Beyond Fixed Calendar: Variabilities and Trends in Local Seasons

Precise astronomical dates and fixed meteorological calendars do not capture the near-surface temporal variability of seasons. This study presents a purely data-driven framework to objectively quantify trends in seasons locally, validated using 45 years (1980-2024) of hourly ERA5 data for a subtropical location in India. By integrating spectral analysis for primary interval detection with unsupervised k-means clustering, the framework reveals dynamic seasonal soft boundaries. Two distinct multivariate feature-spaces were studied: "Air Temperature-Low clouds-Rain" and "Dew Point-Rain-Soil Temperature". The heatmaps show three seasons corresponding to Winter, Pre-Monsoon (summer), and Monsoon. There is also a distinct transitional phase occurring twice: Winter to Pre-Monsoon and Monsoon-to-Winter. Trend analyses using Modified Mann-Kendall (MMK), Trend-Free Pre-Whitening (TFPW), and Innovative Polygon Trend Analysis (IPTA) were performed to investigate the annual and seasonal trends across the years. These trend tests on the seasonal soft boundaries confirm a significant temporal contraction of the seasonal phase of Winter by more than a week per decade, driven by the expansion of surrounding regimes. To gain a deeper understanding of this behavior, the same trend analysis is also performed on individual meteorological parameters such as temperature and rainfall between annual and seasonal intervals. The annual daily mean temperature increases at +0.028 • C/year (consistent with existing work on global warming), and the trend for the Annual Daily Minimum Temperature decreases while the Annual Daily Maximum Temperature continues to rise, signaling the widening of the daily temperature range. From a seasonal perspective, however, the Winter daily maximum and minimum temperatures indicate that nights are becoming warmer while days are cooler. Similarly, there is evidence for decreasing total rainfall in the Monsoon season, yet the amount of minimum rainfall is rising. Tracking such variabilities and trends in seasons can enable better economic planning for various domains such as agriculture, outdoor work productivity, and solar energy harvesting. Thus, the current study conducted at S.A.S. Nagar in Punjab, India should motivate extended multi-station investigations across much wider geographies.