Tracking Land Conversion with the Eye in the Sky
Deforestation is a classic example of land conversion. The world’s forests have in the past absorbed as much as 30% of annual global anthropogenic CO2 emissions, a similar amount as the oceans, so we can say land conversion harms human ability to fight climate change. But how do we track this issue to know what we should do about it?
Unfortunately, there is only one year’s worth of land conversion data in the U.S. available to the public. That is a huge problem if you want to ask, for example, ‘Is the ethanol industry driving land conversion, making it a less viable source of renewable energy?’ or ‘Are laws around the use of natural lands being broken?’
Thankfully, there is an alternative: to create that public record using machine learning techniques and satellite images.
Byron is a data consultant, but earned his ticket to PyCon when he taught himself Python (his first language) just over a year ago and graduate from General Assembly’s Data Science Immersive bootcamp. He is passionate about image recognition and remote sensing as a path to uncovering insights about the world, and will have been using machine learning and satellite images to track changes in agriculture and land conversion.