Remote sensing
Food crises are a regular occurrence in Madagascar. Remote sensing – using technology to measure farmland – is helping the country to tackle this issue head-on. Researchers are utilising satellite images and spatial modelling to estimate crop yields and identify the best ways to combat diseases.
Madagascar is a vast country with many difficult-to-access areas. Remote sensing offers a viable alternative to ground-based agricultural production measurement systems. The system uses satellite imagery, spatial modelling and spatialised databases to gain a better understanding of the layout of environments and activities, optimise the spatial distribution of alternative technologies and available resources, and predict the aggregated impacts of major changes to activity and space management.
Research has been carried out in the Antsirabe region, in the Madagascan highlands, 60 km from the capital Antananarivo. The aim of this research was to produce a land use map, focusing in particular on farmland and cropping systems, using satellite images from different sources. The researchers then developed a set of rice yield estimation methods by analysing satellite images taken at different times and carrying out spatial modelling calculations with a view to identifying the best ways to combat the spread of crop diseases.
The research produced no fewer than three probing results. The first was a highly detailed map of farmland in the area of Antsirabe covered by the study during the October 2014 to May 2015 growing season. This precise information is particularly important given that farmland in the area is highly fragmented, mainly comprising of small plots of land surrounded by natural vegetation. It proved more difficult to produce a map of cropping systems (i.e. all processes used to farm the land), and the result was less accurate and is open to improvement. The researchers were then able to calculate rice crop yield estimates. They used ‘Normalised Difference Vegetation Index’ figures, calculated using the satellite images, to model the crop growing cycle in the region’s rice fields. The first set of rice yield estimation models, produced via remote sensing (estimates of total biomass, straw and seed mass), is encouraging. The researchers, along with students and development partners, were then trained on how to use the quantum geographic information system (QGIC). These research activities – supported by substantial funding from the French national centre for space studies, Centre National d’Etudes Spatiale – were carried out under the auspices of the Joint Experiment for Crop Assessment and Monitoring, an international network that specialises in the use of remote sensing technology to monitor agricultural production. Free data from the European Space Agency’s ‘Sentinel-2’ mission – a constellation of high-resolution land monitoring satellites – will soon be available. This should provide additional capability for monitoring and modelling agricultural systems in Madagascar.
