LONG-TERM CLIMATE AVERAGES FOR THE STATE OF TOCANTINS: SPATIOTEMPORAL ANALYSIS (1970–2025)

Abstract

This study establishes long-term climate averages (series of 34 to 57 years, five stations with ≥ 40 years) for the state of Tocantins, Brazil, using monthly data from six conventional meteorological stations of the INMET network with historical records spanning from 1970 to 2025. The variables analyzed include rainfall, maximum, minimum and mean temperatures, relative humidity, and sunshine hours. Temporal trend analysis was performed using the non-parametric Mann-Kendall test (Yue & Wang, 2004), with serial autocorrelation correction using the yue_wang_modification_test() function from the pymannkendall package (Hussain & Mahmud, 2019), with trend magnitude estimated by Sen's slope (Sen, 1968). Results demonstrate that all stations with long-duration series exhibit statistically significant warming trends (p < 0.001) in annual maximum and minimum temperatures, with Sen's slopes ranging from +0.024 to +0.067 °C year⁻¹. No significant trend was detected in annual total precipitation, except at Peixe (83228): declining trend (Sen = -9.9 mm year⁻¹; p = 0.009). The established long-term climate averages provide a statistically robust foundation for agroclimate planning, Agricultural Climate Risk Zoning, and hydrological model calibration in the Matopiba region.