Remote Sensing of African Rainfall

Abstract Climate data support a suite of scientific and socioeconomic activities that can reinforce development gains and improve the lives of those most vulnerable to climate variability and change. Historical and current weather and climate observations are essential for many activities, including operational meteorology, identifying extreme events and assessing associated risks, developing climate-informed early warning systems, planning, and research. Rainfall is the most widely available and used climate variable. Thus, measurement of rainfall is crucial to society’s well-being. In general, measurements from ground meteorological stations managed by National Meteorological Agencies are the principal sources of rainfall data. The main strength of the station observations is that they are assumed to give the “true” measurements of rainfall. However, the distribution of the meteorological observation network over Africa is significantly inadequate, with declining numbers of stations and poor data quality. This problem is compounded by the fact that the distribution of existing stations is uneven, with most weather stations located in cities and towns along major roads. As a result, coverage tends to be worse in rural areas, where livelihoods may be most vulnerable to climate variability and change. This has resulted in critical gaps in the provision of climate services where it is needed the most. Space-based measurements from satellites are being used as a complement to or in place of ground observations. Satellite-derived precipitation estimates offer good spatial coverage and improved temporal and spatial resolution, as well as near-real-time availability. Moreover, a range of satellite rainfall products are freely available from many sources, and a couple of these products are available only for Africa. However, satellite rainfall products also suffer from many shortcomings that include accuracy, particularly at higher temporal resolutions; coarse spatial resolution; short time series; and temporal inhomogeneity due to varying inputs. This limits the use of the use these products for certain applications. Understanding satellite rainfall estimation errors is critical for deciding which products might be used for specific applications and requires rigorous evaluation of these products using ground observations. The challenge in Africa is lack of availability, accessibility, and quality of rain-gauge observations that could be used for this purpose. Despite these challenges, there have been some validation efforts over different parts of the continent. However, different and inconsistent approaches of validation have created challenges to using these evaluation results. A comprehensive validation of the main operational satellite products at a continental level is needed to overcome these challenges and make the best use of satellite rainfall products in different applications.