Satellite Earth observation to support sustainable rural development

Traditional survey and census data are not sufficient for measuring poverty and progress towards achieving the Sustainable Development Goals (SDGs). Satellite Earth Observation (EO) is a novel data source that has considerable potential to augment data for sustainable rural development. To realise the full potential of EO data as a proxy for socioeconomic conditions, end-users – both expert and non-expert – must be able to make the right decisions about what data to use and how to use it. In this review, we present an outline of what needs to be done to operationalise, and increase confidence in, EO data for sustainable rural development and monitoring the socioeconomic targets of the SDGs. We find that most approaches developed so far operate at a single spatial scale, for a single point in time, and proxy only one socioeconomic metric. Moreover, research has been geographically focused across three main regions: West Africa, East Africa, and the Indian Subcontinent, which underscores a need to conduct research into the utility of EO for monitoring poverty across more regions, to identify transferable EO proxies and methods. A variety of data from different EO platforms have been integrated into such analyses, with Landsat and MODIS datasets proving to be the most utilised to-date. Meanwhile, there is an apparent underutilisation of fusion capabilities with disparate datasets, in terms of (i) other EO datasets such as RADAR data, and (ii) non-traditional datasets such as geospatial population layers. We identify five key areas requiring further development to encourage operational uptake of EO for proxying socioeconomic conditions and conclude by linking these with the technical and implementational challenges identified across the review to make explicit recommendations. This review contributes towards developing transparent data systems to assemble the high-quality data required to monitor socioeconomic conditions across rural spaces at fine temporal and spatial scales.