Geospatial big data methods enable mapping potential habitats

To enhance existing strategies for controlling the Aedes aegypti mosquito, geoinformation scientist Dr Steffen Knoblauch has created a high-resolution environmental suitability map for Rio de Janeiro (Brazil) that can help identify areas most conducive to breeding. It is based on advanced geospatial big data methods – leveraging openly available geodata such as satellite imagery, street view images, and climate data – that the researcher developed at Heidelberg University’s Interdisciplinary Center for Scientific Computing (IWR) and at HeiGIT (Heidelberg Institute for Geoinformation Technology).

The mosquito species Aedes aegypti transmits infectious diseases such as dengue fever, Zika, chikungunya, and yellow fever. Also known as the Egyptian tiger mosquito, the insect favors breeding in stagnant water accumulating in artificial containers including, for instance, water tanks, discarded tires, potted plants, and storm drains, as Dr Knoblauch, a researcher in the Geoinformatics research group at the IWR, explains. Suitable habitat areas are expanding due to global trends such as increasing urbanization. Because the global availability of vaccines for the diseases the mosquito transmits are still limited, with the exception of yellow fever, vector control – eliminating breeding habitats or the targeted spraying of insecticides – is currently the most effective method of control.

“To achieve efficient and cost-effective vector control in heterogeneous urban areas, accurate Aedes aegypti suitability maps are essential,” explains Dr Knoblauch. Yet generating spatially continuous maps is especially challenging for heterogeneous landscapes featuring different topography and built-up environment. The mosquito’s limited flight range, assessed to be below 1,000 meters without the assistance of wind, and the heterogeneous urban landscape, which influences the availability of breeding sites, can lead to a high spatial variability in Aedes aegypti presence. Measuring this variability is difficult using conventional sample-based entomological monitoring systems, as Dr Knoblauch emphasizes.

The researcher hypothesized that openly available geodata – such as satellite and street view imagery – could help overcome these limitations. This hypothesis was supported by findings showing that such data can be used to model up to 75 percent of the spatial variation observed in field-collected entomological measurements in the city of Rio de Janeiro. From the geodata, Dr Knoblauch retrieved 79 environmental suitability indicators for Aedes aegypti with the help of Geospatial Artificial Intelligence (GeoAI) and spatio-temporal modeling. These included among other factors relative breeding container density, urban morphology measures, as well as climate variables like rainwater accumulation and urban heat island effects. Using a Bayesian model, he estimated mosquito presence across space and time while accounting for uncertainty.

From this resulted the first spatially continuous Aedes aegypti environmental suitability map at the habitat scale, offering critical guidance for more targeted interventions, particularly in Aedes-endemic urban areas. “This method of combining openly available geodata with entomological surveillance can be transferred to other similar environmental settings to produce mosquito distribution maps that capture the high spatial variability of their habitats,” says Dr Knoblauch. “This allows us to predict where mosquitoes are likely to breed with much greater precision than before.”

Several research groups from Heidelberg University and Heidelberg University Hospital contributed to the work, including those of Prof. Dr Joacim Rocklöv, Prof. Dr Dr Till Bärnighausen. and Prof. Dr Peter Dambach at the Heidelberg Institute of Global Health, and the Geoinformatics department led by Prof. Dr Alexander Zipf. In addition, researchers from Brazil, the United Kingdom, Austria, Switzerland, Singapore, Thailand, and the USA contributed to the research findings. Funding for the project was provided by the German Research Foundation and the Austrian Science Fund. The results were published in the journal “The Lancet Planetary Health”.