Assessing climate anomalies in the strait of Hormuz using gradient boosting and remote sensing–based environmental parameters

Strait of Hormuz is a climatically sensitive marine transition zone in which interplay of complex air sea interactions, monsoonal forcing, and land ocean thermal contrasts produces a strong impact on the variability of environment in the region. The long term climate behavior of such confined coastal systems is a difficult issue to assess because of the nonlinear connections between the atmosphere and ocean, as well as vertical thermodynamic interactions and the sparsity of in situ measurements. To overcome these limitations this paper proposes Strait-ClimateAssessNet, a data-oriented assessment model of near-surface temperature anomalies on the Strait of Hormuz. The framework combines assimilated environmental variables based on ERA5 reanalysis data between 1971 and 2025 with diagnostics of climatological anomalies, vertical pressure, and lag representations which are maintained in time to maintain physical coherence in climate evolution. Through the combination of remotely sensed atmospheric and oceanic fields, the proposed method is able to capture the long term tendencies of warming as well as the interannual variability superimposed on it through the use of a single learning architecture. The model is executed in a reproducible Python-based setup with the use of the contemporary scientific computing libraries that guarantee transparency and scalability. Despite the improved anomaly reconstruction capability demonstrated by quantitative evaluation, root mean square error of 0.4327 °C and a mean absolute error of 0.3304 °C are more indicative of regional thermal dynamics when non-stationary, as opposed to stationary conditions. The findings emphasize the topography of coupled surface atmosphere processes in defining the variability in temperature in the Strait. In general, Strait-ClimateAssessNet represents a computationally-efficient and physically-consistent model of long-term climate assessment in dynamically constrained coastal zones which is useful in environmental monitoring, climate risk analysis, and regional climate-sensitive planning.