REACT implements a multi-sensor, multi-scale and multi-temporal approach. The project consists of four broad main phases: data acquisition and preparation , submodel generation, final epidemiological modelling for understanding intra-city and inter-city epidemiological variability. The predictability of inter-urban variations of malaria infection risk will be studied at sub-continental scale in a set of about 20 large cities of sub-Saharan Africa using HRRS and L/MRRS, whereas the intra-urban analysis will focus on a subset of 4 cities using VHRRS and HRRS. The methodology is innovative in many aspects.