- Roberto Vera
- Leo Muñoz
- Salvador Del Cos
- Ángeles Cruz
We analyzed the droughts levels across the US and its possible correlation with wildfires throughout the years 2000 to 2016.
- We selected two data sources, one from Kaggle y the other one from MTBS (Monitoring Trends in Burn Severity).
- We cleaned up the datasets, drop columns we weren't going to use and null values too.
- The clean up process for both datasets turned into .py functions to avoid huge jupyter notebooks.
- In the exploration process, we generated different visualization, such as linear and bar plots.
- We also used a library called Basemap, to draw the US map and avoid using the API from Google.
- Based on the droughts levels we generated, throught the year 2000 to 2016, the highest average drought level ocurred in 2012, with a level of 1.4.
- 2.4 has been the highest average drought level, the top 5 counties are located in the state of Nevada.
- Out of the 6 drought levels, the number 5 (which is the exceptional drought) didn't decrease drom 2011 to 2016, as it did in previous years.
- More than 8K wildires had taken place in the US in the same years.
- 2011 was the year with most wildfires, 846.
- 2005 was the year with the most sq. miles burned. However, this year wasnt the year with most wildfires. So, the wildfires quantity doesn't mean more sq. miles are burned.
- In our animated maps, highest drought level locations seem to be located as the wildfires.
-
Droughts & Counties: https://github.com/veraroberto/Project1_Droughts_USA/blob/main/data_extraction.ipynb
-
Fires: https://github.com/veraroberto/Project1_Droughts_USA/blob/main/wildfires_dataframe.ipynb
-
Droughts Intensity: https://github.com/veraroberto/Project1_Droughts_USA/blob/main/Drought_Intensity.ipynb
-
Droughts Location: https://github.com/veraroberto/Project1_Droughts_USA/blob/main/Drought_Locations.ipynb
-
Fires: https://github.com/veraroberto/Project1_Droughts_USA/blob/main/fires_data_maping.ipynb