For 17 years, Canada's RADARSAT-1 earth observation satellite obtained radar images of the Earth's surface. Many of these images remain unprocessed and there is no easy way for interested users to examine the location and availability of currently unprocessed images.
During the Space Apps Ottawa 2019 hackathon, a team of four volunteers tackled a Canadian Space Agency/Natural Resource Canada challenge. The goal was to automate and improve searches on the Government of Canada EODMS system and then position RADARSAT-1 image locations and metadata on a world map, to allow users to easily view and filter against the whole Earth Observation data set.
Project Goals & Objectives
The goal of CSA and NRCan (Natural Resources Canada) is to make more RADARSAT-1 data available. Currently only 2% is processed and open to the public. To reduce high processing costs, the CSA would like to identify and prioritize datasets that could be useful for research purposes.
Because this project was a direct response to a hackathon challenge they posed, it has a single SMART objective: solve the challenge.
- Specific: "Realtor.ca, for RADARSAT-1". Create a script to automate searches in the EODMS database and display the results on a map.
- Measurable: A website must allow users to perform EODMS searches with filters, and display real data results and EO imagery metadata on a world map.
- Achievable: Work is limited to proof of concept including some hard-coded and manual steps to demonstrate "art of the possible".
- Rewarding: Project will demonstrate significantly improved and intuitive EODMS search experience and solve a real challenge posed by CSA and NRCan.
- Time-bound: POC must be completed within 48-hours of project start.
Over a 48-hour period, our volunteer team (from several different companies, including Deploy) created a working website composed of:
- Front end search website using Vue.js and MapBox, served by node.js
- GeoJSON REST API middleware using C#
- Geospatial Metadata database using Postgres and PostGIS.
- GeoJSON extract and intake process from the EODMS system to our metadata database
The resulting proof of concept was well received by CSA and NRCan, subsequently winning the challenge category in Ottawa and then nationally.
We can provide a demo of the working system, upon request. The code is open source under Apache 2 and also available upon request.
One of our team-members Kevin Jones wrote a great blog about the development effort, "Hacking to Liberate RADARSAT-1 Data". It also provides good insight into how hackathons work and why they are a beneficial way to reduce software development risk, cost, and duration.