Earthlab Career Data

The dataset consists of high-resolution unmanned aerial system (UAS)-derived imagery and derived products collected across post-fire landscapes in the Southern Rocky Mountains, USA. Data were acquired between 2023 and 2025 for 34+ transects across 15+ wildfire events that burned between [1994 - 2001]. These data were collected in ponderosa pine (Pinus ponderosa) and Douglas-fir (Pseudotsuga menziesii) dominated forests in the Colorado Front Range. The dataset is designed to quantify factors that influence post-fire vegetation regeneration, with a focus on conifer seedling presence and abundance. UAS imagery was collected using RGB sensors (DJI Phantom 4 Pro and DJI Mavic 3 platforms) over transects and targeted sampling areas within fire perimeters. Imagery was processed using Structure-from-Motion photogrammetry (Agisoft Metashape V 1.8) to generate orthomosaics and dense point clouds. From these products, we derived high-resolution (≈6 cm) digital elevation models (DEMs), Canopy Height Models (CHMs), Tree tops and canopy polygons, and vegetation class for each polygon ( seedling, mature conifer tree, deciduous tree, Shrub, or standing dead tree) in postfire landscapes at the time at data collections (Ilangakoon et al., 2026). Vegetation class was mapped using a supervised classification workflow trained on field-validated plots https://doi.org/10.5281/zenodo.18929294. These products are integrated with fire severity, topographic derivatives (e.g., slope, aspect, topographic position index), and climate covariates to support high resolution modeling of post-fire regeneration processes. To access the full dataset please use the SpatioTemporal Asset Catalogs (STAC) using "https://browser.moregeo.it/external/data.cyverse.org/dav-anon/iplant/projects/earthlab/nfs_career/stac_catalog/catalog.json" Any questions regarding the dataset, dta processing workflow or any issues with data, please reach out to Dr. Ginikanda Ilangakoon via [e-mail].