To get a more precise measure of local greenery for some studies, scientists use Google Street View data and something called the normalized difference vegetation index, which uses satellite imagery to quantify plant density and health in an area of land. A company called Nature-Quant based in Bend, Ore., recently used machine learning to develop NatureScore, which combines multiple datasets on parks, tree canopies, and air, noise and light pollution to develop a score between 0 and 100 as a proxy for greenness for every address in the U.S. (a heavily urban environment would generally score below 30 and a forest above 70).

Read the full article here