MAPPING GEOSOCIAL AND GEOEPIDEMIOLOGICAL PATTERNS OF URBAN MALARIA USING REMOTE SENSING DATA: AN EXAMPLE IN A CHAOTIC URBAN CONTEXT

Urban sprawl in developing areas is rapid and mostly uncontrolled. Limited resources do not allow for extended territorial control and urban planning. Information of demographical and geosocial and geoepidemiological origin is particularly lacking. In such a context, cost effective methods need to be developed. We are presenting a mapping method of urban malaria based on demographical and social data proxy from remote sensing. We developed a syntax-based approach from the geospatial dictionary and human ecology concepts. We used high resolution remote sensing data of the city of Yaoundé together with samples of field data to predict key sociospatial and malarial patterns. We used a machine learning method to test the automation possibility of the method. Results showed that behind the apparent chaotic urban context, the method is able to design the intelligence of the social organization of the city with its consequence on the prevalence of malaria.