Analysis of spatial and temporal patterns in onset, cessation and length of growing season in Nigeria

Abstract

Proper understanding of patterns, distribution and changes in rainfall onset dates (ROD), rainfall cessation dates (RCD) and lengths of growing seasons (LGS) is crucial for agricultural planning, adaptation strategies and food security in Nigeria. This study determines the spatial and temporal changes and examines the relationships among these rainfall attributes in addition to possible synoptic connectivity amongst measurement stations. The data used consist of time series of daily rainfall between 1971 and 2005. ROD and RCD were determined using rainfall-based methods proposed and commonly used in Nigeria. The results show that the mean ROD, RCD and LGS have zonal distributions over Nigeria. Onset dates are delayed from the coast northward leading to shorter LGS in those stations. Variability is generally high below 8°N with coefficient of variation (CV) ranging from 13 to 36% for ROD; 3–12% for RCD and 8–17% for LGS. In the middle belt (8–11°N), spatially CV showed values ranging from 10 to 20% for ROD; 3–8% for RCD and 8–21% for LGS. In the Sahel (above 11°N), rains come later in the year and CV showed values ranging from 2 to 9% for ROD; 2–3% for RCD and 12–30% for LGS. Similarly, of the entire Nigeria, 34.8% and 50.0% exhibit earlier RODs and shorter LGS. Whereas in 40.6% and 68.8% of the stations, delayed RCDs and higher SPI (standardized precipitation index) were observed. The observed association between LGS and ROD was very strong, which may be explored to predict potential LGS. Results from the principal component analysis showed that about ten components contain information on possible geo-environmental processes behind variability in onset and cessation across all the synoptic stations of Nigeria. The signatures of SST, ITD and other features of monsoon systems including large scale forcings of SOI and NAO are discussed together with the implications of the changes in the rainfall attributes examined. This information could provide baseline data for development of appropriate local to zonal adaptation measures as well as future studies related to climate change impacts on rainfall attributes considered in this study.