Background on the MIT Global Change model (EPPA)

• It's written in general algebraic modeling software (GAMS)
• The foundational data is in GTAP (paid data, so they have to make the public model lag behind the most recent data release)
• It takes ~20 min to run
• We'll precompute a set of standardized scenarios, then plot that in the data-viz, and allow people to select the variables they want to plot
• Whenever we want to run the model on a new case, we'd want to precompute more results and add that to the visualization

Use cases

1. Public outreach and communications tool (wouldn't upload everything)
   • Preloaded with a couple scenarios
   • Allow people to upload their own scenarios
2. Help researchers explore results of test runs (sometimes 15 different scenarios a day)
   • Currently they use excel, but it could be faster to use this tool

• energy by type, by region
• electricity use by type, by region

Choosing the best approach

Use the gcam dashboard

GCAM dashboard is a climate data visualizer written in R. I'd adapt the output of EPPA to fit format expected by GCAM. I've forked GCAM dashboard to start changing it to do what we need.

• drawback: not the best option for graphical interface. slow, and lacks customizability. not as easy for other frontend web developers to modify.

Build it using a javascript data-viz library

We could use a library such as d3.js

• drawback: building more from scratch

Projects

• Climate risk map testing https://github.com/cypressf/climate-map-tests
• Climate risk map website https://github.com/cypressf/climate-risk-map
• EPPA dashboard https://github.com/JGCRI/GCAM-dashboard