Adole, Tracy and Dash, Jadunandan and Atkinson, Peter M. (2018) Characterising the land surface phenology of Africa using 500 m MODIS EVI. Applied Geography, 90. pp. 187-199. ISSN 0143-6228
Accepted_Manuscript.pdf - Accepted Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (863kB)
Abstract
Vegetation phenological studies at different spatial and temporal scales offer better understanding of the relationship between the global climate and the global distribution of biogeographical zones. These studies in the last few decades have focussed on characterising and understanding vegetation phenology and its drivers especially using satellite sensor data. Nevertheless, despite being home to 17% of the global forest cover, approximately 12% of the world's tropical mangroves, and a diverse range of vegetation types, Africa is one of the most poorly studied regions in the world. There has been no study characterising land surface phenology (LSP) of the major land cover types in the different geographical sub-regions in Africa, and only coarse spatial resolution datasets have been used for continental studies. Therefore, we aim to provide seasonal phenological pattern of Africa's vegetation and characterise the LSP of major land cover types in different geographical sub-regions in Africa at a medium spatial resolution of 500 m using MODIS EVI time-series data over a long temporal range of 15 years (2001–2015). The Discrete Fourier Transformation (DFT) technique was employed to smooth the time-series data and an inflection point-based method was used to extract phenological parameters such as start of season (SOS) and end of season (EOS). Homogeneous pixels from 12 years (2001–2012) MODIS land cover data (MODIS MCD12Q1) was used to describe, for the first time, the LSP of the major vegetation types in Africa. The results from this research characterise spatially and temporally the highly irregular and multi-annual variability of the vegetation phenology of Africa, and the maps and charts provide an improved representation of the LSP of Africa, which can serve as a pivot to filling other research gaps in the African continent.