Structural Analysis of the Rainy Season Through Gaussian Approximation

ABSTRACT Motivated by the need to quantify the internal structure of the rainy season and thereby improve the understanding of rainfall seasonality, this study develops a new methodology based on daily precipitation data. The approach decomposes each rainy season into rain modes, separately for seasonal and intra‐seasonal timescales. Daily rainfall is approximated by a sum of Gaussians, capturing the gradual onset, peak, and decay of rainy periods and enabling estimation of each mode's timing, amplitude, and duration. Kernel‐density polar histograms depict the timing of the rainy season and allow robust identification of bimodal regimes. The Seasonality Level ( SL ) quantifies the prominence of the rainy season through the mean ratio of seasonal maxima to the intervening minima, providing an alternative to traditional monthly climatological‐based indices by avoiding the use of calendar boundaries and climatological means. The Modality Level ( ML ) measures the separation between successive intra‐seasonal modes. Application to 19 stations across nine climate types shows that SL ranges from ~2 in tropical rainforest and mid‐latitude wet climates to ~90 in Mediterranean climates, with fully dry summers. ML varies from 3 in tropical and mid‐latitude wet climates to over 50 in arid climates. Intra‐seasonal modes exhibit a typical frequency of 1–2 months, making their annual count negatively proportional to SL ( R = −0.74). A moderate positive correlation between SL and ML ( R = +0.42) reflects a tendency of short‐lived rain events to reduce the prominence of modes at both seasonal and intra‐seasonal timescales. The methodology provides a flexible and objective framework for analysing rainfall regimes across diverse climates, revealing structural features obscured by conventional monthly aggregation. Its ability to summarize annual rainfall into interpretable seasonal and intra‐seasonal modes offers new opportunities for climate classification, hydrological applications, and assessments of rainfall changes under future climate scenarios.