Assessment of rainfall variability and trend using variational indices and modified Mann Kendell-Test in Gicumbi, Northern Province of Rwanda

The sustainability of the agricultural sector and livelihoods is largely influenced by rainfall variability and trends, which are key indicators of global climate change. This study assessed rainfall variability and trends in Gicumbi district using rainfall datasets obtained from the Rwanda Meteorological Agency for the period 1983 to 2021. The k-means clustering method was used to identify homogeneous climatic zones. The coefficient of variability, precipitation concentration index, and rainfall anomaly index were used to identify variabilities in rainfall. Trends direction and their magnitude were computed using Modified Mann-Kendal trend test and Sen’s slope estimator. The results indicated that Gicumbi district is climatologically made of three nearly homogeneous climate zones named Southern Highland, Central Highland, Northern Highland. Rainfall variability was observed to be higher during dry seasons with 64% for June–August (JJA) and 33% for January–February (JF) compared to the wet seasons, which exhibited 21% variability for March–May (MAM) and 20% for September-December (SOND). The annual average variability was 18%, indicating that overall variability remained relatively low. No statistically significant trends were observed in monthly, seasonal, or annual rainfall patterns. The findings also indicated that dry years were more frequent than wet years during the dry seasons, which was the opposite during the rainy seasons. Additionally, 2017 was identified as the driest year throughout the study period. These findings provide valuable insights for improving seasonal forecasting, sustainable agricultural planning, and climate adaptation strategies in the Gicumbi District.