Description (V2 - Latest): To enable easy integration in the workflows, we have provided the main datasets in the following formats: Vector dataset: Folder - Vector The global gross estimated rooftop area per FN grid cell for each SSP narrative is provided as a Geopackage (.gpkg) file (Results_Vis.gpkg) with polygon geometries at 1/8-degree spatial resolution in an EPSG:4326 coordinate system. The attribute table of this file contains FN_ID column representing the FN grid cell ID, and other columns representing the FN_ID specific assessed rooftop area. The assessed gross rooftop area columns are sequenced as BF_X_Y with X having values as 1, 2, 3, 4, and 5 for SSP1, SSP2, SSP3, SSP4, SSP5 narratives with Y representing the assessment year having values as 20, 30, 40, and 50 for years 2020, 2030, 2040, and 2050 and with km2 units. In addition, a CF column is added for each FN_ID entry that documents the Capacity Factor for rooftop solar PV based on the World Bank solar atlas. Raster datasets: Folder - Raster The global gross estimated rooftop area per FN grid cell for each SSP narrative is provided as a geotiff (.tif) files with LZW compression in an EPSG:4326 coordinate system. The assessed gross rooftop area datasets are sequenced as BF_X_Y with X having values as 1, 2, 3, 4, and 5 for SSP1, SSP2, SSP3, SSP4, SSP5 narratives with Y representing the assessment year having values as 20, 30, 40, and 50 for years 2020, 2030, 2040, and 2050 and with km2 units. Numerical dataset: Folder - Numerical The global gross estimated rooftop area per FN grid cell for each SSP narrative is provided as a parquet (.parquet) file (Results.parquet). This file contains FN_ID column representing the FN grid cell ID, and other columns representing the FN_ID specific assessed rooftop area. The assessed gross rooftop area columns are sequenced as BF_X_Y with X having values as 1, 2, 3, 4, and 5 for SSP1, SSP2, SSP3, SSP4, SSP5 narratives with Y representing the assessment year having values as 20, 30, 40, and 50 for years 2020, 2030, 2040, and 2050 and with km2 units. In addition, a CF column is added for each FN_ID entry that documents the Capacity Factor for rooftop solar PV based on the World Bank solar atlas. In addition to the main datasets, we have provided additional files to enable generating the vector and numerical datasets from this study: Folder - Models M2_Model.json: This file contains the frozen parameters of the M2 model in .json format generated from XGBoost version 2.0.3 SSP_drivers.parquet: This file contains the driver data used for generating the main dataset in our study FN_MAP.parquet: This file contains the boundary information for each fishnet grid tile in a Well Known Text (WKT) format. Prediction.ipynb: This file provides a python notebook interface to generate inferencing from M2_Model.json using SSP_drivers.parquet file. In addition, this file also generates the numerical dataset and converts it into vector dataset using FN_MAP.parquet file. environment.yaml: This file contains the frozen configuration of python virtual environment used to generate the results presented in this study. Version history: This version corresponds to the revised journal submission (Round 1). The version will be updated upon the completion of the review of the main manuscript. This version V2 is supersedes V1 to correspond with round 1 of review. The database(s) in this version is associated with a Data Descriptor paper manuscript entitled " Global high-resolution growth projections for rooftop area consistent with the shared socioeconomic pathways, 2020-2050 ", submitted to Scientific Reports Journal (https://www.nature.com/srep/) Changelog: The following files from version V1 of this dataset are now archived based on the reviews (Round 1). 1_Geospatial_Dataset_V1.gpkg 2_Countrylevel_gross_rooftop_area_V1.parquet 3_Analytics_Scripts_V1.ipynb