Long-term land use change in Puerto Maldonado, Peru and implications for dengue surveillance and control

Background: Environmental factors remain a major cause of morbidity and mortality throughout the world. Anthropogenic pressures, especially rapid unplanned urbanization, are linked to crowding, air pollution, inadequate water and sanitation systems, and poor housing conditions, which are directly related to poor health conditions. Remote sensing technologies have the ability to measure and predict ground-level environmental changes that are known to contribute to increased diseased incidence. In addition, they can distinguish the ground conditions most likely to create ideal breeding habitats for vectors of diseases such as dengue.

Objective: (1)To characterize land-use change for the region of Puerto Maldonado, Peru between 1986 and 2008, using Landsat TM+ and ASTER imagery; and (2) to describe the implications for dengue surveillance and control.

Methods: The project acquired six satellite images of the study; three Landsat TM+ images (1986, 1996, and 2005) and three ASTER images (2002, 2006, and 2008). Using a supervised pixel-based classification algorithm, we developed a land-use / land-cover classification map for each image and calculated change over time for each land class. Classes were verified through ground-truthing performed in May 2011. We quantified accuracy of the classification in terms of producer's (Type I) and user's (Type II) accuracy, and Cohen's Kappa.

Results: Significant changes in land cover occurred over the study period. The study showed an increase in urban, rangeland/agriculture, and barren land areas, and decreased forested areas between 1986-2005 and 2002-2008.

Conclusions: This study has important implications for the development of novel, cost-effective and efficient broad-scale surveillance technologies for inter-epidemic periods.