Operational, Timely Fire Data on an Open Source Stack.
Presentation | Presented
- Angus Carr, Saskatchewan Environment
In 2007, Saskatchewan Environment's Fire Management and Forest Protection Branch built a system for delivering timely operational fire information for forest fire suppression. The data input requirements are fairly simple: we need lightning, weather, and operations data to be maintained in a database. We use the database to provide our business logic- calculating fire behaviour indices and summary data, emailing warning data, and other requirements.
We used an open source stack of software for sustainability, and flexibility. Data comes in to PostgreSQL/PostGIS from ODBC input programs, XML downloads, and from direct client programs. It is processed to pick out relevant facts, and then notifications are sent out as needed. We use GRASS to produce our raster interpolations of weather between weather stations.
We can view the data with tables and maps using Apache and PHP to drive mapserver to produce mapped output in WMS and WFS forms. The WMS and WFS are used by OpenLayers to produce our map products. We can also feed the data to other GIS systems, and provincial data centres in other provinces.
We improved service quality dramatically. We have reduced downtime from a regular event to practically nil (once a service is debugged). We have added audit trails to all of our data streams, which will help dramatically once the season is over and we debrief significant incidents. We can now interoperate with other jurisdictions throughout Canada with considerable ease.
Our development costs have been staff time of two people over the winter.