Kerala Develops Dengue Forecasting System
Kerala’s AI Model Predicts District-Level Dengue Risk: Scientists at Kerala’s Institute of Advanced Virology (IAV) have developed a Dengue Early Warning System (DEWS) that uses machine learning to identify possible changes in dengue transmission at the district level.
The system combines disease surveillance with meteorological information to generate weekly forecasts for individual districts. It was developed using approximately six years of data from March 2020 to February 2026.
Initial evaluation using Kerala surveillance data from March to July 2026 indicated a positive relationship between the model’s forecasts and observed dengue patterns.
Four Key Inputs
The DEWS integrates epidemiological information from Kerala’s State Surveillance Unit with weather observations supplied by the India Meteorological Department (IMD).
The model considers four principal variables:
- Reported dengue cases
- Rainfall
- Temperature
- Average humidity
Unlike a single state-wide forecast, the system produces district-specific risk assessments. This allows local disease patterns and variations in weather conditions to be incorporated into the prediction.
The resulting risk is grouped into very high, high, moderate and low categories, making the model’s output easier for public-health authorities to interpret and act upon.
Static GK fact: Dengue is a viral infection transmitted mainly by infected Aedes mosquitoes. Aedes aegypti is one of the principal vectors responsible for dengue transmission.
Weather and Dengue Transmission
Weather conditions influence dengue transmission through their effects on mosquito breeding, survival and viral development.
Rainfall can create temporary water-filled breeding sites, while temperature affects mosquito development and the time required for dengue virus to become infectious within the mosquito. Humidity can also influence mosquito survival.
However, climate is only one component of dengue transmission. Urbanisation, human movement, water-storage practices, mosquito density and population immunity can also alter disease patterns.
Static GK Tip: Dengue transmission often displays seasonal variation, with transmission increasing during or after rainy periods in many tropical and subtropical regions.
Evidence From Kerala Research
A separate 2026 study published in GeoHealth examined dengue and climatic data from Kerala covering 2006–2019. It found a clear seasonal association between dengue incidence and the monsoon period.
Approximately 60% of cases occurred between June and September, while the highest incidence was observed during June and July. Temperature, rainfall, relative humidity and ENSO were identified among important climatic predictors.
The study’s machine-learning analysis reported an R² value of 0.72 for the XGBoost model during its test period. It also projected increased dengue incidence under certain future climate scenarios, although such projections remain model-based estimates rather than certain outcomes.
Public Health Importance
The DEWS cannot itself prevent or treat dengue. Its importance lies in providing authorities with advance warning, allowing preventive and healthcare measures to be initiated before disease activity increases.
A district classified as high risk could strengthen mosquito surveillance and control, increase diagnostic and laboratory preparedness, and ensure that healthcare facilities are ready for a possible rise in dengue patients.
Early Warning and Preparedness
The World Health Organization (WHO) also recognises early-warning and response systems as useful tools for anticipating dengue outbreaks and improving preparedness.
Kerala’s initiative demonstrates how machine learning, disease surveillance and weather data can be combined for geographically targeted public-health planning. However, DEWS remains a research system and requires continued validation, updating and refinement before wider operational use.
Static Usthadian Current Affairs Table
Kerala’s AI Model Predicts District-Level Dengue Risk:
| Fact | Detail |
| System | Dengue Early Warning System (DEWS) |
| Developed By | Institute of Advanced Virology, Kerala |
| Technology | Machine learning |
| Data Period | March 2020 to February 2026 |
| Forecast Frequency | Weekly |
| Forecast Level | District-wise |
| Disease | Dengue |
| Main Disease Input | Reported dengue cases |
| Weather Inputs | Rainfall, temperature and average humidity |
| Surveillance Source | Kerala State Surveillance Unit |
| Weather Data Source | India Meteorological Department |
| Risk Categories | Very high, high, moderate and low |
| Initial Evaluation | March to July 2026 |
| Earlier Study Period | 2006–2019 |
| Peak Dengue Incidence | June–July |
| Monsoon-Period Cases | About 60% during June–September |
| Earlier ML Model | XGBoost |
| Reported R² | 0.72 |
| Major Vector | Aedes mosquitoes |
| Main Objective | Early warning and public-health preparedness |
| Current Status | Research system requiring further validation |
