Inspiration
Four years ago, there was a huge fire just miles from where we live. The smoke from the fire made the air quality reach its worst point in decades. The sun was red, sky was gray, and breathing got difficult. As someone with asthma, I was having a difficult time breathing due to the smoke and had to go on a ventilator. Because of that, I wanted to make a website that helped raise sustainability awareness and preparation for what could occur, including other natural disasters like droughts.
What it does
With data analysis and predictive modeling, we highlight and inform about possible droughts, and wildfires. We have put our programs onto the NatureAlert website.
1) Predict Total Acres to be Damaged In 2025 We take fire burning statistics from 2015 to 2020 and have calculated a prediction of the total acres of fire burning in 2025 using a linear regression model trained.
2) Fire Hotspot Map We have a map that identifies hotspots with higher fire vulnerability, allowing authorities or relevant stakeholders to prioritize fire prevention and management efforts in those locations. These hotspots have been identified from historical data from 2015 to 2017
3) California Drought Area Graph We have built a bar graph that represents the area in drought in California over different weeks since January 2017 to 2022. This interactive graph is embedded in an HTML file and includes tooltips and click events for enhanced interactivity. This program allows for future development of strategies for sustainable water resource management
4) Population Affected by Droughts in Specific Months Line Plot The line plot shows the total population from 2017 to 2023 affected by droughts in California for each month. This program allows awareness of when water conservation effort should be applied based on the graph spikes.
How we built it
Using HTML, CSS, and Javascript, we made the website interface, and using python flask, we projected it locally and included python in the backend. We used pandas, geopandas, matplotlib, and linear regression modeling, we trained the models from data in csv files on different topics.
Challenges we ran into
We were unable to add the interactive maps and plots to the website and had to resort to saving them as PNG files. We hope to make it more interactive and not PNGs. We also wanted to add more statistics such as the increases in average water bills for Californians so it has more depth.
Accomplishments that we're proud of
We are proud of creating interactive graphs (some are PNGs but run in the backend can be interactive), especially the fires burned map which has different sizes and colors for the fires burned in California based on severity. We are also proud of our prediction for the area that will be burned in 2025 based on past data using linear regression models.
What we learned
We learned how to create plots using matplotlib and how to parse through data to make graphs and statistics out of it. We also learned how to show different data in different ways such as bar graphs, scatterplots, and maps.
What's next for Nature Alert
In the future, we'd like to add more interactivity with our statistics and ways to prepare and prevent fires and droughts, as well as expand to other states outside California and include more disasters that occur such as hurricanes.
Log in or sign up for Devpost to join the conversation.