# Mapping landscape changes to address dengue fever risk in Laos and Thailand

## TL;DR

This study maps land use changes in Laos and Thailand to understand how they affect dengue fever risk by creating better conditions for mosquitoes.

## Contribution

The study uses satellite data and machine learning to link land use changes to dengue vector proliferation in Laos and Thailand.

## Key findings

- Forest cover decreased while settlements and rubber plantations increased in four provinces.
- Urbanization and monoculture plantations may create new mosquito breeding habitats.
- Built-up areas expanded around provincial capitals, increasing artificial breeding sites.

## Abstract

Land use and land cover (LULC) are changing rapidly worldwide, with climate change making living conditions uncomfortable for humans while creating favorable environments for dengue vectors, Aedes aegypti and Aedes albopictus. Laos and Thailand are endemic to dengue, with frequent outbreaks reported. Cleared forests allow vectors to interact with nearby human hosts, potentially transferring viruses from zoonotic reservoirs. Urban areas are often built at the expense of natural hydrology, forests, or farmland. These areas, known as urban heat islands, absorb heat because of construction materials that retain warmth. This, along with water containers or paddles, creates ideal breeding grounds for dengue vectors. Additionally, rubber plantations have been identified as breeding sites for vectors, which can worsen endemic conditions. This study aimed to map LULC changes associated with dengue vector proliferation from 2001 to 2025 using Landsat satellite images (1 to 8). Random Forests classification, a supervised machine learning method, was used to delineate LULC patterns. Satellite-derived LULC data showed a significant decrease in forest cover, an increase in settlements, and a shift in agricultural crops from food crops to cash crops, specifically rubber, confirming changes previously linked to dengue vector ecology. These findings raise serious concerns for public health agencies in both countries and provide valuable insights for cross-sectoral policies in public health, urban planning, forestry, and agriculture sectors.

Image 1

•Random Forest classifier achieved 80–84 % accuracy in classifying LULC changes in Laos and Thailand during 2001–2025.•Forest cover declined in all four provinces studied, with increases in settlements and cash crop (rubber) plantations.•Shift to monoculture plantations may reduce biodiversity and create new mosquito breeding habitats.•Built-up areas expanded around provincial capitals, particularly in Savannakhet (Laos) and Ubon Ratchathani (Thailand).•Urbanization may increase artificial breeding sites and may influence mosquito biting and virus replication.

Random Forest classifier achieved 80–84 % accuracy in classifying LULC changes in Laos and Thailand during 2001–2025.

Forest cover declined in all four provinces studied, with increases in settlements and cash crop (rubber) plantations.

Shift to monoculture plantations may reduce biodiversity and create new mosquito breeding habitats.

Built-up areas expanded around provincial capitals, particularly in Savannakhet (Laos) and Ubon Ratchathani (Thailand).

Urbanization may increase artificial breeding sites and may influence mosquito biting and virus replication.