Coupling population and the environment: exploring the use of satellite data as a means of estimating livelihoods

Relationships are often found between rural livelihoods and environmental variables (e.g. woodland and non-timber forest products can have a positive impact on rural incomes). In many developing countries the poorest populations are often located in marginal locations that are unsuitable for sustained habitation such as those affected by flooding and erosion. Satellite data is uniquely placed to provide fine temporal and spatial resolution data for rapid assessment of environmental conditions. Therefore, if relationships can be found between livelihoods and environmental metrics derived solely from satellite data, it may be possible to create a basic estimate of local livelihoods using remotely sensed data. Data from the Millennium Villages Project (MVP) were used to explore the relationships between socioeconomic conditions and environmental variables derived from satellite data. Statistical associations from spatial regression techniques were identified between satellite-derived environmental metrics and socioeconomic conditions which suggest that there is potential to infer basic livelihood conditions from satellite data. This could provide useful information for identifying vulnerable populations to improve geographic targeting of development assistance, monitoring progress towards the MDGs.