Analyzing Satellite-Derived 3D Building Inventories and Quantifying Urban Growth towards Active Faults: A Case Study of Bishkek, Kyrgyzstan

#############################################################################################################
Datasets supporting the publication:
Analyzing satellite-derived 3D building inventories and quantifying urban growth towards active faults:
a case study of Bishkek, Kyrgyzstan.
https://doi.org/10.3390/rs14225790

-Please refer to the publication for details on the production of each dataset.
-Datasets are ordered following the publication figures.
-Please cite the publication and this dataset repository when using the data.
#############################################################################################################

------------------------
Structure:
File ID
-[fields:] description
------------------------

KH9_1979_builtup.shp
-KH9 1979 built-up area classification

S2_2021_builtup.tif
-Sentinel-2 2021 built-up area classification.

S2_2021_corine_land_cover_class.tif
-Sentinel-2 2021 land cover classification in Corine 2018 land-cover classes.

S2_KH9_DN_change_aggregated.shp
-Proportional DN change aggregated to a 1 km^2 grid for areas ≥50% built-up.

building_characteristics.shp
-build_count: building count in 500 m square grid cell.
-mean_area: mean building size (m^2) in  500 m square grid cell.
-median_area: median building size(m^2) in  500 m square grid cell.
-cell_coverage: %building coverage of 500 m square grid cell.

pleiades_buildings_all.shp
-All building detections from Pleiades data. Confidence values are output from the deep learning model.

pleiades_buildings_heights.shp
-Building detections from the Pleiades data that were allocated heights (m).
-Zmean, Zmedian,... refer to heights (m)

wv2_buildings_all.shp
-All building detections from WorldView-2 data. Confidence values are output from the deep learning model.

wv2_buildings_heights.shp
-Building detections from the WorldView-2 data that were allocated heights (m).
-Zmean, Zmedian,... refer to heights (m)

trained_rcnn.zip
-ArcGIS Pro deep learning model (DLPK) used to extract building footprints.