INTEGRATION OF REMOTE SENSING AND STATISTICAL METHOD FOR MONITORING MALARIA PARASITE

Malaria cases and its consequent deaths have been predominant unsolved public health issues in Thailand for decades, despite cases have fallen gradually since 1999. Malaria transmits through three possible mediums; malaria parasite, human hosts, and Anopheles mosquito. One possible way to solve the malaria problem is to have intervention on any of these mediums. This study focuses on Anopheles mosquito medium, a part of the malaria transmission cycle. As the malaria control methods depend on many setting-specific factors such as endemic, vector species and behavior, seasonality, disease patterns, health service factors and more, which they have not been distributed equally in spatial, therefore the accuracy of these predicted information at timely manner are necessary requirements for effective malaria control planning and preparations. Thus, increasing of spatial accuracy and information updates on the vector density are the main issues for the malaria control. In order to support these requirements, Geo-informatics technology is used to develop the model for predicting Anopheles mosquitoes, which is called “Anopheles Mosquito Density Predictive Model (AMDP model)”.

Remotely sensed data and statistical model are integrated to develop the model for predicting Anopheles mosquitoes, which is called “Anopheles Mosquito Density Predictive Model (AMDP model)”. The results found that NDVI values that are higher than 0.501, temperature values with the range of 25-29ºC, relative humidity values with the range of 81-85%, and deciduous forest land cover are the best predictors of the Anopheles mosquito density classes in wet season, while NDVI values that are higher than 0.501, temperature values with the range of 25-29ºC, deciduous forest land cover, and elevation 400-700 meters interval are the best predictors for the Anopheles mosquito density classes in dry season. AMDP model was able to predict correctly 79.7% and 73.8% in wet and dry seasons. This model has passed the model calibration and validation procedures. The results indicate that the model could be applied for prediction of the Anopheles mosquito density in other areas, malaria cases and a tool for decision making system for malaria control planning.