High resolution UAS imagery derived metrics for conifer seedling detection in a postfire Southern Rockies landscapes

This dataset contains the spatial data products and derived variables used to support the analyses presented in the associated manuscript titled "Conifer Seedling Mapping with UAS Data and Supervised Learning in Postfire Southern Rockies Landscapes" (in review). The dataset includes high resolution raster layers generated from drone imagery using the Structure-from-Motion workflow implemented in Agisoft Metashape (base_raster_layers.7z). These products include orthomosaics and digital surface and terrain models and canopy height models. From the Metashape outputs, multiple vegetation indices and texture metrics were calculated to characterize vegetation structure and spatial heterogeneity. These raster layers provide quantitative representations of vegetation condition and spatial patterns across the study area and can be found at all_layers.zip folder. The dataset also includes vector polygon layers representing mapped features within the study sites. For each polygon, summary statistics of vegetation indices, texture measures, and other raster derived variables were extracted to support subsequent statistical and spatial analyses. These polygon level attributes form the primary analytical dataset used in the manuscript and can be found at var_polygon_stats.7z folder. In addition, the model outputs and final model results can be found at model_outputs.7z and full_data_table_with_rf_predictions.csv files. All files are provided with accompanying data sources, processing steps, variable definitions, and coordinate reference systems. The r codes used to make these data and analsytics used in the manuscript are at Github Repository (https://github.com/Chathu84/Conifer_seedling_mapping.git). The dataset enables transparency, reproducibility, and reuse of the spatial analyses presented in the study and may support future research on high-resolution remote sensing applications for vegetation and ecosystem monitoring. Any users are required to acknowledge the data source and works using the DOI and the co-authors of this data repository.