RSEI in motion, and a few tips on Torque

'
 

The US Environmental Protection Agency (EPA) has an incredible amount of data available to people  and the Risk-Screening Environmental Indicators (RSEI) microdata is a whopper of a dataset. RSEI is an air model which organizes results from environmantal data collected. Specifically  it tallies the quantities of release and transfer of chemicals  and is used to rate the air you breathe. The EPA is re-releasing this data by way of Amazon’s public data portal for anyone to use and visualize.

We took a look at the data and decided our Torque engine would be a great way to show how the readings have changed over time. With 27 years of data for 400 listed chemicals over 45 000 facilities  there’s a lot to show.
We narrowed it down to a small area (SF Bay Area) and Joined the RSEI 1km grid to the

What we have done exactly here is assign the ‘toxconc’ value to each grid cell for the 22 years of data we are using and merged that with the year of data  then visualized them  colored over a range. Since our values went from .000004 to 13 million  we used a custom torque aggregation function that mapped the values across a ramp:

-torque-aggregation-function:"avg((log(toxconc)+5.36)*100.0/12.49)";

‍
The parameters for this function are:

avg((log(toxconc)+ (-min_toxconc) )*100.0/(max_toxconc-min_toxconc)<br>

The log transforms our values from .000004 to 13 million to -5.36 to 7.13. Then we simply shift the origin and scale so each value is between 0 and 100.

This gave us a balanced look at the data in the grid and made it possible to understand the nuances. Using chroma.js  we developed a color ramp based on a perceptual color space that assigned a color to each value to give it a continuous tone.

Using the power of CartoDB  we’ve brought the history of these 748K points into and can see the story of the air we breathe much more clearly.

Happy mapping (and breathing!)