Utah, USA, demonstrates analogous mechanisms where orography-induced ridges, vortices, and jet stream meanders create quasi-stationary highs that block mid-latitude storms, highlighting the transferability of deep learning-based insights across semi-arid mountainous regions. Key findings indicate that precipitation blocking frequency and intensity are highest during cold-season episodes when high-pressure persistence, vortex reinforcement, and jet stream anomalies coincide, and that Lake Urmia’s desiccation exacerbates these effects. This study provides a mechanistic and predictive framework for anticipating blocked precipitation, with direct implications for water resource management, drought mitigation, and climate adaptation in semi-arid transboundary regions. By combining advanced computational methods with cross-regional comparison, the research advances understanding of how complex interactions among vortices, pressure systems, and topography control precipitation distribution under rapid environmental change.

کلیدواژه ها:

Keywords: Precipitation Blocking ، Deep Learning ، Lake Urmia ، Jet Stream ، Vortices ، High- and Low-Pressure Systems ، Northwestern Iran