Inspiration
Research has shown that short- and long-term exposure to air pollution can lead to a wide range of diseases including stroke, chronic obstructive pulmonary disease, trachea, bronchus, and lung cancers, and lower respiratory infections. Accurate forecasting of air quality helps people to plan ahead, decreasing the effects on health and the associated costs. If people are aware of the variations in the quality of the air they breathe, the effects of pollutants on health as well as concentrations likely to cause adverse effects, they can take action in advance to reduce exposure. The awareness also has the potential to create a cleaner environment and a healthier population.
What it does
The models in this project are able to predict the quality of air in Bakersfield, Phoenix, Fresno, and Visalia areas based on the National Ambient Air Quality Standards (40 CFR part 50) for six principal pollutants ("criteria" air pollutants). Specifically, the thresholds for particulate matter (2.5 and 10 microns) and ozone were considered for the regions listed above.
How we built it
We trained classification models that can make air quality predictions in a given scenario based on weather features such as windspeed, precipitation, dew point, maximum and minimum temperatures, and the month of the year.
Challenges we ran into
No challenges
Accomplishments that we're proud of
The ability to access and leverage both the NOOA GSOD WEATHER DATA and the OPENAQ data for this project
What we learned
The most important weather factor or feature affecting air quality varies with the scenario. In some scenarios, it is Maximum temperature while in others, dewpoint or precipitation has the greatest influence.
What's next for Air Quality
Deployment and starting forecasting as a service.
Log in or sign up for Devpost to join the conversation.