Sudden Aspen Decline is microsite-driven, so a blanket species flag is wrong. Each aspen and cottonwood here gets its own call by fusing two independent lines of evidence.
How we built this. Trees were mapped from satellite with no inventory (NAIP 0.3 m aerial imagery + a tree-crown detector). Each tree's 2050 outlook combines a two-sided climate envelope (LOCA2, a 4-model ensemble for the Steamboat grid cell, +4 °C) with, for aspen & cottonwood, a per-tree drought-vulnerability fusion: a terrain model (heat-load, elevation, distance to the Yampa from a 30 m DEM) × an observed satellite signal (Sentinel-2 2018–2024 late-summer NDVI trend — the published fingerprint of aspen dieback). Replace = exposed microsite and observed decline; Monitor = one signal only; Keep = cool/moist refugium holding steady. The satellite axis is 10 m, so it is strongest for stands and weaker for lone street trees (flagged per tree). Dollar figures are conservative estimates. Built by Raiho Consulting.
Imagery: USDA NAIP, ESA Sentinel-2, Copernicus DEM (all open). Climate: LOCA2 (UCSD). Basemap: CARTO / OpenStreetMap.
Get a forecast for your town →