A simple Dash application to examine CMIP5 downscaled climate model outputs over the Northwest Territories, Canada.

All work is funded through SNAP at the University of Alaska, Fairbanks.

To extract the data for NWT locations, run data_prep/extract_profile_snap_deltadownscaled_rasters.py on Atlas. The data.pickle file should be generated locally with make_pickle.py after the extraction is complete.

To run the application locally, install pipenv. This app needs python3 to run; if that's not your default python, adjust the command below (i.e. python3 instead of python).

cd /path/to/this/repo
pipenv install
export REQUESTS_PATHNAME_PREFIX='/' # see below for more info
export MAPBOX_ACCESS_TOKEN='' # <-- insert a valid mapbox token here
pipenv run python application.py

The application will be available at http://127.0.0.1:8080/.

Deployment on AWS

Before deploying, update the requirements.txt file:

pipenv clean
pipenv run pip freeze > requirements.txt
git commit -am'updating requirements.txt'

When deploying on AWS Elastic Beanstalk, a few environment variables must be set using eb setenv:

• MAPBOX_ACCESS_TOKEN: token for API access for Mapbox, no default value.
• REQUESTS_PATHNAME_PREFIX: Path prefix on host, should be / for local development and /tools/nwt-climate-explorer/ for current deploy on AWS.
• DASH_REQUESTS_PATHNAME_PREFIX: URL for file requests, must start and end with /. Should be /tools/nwt-climate-explorer/ for current deploy on AWS.
• eb printenv displays the current environment variables.
• eb deploy deploys the source bundle from the initialized project directory to the running application (e.g. bob-nwt-dash-app-dev).
• eb open will open the URL in your browser.