UAV-LiDAR, field inventory and coastline data for two Avicennia germinans chronosequences, Coronie, Suriname (2024)

Abstract

Field, wood, soil, UAV laser scanning (ULS) and coastline data, with the analysis code, from two coast-perpendicular Avicennia germinans chronosequences on the mudbank coast of Coronie, Suriname. The data underlie: Feyen J. et al. Structural development of Avicennia mangrove forests along a successional chronosequence on the Suriname coast revealed by UAV laser scanning and field inventory (in preparation). Study design35 circular plots (12.5 m radius) were laid out at 100 m intervals from the seaward mangrove edge along two transects: T1 near Perseverance (20 plots, crossing a chenier ridge) and T2 near Jenny (15 plots). Stand ages range from 5 to ≥ 36 years, derived from annual Landsat classifications (1988–2024). Fieldwork and UAV flights took place in September 2024. Contents Coronie2024_code.zip data/ Plot data (Coronie2024_field_plot_data.csv): stand age, seaward distance, chenier position, stem density, basal area, diameters, heights, species composition, aboveground biomass with Monte Carlo uncertainty (Chave et al. 2014; BIOMASS), basic wood density and wood anatomy from increment cores, soil organic carbon to 30 cm and 1 m, soil EC and pH per depth, and porewater salinity. Tree data (Coronie2024_tree_data.csv): 1321 stems with species, DBH, measured and modelled height, decay class and assigned wood density. ULS metrics (Coronie2024_ULS_MC_iterations.csv, Coronie2024_ULS_plot_data.csv): over 100 structural metrics (height, vertical heterogeneity and complexity, plant area, canopy cover, gaps, canopy volume) for 100 Monte Carlo realisations of each plot centre, which propagate the ~5 m GPS uncertainty, plus plot medians and SDs. Shorelines (Coronie2024_shorelines.gpkg): the annual seaward mangrove edge, 1988–2025, from Landsat imagery. Spatial layers (Coronie2024_spatial.gpkg, Coronie2024_plot_centres.csv): centres of all 35 plots, 50 m clip footprints, and a tile index showing which point-cloud tile covers each plot. Data dictionary (data_dictionary.csv): column definitions and units for all tables; Data_S1_field.xlsx and Data_S2_ULS.xlsx hold the same plot tables formatted as journal supplements. R/, python/: scripts that reproduce every statistic, table and figure of the article from the tables in data/. lidar/: scripts that clip the point clouds around the plots and extract the ULS metrics. README.md (methods, conventions, instructions), CITATION.cff, licences. Coronie2024_transect_T1.zip, Coronie2024_transect_T2.zip Point clouds: height-normalised point clouds of both full transects, homogenised to 750 pulses m⁻²; the ULS metrics were computed from plots clipped from these clouds. Acquired with a DJI Matrice 300 RTK carrying a DJI Zenmuse L2 (47 m altitude, 240 kHz, ~3300 points m⁻² before thinning), RTK-corrected. Licences and fundingData and point clouds: CC BY 4.0. Code: MIT. Funded by the Global Climate Change Alliance (GCCA+ 2) project “Strengthening the Mangrove Forest Monitoring System” (SUR/2021-019) and the PRSUMF project (SUR-2025-034). Field and logistical support: Foundation for Forest Management and Production Control (SBB) and the Centre for Agricultural Research in Suriname (CELOS).