An environmental data set for vector-borne disease modeling and epidemiology

Area of Science:

• Environmental science and public health
• Climate data analysis for disease modeling

Background:

• Vector-borne diseases disproportionately affect low- and middle-income countries.
• Sparse and difficult-to-reconstruct weather data hinders disease transmission studies in these regions.
• Accurate environmental data is crucial for understanding disease ecology.

Purpose of the Study:

• To develop and test methods for assembling high-resolution gridded time series datasets of key environmental variables.
• To address data scarcity for air temperature, relative humidity, land temperature, and rainfall in data-limited regions.
• To improve the foundation for vector-borne disease modeling in low- and middle-income countries.

Main Methods:

• Statistical interpolation of weather station measurements for air temperature and relative humidity.
• Fourier decomposition and time-series analysis to fill gaps in the MODIS11 remote sensing land temperature product.
• Characterization of the RFE 2.0 remote sensing rainfall estimator against interpolated rainfall products.

Main Results:

• Developed methods produced gridded time series data for Madagascar with median 95th percentile absolute errors of 2.75°C for air temperature and 16.6% for relative humidity.
• Successfully estimated missing land surface temperature pixels, offering an alternative to aggregated remote sensing products.
• Identified significant differences in temporal and spatial heterogeneity of rainfall estimates, impacting vector-borne disease modeling.

Conclusions:

• The described methods are effective in generating high-resolution environmental datasets for data-scarce regions like Madagascar.
• These datasets are valuable for enhancing the accuracy and resolution of vector-borne disease transmission models.
• Addressing environmental data gaps is critical for improving public health outcomes in vulnerable populations.