Geospatial Case Files #01: Northeast Smoke Week (Free Practice Dataset Kit)
Free Practice Dataset Kit
Free
A free, real-world geospatial dataset for practice, not toy data.
In early June 2023, smoke from Canadian wildfires drifted over the northeastern United States and turned the sky orange over New York. This kit lets you explore that event with real data, joined into one table. This is Kit 01 of Geospatial Case Files.
What's inside
1,805 rows, one per air-quality station per day (153 stations, 11 US states, 1 to 14 June 2023)
Air quality (EPA), satellite fire detections (NASA FIRMS) and weather (NASA POWER) in one table
Fire features within 500, 1,000 and 1,500 km, plus fire counts from 1 to 5 days earlier
CSV, GeoPackage, GeoJSON and Parquet. GeoPackage and GeoJSON open directly in QGIS and ArcGIS Pro.
A guided Jupyter notebook, with a PDF copy
A data card (every column, how it was built, known limits) and a sources file
What you can practice
Mapping a spatial event over time
Distance and time-lag features
Testing an idea that fails, then working out why
Regression with random vs time-based validation, and why random splits can mislead
Opening the same data in Python, R or GIS software
Who it's for
Learners and beginners building a geospatial data science portfolio. Basic Python helps for the notebook. If you use GIS software, the GeoPackage works without coding.
Honest limits
It covers 14 days and one event. Station coverage is uneven (105 of 153 stations have all 14 days), wind direction is not included, and the weather is on a coarse grid. Fire counts are satellite detections, not fires. This is a learning dataset: validate it against the original sources before using it for research, and don't use it for health or safety decisions.
Credits and terms
Data from the U.S. EPA Air Quality System, NASA FIRMS and NASA POWER. This product is not affiliated with or endorsed by NASA or the EPA. Personal and educational use, no reselling the data. See LICENSE.txt in the download.
Which event should Kit 02 cover? Message me on Instagram: @geospatial.bff