Mosquito Breeding Hotspot Prediction Tool
The city of Hyderabad is now an urban sprawl of about 650 sq. kms. delineated into several administrative zones and municipal wards. Onset of monsoon each year is characterized as the beginning of a high risk period for vector-borne diseases, leaving the city residents and its health infrastructure on their toes. For management of this public health issue, a geospatial dashboard was envisaged under the aegis of a grant from US-India Science & Technology Endowment Fund (USISTEF). This dashboard would be used to identify 'high-risk' areas for mosquito breeding or in short - mosquito breeding hotspots.
● A detailed research was carried out on environmental and other factors that contribute to vector proliferation.
● A long list of such variables was created and thereafter a shortlist was made with the variables that were to be used as predictors for identifying potential breeding hotspots.
● Data pipeline from Satellite platforms to open source weather platforms was brought together onto a single platform, in a unified format and aggregated to a grid cell size of 100m x 100m.
● A risk score was calculated for each grid cell followed by spatial clustering analysis that would identify clusters of high risk score cells bunched together.
● Data from a partner organization's mosquito traps was used to know mosquito type and their count. This was fed back into the risk analysis loop to fine-tune the weightage each variable got and thus, make the algorithm even more robust.
● Precision strategies being developed by using a tool of such granularity and detailing.
● Resources can be focused on high-risk areas thereby reducing the mosquito populations, and mosquito-borne diseases
● Avoiding strategy of insecticide carpet-bombing that were deployed otherwise, in the absence of such a data-driven precision tool.
