Space and MATE project
The emergence of dengue as a major public health problem has been dramatic in the American region.
In Argentina, the first Dengue Epidemic was recorded between February and March 1916. During 70 years the country was free of Dengue until April 1997 when DEN-2 cases were detected at Oran, Salvador Maza, Guermes and Tartagal cities.
In 1998, and during 2002, epidemic took place in Salta, Misione and Formosa provinces.
In this context, the National Health Minister and CONAE have established a cooperation program with the objective of develops new tools of vectorial and sanitary surveillance using GIS and environment parameters monitored from the space.
During 2004 a particular effort was done to create statistical model for predicting vectorial indices of Aedes in the Nord Argentine.
The objective is to build a predictive model of the spatial distribution of cases based in remote sensing technology.
An environmental risk prediction model was developed based on synthetic multiband image created from Landsat 5TM data (satellite images) and its correlations with 2004 dengue epidemy risk map. Results show that heterogeneity in space and time epidemy distributions are directly related to environmental variations measurable by remote sensing. This fact demonstrates the importance and potential of the use of Remote sensing data (Landscape Epidemiology tools) and spatial statistics for elaborating and testing the efficiency of dengue fever surveillance strategy and dengue prevention programs.
Tartagal city (Salta) was choose as a pilot site.
The city of Puerto Iguazu, Argentina (25 ° 36’S - 54 ° 35’O), is a other study area, in the MATE project. The main aims is to use satellite imagery high spatial resolution, to create products for the epidemiological surveillance of dengue. For that, SPOT 5 satellite image, is used to do a classification based on vegetation indice and allows to obtain a map of land use.
Various environmental variables is extracted such as : map of distance to each class (buffer), percentage of each class by area (type of soil, vegetation, water, etc.)and statistical indices by area.
The pockets of positive vector is used as training areas for a supervised classification of environmental variables derived.
The resulting map, shows the predictive potential areas for the development of the vector in study and the abundance of positive foci per hectare, predicting 87% of the positive areas between the Middle and High levels of environmental sensitivity.
All the study is explain in specific article (in Spanish).