https://apps.nrs.gov.bc.ca/gwells/
The Ministry of Environment receives and processes groundwater data and information related to the construction, alteration and decommissioning of groundwater wells. Well construction and reporting requirements are regulated under the Water Sustainability Act and Groundwater Protection Regulation. The information collected and stored is used by government and other users to help inform decisions related to the management of the groundwater resource in B.C.
GWELLS, the new groundwater data repository, aims to improve the user experience when submitting and searching for well information, to improve the quality of the data being submitted, and to improve the overall functionality of the system to meet user and regulatory requirements.
The application is being developed as an open source solution.
GWELLS maintains a REST API where public data relating to wells is made available for other applications and services.
Our Swagger documentation is available at https://apps.nrs.gov.bc.ca/gwells/api/. Some examples of GWELLS endpoints:
Wells:
# all wells (paginated)
curl https://apps.nrs.gov.bc.ca/gwells/api/v1/wells
# all wells in an area bounded by sw_lat, sw_long, ne_lat, and ne_long
curl https://apps.nrs.gov.bc.ca/gwells/api/v1/wells?sw_lat=51.599253&sw_long=-128.220340&ne_lat=52.108192&ne_long=-127.150547
Aquifers:
# all aquifers
curl https://apps.nrs.gov.bc.ca/gwells/api/v1/aquifers
See the Swagger documentation for more examples.
GWELLS Api is used by DataBC to export model information and provide well information on the DataBC website.
The command located in app/backend/gwells/management/commands/export_databc.py is used to pull data from the gwells database and export it for consumption in DataBC.
If a new field(s) are needed for export, this export_databc.py has raw sql queries that need to be updated with the required fields. Descriptions are also needed for the swagger documentation that are pulled from the db_comment fields in the django models. Both the raw sql queries and the db_comment field on the model need to be updated for new fields.
Clone the GWELLS repository and run the application with Docker:
cd gwells
docker-compose up
Visit the following links to browse the API and frontend applications:
- Django REST API development server: http://localhost:8000/gwells/api/
- Vue frontend development server: http://localhost:8080/
Some GWELLS pages (submitting new well reports, adding or editing aquifers, or adding or editing qualified well drillers to the registry) require authentication. Authentication uses the Province's Single Sign-On system. A GWELLS team member can request access for collaborators if needed.
Django unit tests:
cd app/backend
docker-compose exec backend python manage.py test
Vue unit tests:
cd app/frontend
docker-compose exec frontend npm run test:unit
Postman API tests:
Import the json test collections in the api-tests/
folder into Postman.
Single shapefile, with aquifer ID specified in CLI.
docker-compose exec backend python manage.py import_shapefile 2 aquifers/fixtures/shp/shapefile.zip
Bulk import, requires the AQ_NUMBER attribute on each polygon. Requires a folder with shapefiles to be prepared and passed in (zipped or not). Note: if DEBUG=True, all geometries will be uploaded to a random aquifer instead of the one matching its' number, so we can test locally with a development database.
mkdir app/backend/bulk
mv DATABC_EXPORT_FILE.zip app/backend/bulk/
docker-compose exec backend python manage.py import_bulk_shapefile bulk
To download new licence data from DataBC and merge it into your DB, do
docker-compose exec backend python manage.py import_licences
Pull requests made from branches in the bcgov/gwells repo will kick off a dev pipeline, which runs tests and deploys a preview environment containing your changes. The pull request page will update when the environment is deployed and mandatory checks (unit tests, etc.) pass or fail.
GWELLS uses PostgreSQL (with PostGIS), Django REST Framework, and Vue.js. We also use both AWS S3 and a self-hosted Minio service for storing documents.
Our production and staging environments run on an OpenShift container platform cluster. OpenShift templates for services are located in the openshift/
folder, along with more information about dev and staging environments on our cluster.
Government employees, the public and members of the private sector are encouraged to contribute. Please read and follow our Code of Conduct.
All contributors retain original copyright, but are granting a world-wide, royalty-free, perpetual, irrevocable, non-exclusive, transferable license to all users. This project is covered by an Apache v2.0 license.
Many of our features have been developed by members of the community. Check the BC Developer Exchange website, where paid opportunities to build features for GWELLS and other applications are posted.
Issues are tracked on the Water Jira board. Note this is a restricted issue tracker and you will require permission to be able to view it.
Code released under the Apache License, Version 2.0.