Investigating the variation of Intra-seasonal rainfall characteristics in Sierra Leone

Abstract

The rainy season in Sierra Leone is unimodal from April to October; an understanding of the inter-annual variability of rainfall in Sierra Leone is of importance to economic sectors such as fisheries, agriculture, infrastructure, hydro-electric power generation (HEP) and water resources. This study investigated the variation of the intra-seasonal rainfall characteristics in Sierra Leone. Daily rainfall data from 4 synoptic stations was obtained from the Sierra Leone Meteorological Department from 1990-2014. The relationship between onsets, cessation, and duration to seasonal totals was determined using Pearson correlation, variability of intra-seasonal rainfall characteristics was determined using coefficient of variation. The magnitudes of the intra-seasonal rainfall characteristics were obtained from the slopes of linear regression lines and the statistical significance was obtained by non parametric method (Mann-Kendall). The results from the study showed that the onset of the rainy season starts early and ends late in the Freetown and Lungi stations resulting to longer durations. The seasonal total of rainfall was also higher at Freetown ranging from 3000-3500mm. On the other hand, at BO station the rainy season starts a bit late and ends late while at Daru, the onset start late and ends early indicate a short duration with seasonal totals of rainfall ranging from 1500-2500mm. The observed phenomenon can be attributed to the fact that Freetown and Lungi are near the coastline hence there is enough source of moisture, while BO and Daru are further inlands hence somehow dry. The results for the correlation analysis indicated negative and significant coefficients for the onsets at Daru. Implying that early onsets results to higher amounts of rainfall and late onsets results to lower amount of rainfall. Both BO and Daru stations depicted positive and significant correlation coefficients between duration of the rains and seasonal rainfall totals. This suggests that as the duration increases the seasonal rainfall amounts increases and vice versa. The spread of onset dates at BO is considerably larger compared to the other three stations hence it is variability is slightly higher (12.4%) compared to the three other stations. The onsets are skewed to the right that is most of the onsets occurs much later. At Freetown the variability of cessation is considerably larger compared to the other three stations. The cessation dates are skewed to the right that is most of the cessation dates occur much later. The spread of duration at Daru is considerably lower with a variability of 4.4%. The variability of seasonal rainfall at Freetown is slightly higher than the other three stations, with a variability of 9.9%. It is also further noted that the variability of the intra-seasonal rainfall characteristics at BO station was slightly higher as compared to the other stations though not significant. The observed trends for onset and cessation dates, as well as duration of the season were not statistically significant except for cessation dates at BO station. The seasonal rainfall totals depicted insignificant decreasing trends for all the stations except Freetown station which indicated an insignificant increasing trend. The results from this study will be useful in addressing the problems associated with late onset and early cessation of rains which has a detrimental effect to farmers in the area of agriculture in terms of food security in the country and also combating the problem of water shortage in industries, dams, reduced stream flow in rivers and watercourses and many more among others.