Dataset for: <b>Forest structure types in the southern Amazon Basin are influenced by anthropogenic edges and physiography</b>

<p dir="ltr">This repository contain all the dataset and extended results for the <b>Forest structure types in the southern Amazon Basin are influenced by anthropogenic edges and physiography</b>. </p><p dir="ltr">All the main analysis reported in the paper plus extended result (here collected as .zip) can be reproduced following the script sequence available at github.com/monviso/XinguBasin. Necessary file to run the scripts are the five 2019-2023 .parquet files, the Xingu_GAM_prepared.rds and the ALS.zip file (only useful for the ALS characterization)</p><p dir="ltr">The GEDI_RFLD_sampling_frame_YYYY.parquet are the original GEDI V3 shoot over the Xingu basing, already quality filtered according to the filtering criteria reported in the paper. </p><p dir="ltr">Xingu_GAM_prepared.rds is the file necessary to fit all the GAM model; it include all the predictors set.</p><p dir="ltr">GAM_outupus.zip include all the extended GAM results, including: 70 models, models statistics and partil effect plots</p><p dir="ltr">step_01c_extended_robustness.zip report all the extnded result of the NMF and k-means robustness tests perfomed to check forest architecture classes clustering stabilty. </p><p dir="ltr">step_03_radiometric_analysis_v3.zip reports all the radiometric characterization results</p><p dir="ltr">step_08_ALS_characterization.zip reports the GEDI-based architecture classes characterization based on high-resolution ALS data.</p><p dir="ltr">ALS.zip is a large collection of laz files that spatially overlaps with retained GEDI footprints in the Xingu basin, used to obtain the ALS characterization. </p><p dir="ltr"><br></p><p dir="ltr"><br></p>