Remote Sensing-Enhanced Spatio-Temporal Model for Precision Prediction of Anopheles Larval Habitats

This repository contains a dataset and accompanying code for a novel spatio-temporal model that leverages high spatial remote sensing data and weather data to predict Anopheles mosquito breeding locations. The model was developed to address the significant global health threat of malaria, which is transmitted by infected female Anopheles mosquitoes.

The dataset includes various independent variables such as geometric features of depressions, river buffers, land cover, soil types, rainfall, and the average Land Surface Temperature (LST) during the Anopheles larval development stage, influenced by Growth Degree Days (GDD). These features were extracted through data preprocessing of high-resolution remote sensing and weather data.

The spatio-temporal prediction model is based on the Logistic Regression Model (LRM) as a Machine Learning (ML) tool. The model was trained and validated to accurately classify high-risk, moderate-risk, and low-risk Anopheles mosquito breeding sites. Statistical analysis methods and field observations were used to evaluate the model's effectiveness, revealing a negative correlation between high-risk depressions and both average LST and depression areas.

This research provides a valuable tool for targeted interventions and preventive measures against malaria transmission. The dataset and code in this repository can be used by researchers and public health professionals to further develop and apply the spatio-temporal model in different contexts, contributing to the enhancement of population control strategies and public health management.