An EverImpact’s customer, the city of Dijon is willing to estimate the amount of carbon directly emitted over its territory by urban activities. We have equipped the city with a spatially distributed network of IoT sensors that measure in real time the CO2 concentration in the atmosphere close to ground level (3-5 meters above ground surface).
The assessment consists on:
- Clean and organize the dataset;
- Plot statistics (mean, min, max, etc...) for each sensors on a map;
- Interpolate/extrapolate the discrete measurement data to have a heatmap of the CO2 concentration over the Dijon Metropole;
- Play with the data and impress us on what you would do with such amazing ressources.
Explain how you would convert the CO2 concentration recordings into flux (in tCO2 by unit of surface by unit of time) of CO2 emitted by the road network of the city.
You can reach the original notebook on kaggle here.
You can reach the training runs on wandb here.
You can download my pdf presentation here.