Spatio-temporal Trends and Change Detection of Surface Air Temperature in Tanzania: October to April Season.

Abstract

The paper examined the spatio-temporal variability and trends of minimum temperature (T-min), mean (T-mean), and maximum temperature (T-max) in Tanzania from October to April using the Climatic Research Unit Time Series version 4.06 from 1951-2021. The traditional Mann-Kendall (MK) test, modified MK, and Sen’s slope test were used to determine the trend direction and magnitude of the T-mean, T-max, and T-min trends, respectively. Sequential MK, standard normal homogeneity, and Pettitt tests were used to detect abrupt changes, while the Hurst coefficient (H) assessed future trend behavior. Results show a statistically significant increasing trend of 0.14 °C per decade, 0.18 °C per decade, and 0.16 °C per decade in T-mean, T-max, and T-min, respectively, at a 95% confidence level. The H values of T-min, T-max, and T-mean are above 0.7, indicating their future increasing trends. The change points for T-mean and T-max were in the early 1990s, and for T-min in the late 1990s. After 1991, more locations showed significant trends (positive or negative) for T-max than T-mean, and T-min had no significant trend after 1998. The 1960s had the lowest T-max and the 2010s had the highest T-min. Therefore, this study provides valuable information for immediate planning and future climate monitoring in Tanzania.