Skip to content

ClimAID - Climate Change impact using AI on Diseases

ClimAID is an integrated toolkit for modeling, forecasting, and projecting climate-sensitive diseases such as dengue and malaria using machine learning and climate model ensembles.

  • ClimAID has inbuilt climate data for South Asian countries, namely India, Nepal, Bhutan, Sri Lanka, Myanmar, Afghanistan, Pakistan and Bangladesh.

  • ClimAID support data from other countries through the global mode on the browser interface.

What you can do

  • Analyze historical disease patterns
  • Integrate climate variables (temperature, rainfall, humidity)
  • Automatically detect optimal lag effects using AutoML
  • Train hybrid ML models
  • Generate CMIP6-based future projections
  • Identify outbreak risk under climate change scenarios
  • Generate automated policy reports using the integrated C-DSI or local LLM models.

Quick Example using Codes

from climaid.climaid_model import DiseaseModel

dm = DiseaseModel(
    district='IND_Mumbai_MAHARASHTRA',
    disease_file="dengue_data.xlsx",
    disease_name="Dengue"
)

dm.optimize_lags()
dm.train_final_model()

Workflow

ClimAID has two interfaces,

  • ClimAID Browser Interface (For both South Asian + Global countries)

    For initialisation through terminal, use 

        climaid browse
    ClimAID Workflow

    Figure 1: Guide to use the ClimAID Browser Interface.

  • ClimAID Wizard Interface (For South Asian countries)

    For initialisation through terminal, use 

        climaid wizard
    ClimAID Workflow

    Figure 2: Guide to use the ClimAID Wizard Interface.

Documentation

Use the sidebar to explore:

  • API reference for all modules
  • Modeling pipeline
  • Projection tools
  • Reporting system

Designed for

  • Epidemiologists
  • Climate scientists
  • Public health analysts
  • Data scientists

Research-ready

ClimAID supports:

  • Reproducible workflows
  • Multi-model ensembles (CMIP6)
  • Policy-oriented outputs

Built for climate-health intelligence

Development Underway

  • Support for more models

    • Spatiotemporal ML models
    • Integrated SEIR-ML hybrid framework

  • More user choices (Priority)

    • Support for selection of covariates
    • Additional covariates may be added by the user
    • Support for selection of LLMs through Browser Interface

  • For technical feedback, please email: avik.sam@iitb.ac.in

Dependencies & License

Dependencies

  • Core Requirements

    pandas  
numpy  
geopandas  
matplotlib  
scikit-learn  
xarray  
regionmask  
plotly  
xgboost  
optuna

  • Additional Utilities

    These packages support extended functionality and will be auto-installed. 

    requests
joblib
fastapi
uvicorn
typer
markdown
fastparquet
python-multipart
seaborn
openpyxl

  • Optional (LLM Support)

    To enable local LLM-based report generation:

    pip install climaid[full]

    Includes: 

    ollama

License

Designed by Avik Kumar Sam & Harish C. Phuleria as an open-access software.