Policy Bear

Our contribution to Hack the Arctic, a hackathon to design solutions for environmental challenges.

The Policy Bear is up and running at www.policybear.gregl.it.

• We developed a prototype for a web interface to generate short and relatable sentences from climate data sets to use them in the next conference or meeting (that starts in only 10 minutes).
• The web interface enables policy-makers to create nice looking text graphics with a short sentence about the data and a comparison to a day to day measurement.

Policy-makers are able to ...

• compare concentrations of greenhouse gases (carbon dioxide and methane) between different dates and observation stations within the ICOS Atmosphere network (Integrated Carbon Observation System).
• select a timeframe with either annual or monthly steps (range depends on the dataset availability)
• compare absolute or relative changes
• compare change with day-to-day measurements (we included cows and cars)
• export results as pdf or jpeg in three different beautiful themes
• share results with colleges and friends

Technical details

Backend

• The data is processed in a flask-powered Python backend that we hosted on a Heroku server.
• The data is loaded directly via a Python Api of the ICOS data portal.
• The flask server can process API-calls to access specific values from the chosen dataset.
• Our python-based backend makes it easy to add all of your favourite datasets in the future.

Frontend

• The frontend is a Vue.js web app with Bootstrap styling hosted on a netlify server.
• All relevant information about available datasets, timeframes and other options are retrieved vie API-calls so that no editing in the frontend is needed to add new data!
• The API-calls are made dynamically so that changes to the generated sentences can be seen instantly!

Data

• We processed the data by taking means over all datapoints within the target period.
• Although data is available at different heights our algorithm is currently still limited to the measurement at lowest height.
• The day to day measurements comparisons are very rough estimates on the basis of general values so don't expect the most meaningful values.

Further Development

• Select available ICOS stations via map viewer.
• Adding more and improving the currently available parameters to compare to.
• Including gridded datasets and more parameters (such as sea ice cover)
• Better algorithm to make comparisons more reliable.
• Improving interface and user experience of our website.