Distribution of workplaces in Warsaw 2015-2025 in high spatial resolution

Description

This data presents a 100 square meters spatially disaggregated dataset of workplace locations in Warsaw for the years 2015 and 2025. The dataset was created by integrating employment data from the Warsaw Traffic Survey (WTS) 2015 with high-resolution geospatial data from the Global Human Settlement Layer (GHSL, R2023), taking into account areas with missing data from OpenStreetMap (OSM). In 2015, WTS reported 804,760 workplaces in 800 transport zones. To enable fine-scale analysis, these values were redistributed to a 100 m GHSL grid using non-residential built-up volume and height as allocation factors. Areas without employment, such as parks, cemeteries, airport runways, the Vistula river, and water areas, were excluded using OSM data. produced The 2025 dataset by updating the 2015 distribution with changes in non-residential built-up volume between 2015 and 2025, as well as incorporating newly constructed skyscrapers not included in GHSL 2018 height data. The result is two harmonised datasets representing workplace distribution across Warsaw at 100 m resolution. These datasets provide valuable input for studies of urban accessibility, transport equity, and land-use–transport interactions, and the methodological framework can be replicated in other urban contexts.

Steps to reproduce

The procedure consisted of the following steps: Step 1. Data preparation - Input data: GHSL built-up height (2018), GHSL built-up volume (2015, 2025), boundary of Warsaw. - Operation: Clip raster datasets to the Warsaw boundary, convert raster TIFF files to vector shapefiles. - Output: Vector layers of GHSL built-up height (2018) and volume (2015, 2025) for Warsaw. Step 2. Adjustment for new buildings - Input data: GHSL built-up height (2018), OSM building footprints. - Operation: Identify non-residential buildings constructed after 2018 (especially skyscrapers). Assign estimated building height values to those features. - Output: Updated building height dataset including post-2018 skyscrapers. Step 3. Height recalibration - Input data: Updated building height dataset (2018 + skyscrapers). - Operation: Recalculate average building height per 100 m grid square, weighting skyscraper height by its share of the grid cell area. - Output: Corrected 2025 building height values per grid cell. Step 4. Merging built-up volumes - Input data: GHSL built-up volume 2015, GHSL built-up volume 2025. - Operation: Merge both datasets into one table, assign a grid code for each year. - Output: Integrated dataset with volume values for 2015 and 2025. Step 5. Integration of height and volume - Input data: Dataset from Step 4, building heights (2018, 2025). - Operation: Add height attributes. For 2025, recalculate grid codes by adjusting 2015 values with height differences (2018 vs. 2025). - Output: Dataset with combined information on built-up volume and height for 2015 and 2025. Step 6. Exclusion of non-employment areas - Input data: Dataset from Step 5, OSM polygons of cemeteries, parks, airport and water areas. - Operation: Remove grid cells intersecting with non-employment land uses. - Output: Refined dataset containing only areas with potential employment. Step 7. Allocation of workplaces (2015) - Input data: WTS 2015 workplace counts (804,760 jobs in 800 zones), dataset from Step 6. - Operation: Intersect WTS zones with GHSL grid. Distribute jobs across grids based on (a) share of 2015 non-residential built-up volume and (b) share of grid area within the WTS zone. Average the two proportions to assign job counts. - Output: Workplace distribution for 2015 at 100 m resolution. Step 8. Projection to 2025 - Input data: Dataset from Step 7, GHSL built-up volume change (2015–2025), skyscraper updates. - Operation: Update 2015 workplace values by applying differences in grid codes (2015 vs. 2025). For grids without 2015 values, allocate jobs based on the share of the 2025 grid within the WTS zone relative to the total projected workplaces. - Output: Workplace distribution for 2025 at 100 m resolution.

Institutions

Categories

Funding

Licence