Forecast dengue outbreaks. Part of 510's impact-based forecasting portal. Built to support Philippines Red Cross.
This model:
- calculates expected mosquito abundance given meteorological data
- forecast dengue risk given mosquito abundance
- determines if dengue risk is anomalously high and, if so, gives an alert
- calculates the expected number of dengue cases given dengue risk
See details in the IBF-dengue technical note.
If you want to run this model on Azure, follow these instructions.
Generic requirements:
For 510: project service account accessible here, login credentials in Bitwarden
mosquito_model/input/
- Copy your Google Earth Engine service account credentials (stored as .json) and IBF-system credentials (stored as .env) in
mosquito_model/credentials/
- Build the docker image from the root directory
docker build -t rodekruis/mosquito-model .
- Create a docker container
docker run --name mosquito-model rodekruis/mosquito-model
- Run and access the container
docker run -it --entrypoint /bin/bash rodekruis/mosquito-model
- Check that everything is working by running the model (see Usage below)
Specific requirements:
- Move to project root
cd mosquito_model
- Install project
pip install .
- Download input data from here and move it to
mosquito_model/input/
- Copy your Google Earth Engine service account credentials (stored as .json) and IBF-system credentials (stored as .env) in
mosquito_model/credentials/
Usage: run-mosquito-model [OPTIONS]
Options:
--countrycode TEXT country iso code
--vector TEXT vector file with admin boundaries
--admincode TEXT name of admin code in vector file
--temperaturesuitability TEXT table with suitability vs temperature
--thresholds TEXT table with thresholds and coefficients (risk vs dengue cases)
--demographics TEXT table with demographic data
--credentials TEXT credentials directory
--data TEXT input data directory
--dest TEXT output data directory
--predictstart TEXT start predictions from date (%Y-%m-%d)
--predictend TEXT end predictions on date (%Y-%m-%d)
--storeraster store raster data locally
--verbose print output at each step
--help show this message and exit