Dataset: "Soils and topography control natural disturbance rates and thereby forest structure in a lowland tropical landscape"

This dataset contains raw data, processed data, and code associated with the manuscript "Soils and topography control natural disturbance rates and thereby forest structure in a lowland tropical landscape". This project quantifies canopy disturbances between 2015 and 2020 across Barro Colorado Island, Panama, using drone photogrammetry.<br><b>Raw image/point cloud data</b>- Raw drone images are in folders "DroneImages_YEAR". Agisoft Metashape project files referencing files as organized in these folders are also included in these folders.- Raw image orthomosaics are in folder "DroneOrthomosaics"<br>- Raw (unaligned) point clouds are in the folder "PointClouds/Raw"<br><b>Processed point cloud data</b>- Processed (aligned, tiled) point clouds from lidar (2009) and photogrammetry (2015, 2018, 2020) are in the folder "PointClouds/Processed"<br><b>Analysis data</b>All data files directly used in analyses are included in folders starting with "Data_".<br>- Data_Ancillary: shapefiles for soils (BCI_Soils), forest age (Enders_Forest_Age_1935), streams (StreamShapefile), 50 ha plot outline (BCI50ha, and island outline minus 25 m buffer (BCI_Outline_Minus25); information for blocks used for bootstrapping size frequency distributions (bootstrapBlocks.csv) and for aligning data in CloudCompare (gridInfo.csv)<br>- Data_GapShapefiles: shapefiles for canopy disturbance in each period created in Code_ProcessHeightData/DefineGaps.R<br>- Data_HeightRasters: height rasters produced in Code_ProcessHeightData/MakeDSMs.R and Code_ProcessHeightData/DefineGaps.R. Also includes previously created digital elevation model from 2009 lidar (LidarDEM_BCI.tif) and a digital surface model from higher-res 50 ha plot data (DSM_50haPlot_20150629_geo).<br>-Data_INLA: input data for INLA models created in Code_INLA/setupINLA.R (INLA_40m.RData) and .RData objects with results from all INLA models (described in Code_INLA/setupINLA.R). Results from sensitivity analysis for best curvature/slope smoothing scale (INLA_SmoothingScaleResults.csv, also output from Code_INLA/setupINLA.R).<br>- Data_QAQC: All data used to create cloud and QAQC masks (cloud raster products output from ArcGISPro, other rasters from Code_ProcessHeightData/MakeDSMs.R) in file Code_QAQC/AnnualQualityMasks, and resulting mask rasters. Also includes manually annotated gap shapefiles from the 50 ha plot for 2015-2018 from this project (QAQC_IslandData) and from the higher-res monthly data for the 50 ha plot (QAQC_50haData)<br>- Data_TopographyRasters: lidar DEM smoothed at different scales (output from Code_ProcessHeightData/SmoothDEMs.R), resulting curvature and slope smoothed rasters (output from ArcGISPro), and height above nearest drainage raster (distAboveStream_1000.tif, output from ArcGISPro).<br><b>Code</b><br>All code are in zipped copy of the GitHub repository https://github.com/kccushman/BCI_Photogrammetry, saved at the time of publication (Code_GitHubRepository.zip). Code scripts reference all data files from the folders in which they are organized here.<br>- Code_AlignDroneData: R script for tiling raw point cloud data, and .bat files for aligning point cloud tiles in CloudCompare's command line tools.<br>- Code_GapSizeFrequency: R scripts for fitting and plotting gap size frequency data from gap rasters and shapefiles.<br>- Code_INLA: R scripts for configuring data for INLA models, running INLA models, and analyzing INLA results.<br>- Code_MakeFigures: R scripts for making main and supplemental figures.<br>- Code_ProcessHeightData: R scripts for making canopy height rasters from point cloud data, defining gap rasters/polygons from canopy height data, and smoothing digital elevation model (DEM) topography data from 2009 lidar.<br>- Code_QAQC: R scripts to make data quality masks based on cloud and photogrammetric reconstruction quality, and to find the optimal height correction for 2015 data with lower image overlap.<br>