We show that the synergistic use of Sentinel-1 and -2 gives superior classification results, allowing for the identification of residential infrastructure vs. other infrastructure types, and offers improvements spatially, temporally, and thematically over existing LULC datasets.

We compared our product with existing land use and land cov3r (LULC) datasets: National Land Cover Dataset (NLCD), Microsoft Building Footprints, and the Dynamic World dataset.

Using Google Earth Engine, we applied our Random Forest algorithm to create annual maps of built infrastructure from 2015 to 2024, delivered at 10 m resolution, for across the Snake River Plain, in southern Idaho.

I am excited to share our new paper on mapping built infrastructure using remote sensing data integration and open-source cloud computing.

doi.org/10.1016/j.rs...

After graduating, unsure of my next steps, I chose to go on an adventure and spent six months thruhiking on the Appalachian Trail. On this trail, I experienced two life changing events: meeting the man who would become my husband and my first #wildfire.

A decage ago, I started my academic journey at @rhulgeography.bsky.social. At the same time I was completing my Gold DofE (@dofeuk.bsky.social), which took me to the Amazon Rainforest, Peru. There I volunteered with the Crees Foundation (www.crees-foundation.org) monitoring macaws and parrots.