Images from Google Earth Engine Classification Algorithm Development for Riparian Zones
On this page:
When considering a classification of riparian areas, high resolution (1 meter) offers significantly more in the way of detail, adding 900 one-meter pixels for every pixel of a 30-meter product (e.g., National Land Cover Dataset). As cloud-based computing continues to be mainstreamed into remote sensing, platforms like Google Earth Engine showcase powerful parallel processing abilities and on-the-fly data visualizations that lend themselves well to classifying large geographic areas quickly, but with steep learning curves and challenges that can make getting a refined product difficult. This project, still incomplete, sets out to classify functional and fixed-width riparian zones using Google Earth Engine and to test the capabilities of Google Earth Engine for developing high resolution land cover classifications. Results from this project are compared to products developed for EPA’s EnviroAtlas ecosystem service web tool.