Second of four zipfiles providing all data and Python code necessary to replicate any of the 13 development potential indexes (DPIs) described within Oakleaf et al. (2019), “Mapping global development potential for renewable energy, fossil fuels, mining and agriculture sectors”. A README.pdf guides users on setting up environment necessary to use data and run Python code.

To run Python code with accompanying spatial data, 64 GBs of disk space is required. Additionally ArcPY, a python module associated with ESRI’s ArcGIS Desktop, and an accompanying Spatial Analyst extension license are required to run Python code. All code was created by J.R. Oakleaf during 2018 and is licensed under Creative Commons Attribution-NonCommercial 4.0 International License http://creativecommons.org/licenses/by-nc/4.0/.

This repository contains the data used for the book chapter by De Sabbata et al (2025), including the Free Gazetteer Data made available by GeoNames under CC BY 4.0, a file containing the names of the Italian provinces was made available by Michele Tizzoni under CC BY 4.0, and data derived from them using Mistral-7B-Instruct-v0.2, which was made available by the Mistral AI Team under Apache License 2.0. The code used to process the data is available via our related GitHub repository under MIT Licence.The Author Accepted Manuscript of "Geospatial Mechanistic Interpretability of Large Language Models" available on arXiv (arXiv:2505.03368).De Sabbata, S., Mizzaro, S. and Roitero, K. (2025) “Geospatial mechanistic interpretability of large language models,” in Janowicz, K. et al. (eds.) Geography according to ChatGPT. IOS Press (Frontiers in artificial intelligence and applications).Abstract: Large language models (LLMs) have demonstrated unprecedented capabilities across various natural language processing tasks. Their ability to process and generate viable text and code has made them ubiquitous in many fields, while their deployment as knowledge bases and ``reasoning'' tools remains an area of ongoing research. In geography, a growing body of literature has been focusing on evaluating LLMs' geographical knowledge and their ability to perform spatial reasoning. However, very little is still known about the internal functioning of these models, especially about how they process geographical information.In this chapter, we establish a novel framework for the study of geospatial mechanistic interpretability -- using spatial analysis to reverse engineer how LLMs handle geographical information. Our aim is to advance our understanding of the internal representations that these complex models generate while processing geographical information -- what one might call "how LLMs think about geographic information" if such phrasing was not an undue anthropomorphism.We first outline the use of probing in revealing internal structures within LLMs. We then introduce the field of mechanistic interpretability, discussing the superposition hypothesis and the role of sparse autoencoders in disentangling polysemantic internal representations of LLMs into more interpretable, monosemantic features.In our experiments, we use spatial autocorrelation to show how features obtained for placenames display spatial patterns related to their geographic location and can thus be interpreted geospatially, providing insights into how these models process geographical information. We conclude by discussing how our framework can help shape the study and use of foundation models in geography.

The data set "20m height lines Wuppertal 2015 (with exemption)" contains contour lines with an equidistance of 20 m and exemption at the buildings from the official property cadastre information system ALKIS for the entire urban area of Wuppertal. There are 5 different characteristics with regard to data format and coordinate system. The dataset is based on the Digital Terrain Model DGM1 provided by Geobasis NRW with the predominant data status in 2015 (stands 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal urban area). The DGM1 is a regular point grid generated by aircraft-based laser scanning (Lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. The contour lines were generated with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension by interpolation in DGM1 without taking into account additional terrain fracture edges. In addition to the standard behavior of the tool "Surface - Contour Line" available in the Spatial Analyst, no explicit further smoothing of the curve of the contour lines took place. The dataset will not be continued. Following an update of the DGM1, the Wuppertal contour lines are re-derived under other product names. The dataset is available under an Open Data license (CC BY 4.0).

The data set "100m-Höhenlinien Wuppertal 2015" contains contour lines with an equidistance of 100 m for the entire urban area of Wuppertal. There are 5 different characteristics with regard to data format and coordinate system. The dataset is based on the Digital Terrain Model DGM1 provided by Geobasis NRW with the predominant data status in 2015 (stands 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal urban area). The DGM1 is a regular point grid generated by aircraft-based laser scanning (Lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. The contour lines were generated with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension by interpolation in DGM1 without taking into account additional terrain fracture edges. In addition to the standard behavior of the tool "Surface - Contour Line" available in the Spatial Analyst, no explicit further smoothing of the curve of the contour lines took place. The dataset will not be continued. Following an update of the DGM1, the Wuppertal contour lines are re-derived under other product names. The dataset is available under an Open Data license (CC BY 4.0).

The data set “20 m-height lines Wuppertal 2015” contains elevation lines with an equidistance of 20 m for the entire city area of Wuppertal. There are 5 different types of data format and coordinate system. The data set is based on the digital terrain model DGM1 provided by Geobasis NRW with the predominant data level in 2015 (stands 2012 and 2017 in some northern and eastern peripheral areas of Wuppertal’s urban area). The DGM1 is a regular point grid generated by aircraft-assisted laser scanning (Lidar) with a mesh width of 1 m and a height accuracy of a single point of ± 2 dm. The vertical lines were created with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension by interpolation in the DGM1 without taking into account additional ground breaking edges. Beyond the standard behavior of the “surface contour line” tool available in the Spatial Analyst, there was no explicit further smoothing of the curvature of the vertical lines. The record will not be continued. After an update of the DGM1, the Wuppertal height lines are newly derived under other product names. The dataset is available under an open data license (CC BY 4.0).

The data set "5m contour lines Wuppertal 2015" contains contour lines with an equidistance of 5 m for the entire urban area of ​​Wuppertal. There are 5 different specifications regarding data format and coordinate system. The data set is based on the digital terrain model DGM1 provided by Geobasis NRW with the predominant data status of 2015 (statuses 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal city area). The DGM1 is a regular grid of points generated by airborne laser scanning (lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. The contour lines were created with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension through interpolation in DTM1 without considering additional terrain break edges. Beyond the standard behavior of the "Surface - Contour Line" tool available in Spatial Analyst, there was no further explicit smoothing of the curvature of the contour lines. The record is not continued. After an update of the DGM1, the Wuppertal contour lines are re-derived under different product designations. The data set is available under an open data license (CC BY 4.0).

The data set "100m contour lines Wuppertal 2015" contains contour lines with an equidistance of 100 m for the entire urban area of ​​Wuppertal. There are 5 different specifications regarding data format and coordinate system. The data set is based on the digital terrain model DGM1 provided by Geobasis NRW with the predominant data status of 2015 (statuses 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal city area). The DGM1 is a regular grid of points generated by airborne laser scanning (lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. The contour lines were created with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension through interpolation in DTM1 without considering additional terrain break edges. Beyond the standard behavior of the "Surface - Contour Line" tool available in Spatial Analyst, there was no further explicit smoothing of the curvature of the contour lines. The record is not continued. After an update of the DGM1, the Wuppertal contour lines are re-derived under different product designations. The data set is available under an open data license (CC BY 4.0).

WUImap est un logiciel de cartographie des interfaces habitat-forêt. Cet outil est une extension du logiciel ©ESRI ArcGIS Desktop® ArcMap™. Il permet de cartographier les types d’interfaces habitat-forêt définis dans le document, en référence à l’article Lampin-Maillet et al. (2010). Il permet de produire une cartographe des types d’habitat résidentiel. Il ne fonctionne qu’avec une licence valide ArcGIS Desktop® version10.x et une licence pour le module « Spatial Analyst ».

In 2009, the Kentucky Water Science Center completed the Water Availability Tool for Environmental Resources (WATER-KY), which provided the ability to simulate streamflow for the period 1980-2000. This model integrated TOPMODEL (Beven and Kirkby, 1979) for pervious portions of the landscape with simulation of flow generated from impervious surfaces (USDA, 1986). Associated products included a flow-duration curve, load-duration curves when water-quality data were available, and general water balance. WATER-KY required a dedicated ArcGIS license with the Spatial Analyst extension, which made it difficult to use for some cooperators and limited integration with other hydrologic approaches. This new version translates the abilities of WATER to a format that can be used without proprietary software or local updating of software. Additional functionality has also been added to include hydrologic response units (HRUs) that are defined based on three fundamental land-use categories: forest, agricultural land, and developed areas, based on subsequent development of WATER for the Delaware Basin (Williamson and others, 2015). Beven, K.J., and Kirkby, M.J., 1979, A physically based, variable contributing area model of basin hydrology / Un modèle à base physique de zone d'appel variable de l'hydrologie du bassin versant: Hydrological Sciences Bulletin v. 24, p. 43-69, http://dx.doi.org/10.1080/02626667909491834. U.S. Department of Agriculture [USDA], 1986, Urban hydrology for small watersheds: Natural Resources Conservation Service, Conservation Engineering Division, Technical Release 55, Revised June 1986, Update of Appendix A January 1999, https://www.nrcs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb1044171.pdf. Williamson, T.N., Lant, J.G., Claggett, P.R., Nystrom, E.A., Milly, P.C.D., Nelson, H.L., Hoffman, S.A., Colarullo, S.J., and Fischer, J.M., 2015, Summary of hydrologic modeling for the Delaware River Basin using the Water Availability Tool for Environmental Resources (WATER): U.S. Geological Survey Scientific Investigations Report 2015–5143, 68 p., http://dx.doi.org/10.3133/sir20155143.

Based on open access data, 79 Mediterranean passenger ports are analyzed to compare their infrastructure, hinterland accessibility and offered multi-modality categories. Comparative Geo-spatial analysis is also carried out by using the data normalization method in order to visualize the ports' performance on maps. These data driven comprehensive analytical results can bring added value to sustainable development policy and planning initiatives in the Mediterranean Region. The analyzed elements can be also contributed to the development of passenger port performance indicators. The empirical research methods used for the Mediterranean passenger ports can be replicated for transport nodes of any region around the world to determine their relative performance on selected criteria for improvement and planning.

The Mediterranean passenger ports were initially categorized into cruise and ferry ports. The cruise ports were identified from the member list of the Association for the Mediterranean Cruise Ports (MedCruise), representing more than 80% of the cruise tourism activities per country. The identified cruise ports were mapped by selecting the corresponding geo-referenced ports from the map layer developed by the European Marine Observation and Data Network (EMODnet). The United Nations (UN) Code for Trade and Transport Locations (LOCODE) was identified for each of the cruise ports as the common criteria to carry out the selection. The identified cruise ports not listed by the EMODnet were added to the geo-database by using under license the editing function of the ArcMap (version 10.1) geographic information system software. The ferry ports were identified from the open access industry initiative data provided by the Ferrylines, and were mapped in a similar way as the cruise ports (Figure 1).

Based on the available data from the identified cruise ports, a database (see Table A1–A3) was created for a Mediterranean scale analysis. The ferry ports were excluded due to the unavailability of relevant information on selected criteria (Table 2). However, the cruise ports serving as ferry passenger ports were identified in order to maximize the scope of the analysis. Port infrastructure and hinterland accessibility data were collected from the statistical reports published by the MedCruise, which are a compilation of data provided by its individual member port authorities and the cruise terminal operators. Other supplementary sources were the European Sea Ports Organization (ESPO) and the Global Ports Holding, a cruise terminal operator with an established presence in the Mediterranean. Additionally, open access data sources (e.g. the Google Maps and Trip Advisor) were consulted in order to identify the multi-modal transports and bridge the data gaps on hinterland accessibility by measuring the approximate distances.

This spatial layer displays Range Tenures (grazing and hay cutting licence and permits) administered by the Ministry of Forests, Lands and Natural Resource Operations. A Range Tenure is an area of Crown rangeland where a Range Act tenure applies. Tenure holders access a defined amount of forage through grazing (measured in Animal Unit Months) or hay (tonnage). Range Tenures apply only to Crown Land. In some cases, digital boundaries may overlap Private Land but these lands are not part of the Grazing area (as described in the legal description). Grazing may overlap waterbodies during drawdown (also described legally in the Tenure documents and where applicable, Range Use Plan). Livestock may graze islands and large bodies of water may act as Natural Range Barriers. March 3, 2023: Updates to the Range Tenure attribute table as described. Please contact Nancy.Elliot@gov.bc.ca if you have questions. To guide your use: 1. Layer contains RETIRED, PENDING, and ACTIVE Tenures (=Licenses and Permits); Select ACTIVE under attribute LIFE_CYCLE_STATUS_CODE to isolate all active tenures (status refers to spatial boundary status - therefore, a spatial boundary must be Retired in the Forest Tenure Administration system to have a status of retired); 2. Unique ID for Polygon is by RAN# in attribute FOREST_FILE_ID. Each tenure has own its RAN# (e.g. RAN07777). Multiple areas (polygons) belonging to the same tenure may have same RAN# with unique map block ids (e.g. RAN07777 A, RAN07777 B) 3. The field SUM_TENURE_ACTIVE_AREA_HA will provide, where LIFE_CYCLE_STATUS_CODE is ACTIVE, the total tenure area in Ha. For single block (polygon) tenures, this will be the same area as the polygon. Where there are multiple blocks, this will be the total sum area of all Active blocks. (Note this does not include PENDING or RETIRED tenure areas. Area is for Approved and Active Tenure boundaries. A block may be Approved but Pending, and therefore is not included). Review data for multipart vs multi polygons 4. AUTHORIZED_USE and TOTAL_ANNUAL_USE are for the entire Tenure; where there are multiple blocks, the total is over all blocks, seasonally distributed through different pastures 3. FILE_TYPE_CODE contains information on the type of Permit: • E01 - Grazing License • E02 - Grazing Permit • H01 - Haycutting License • H02 - Haycutting Permit 4. IF YOU ARE DOING SPATIAL ANALYSIS based on AREA --- Please note that tenures overlap either partially or wholly amongst tenure holders. One area may be shared by more than one Tenure holder and therefore there will be multiple congruent or partially overlapping polygons (multipart). If you want to do SPATIAL ANALYSIS based strictly on area, you must collapse or flatten the data using DISSOLVE so that the polygons are 1:1 with the land base (suggested approach)

To create this layer, OCTO staff used ABCA's definition of “Full-Service Grocery Stores” (https://abca.dc.gov/page/full-service-grocery-store#gsc.tab=0)– pulled from the Food System Assessment below), and using those criteria, determined locations that fulfilled the categories in section 1 of the definition.Then, staff reviewed the Office of Planning’s Food System Assessment (https://dcfoodpolicycouncilorg.files.wordpress.com/2019/06/2018-food-system-assessment-final-6.13.pdf) list in Appendix D, comparing that to the created from the ABCA definition, which led to the addition of a additional examples that meet, or come very close to, the full-service grocery store criteria. The explanation from Office of Planning regarding how the agency created their list:“To determine the number of grocery stores in the District, we analyzed existing business licenses in the Department of Consumer and Regulatory Affairs (2018) Business License Verification system (located at https://eservices.dcra.dc.gov/BBLV/Default.aspx). To distinguish grocery stores from convenience stores, we applied the Alcohol Beverage and Cannabis Administration’s (ABCA) definition of a full-service grocery store. This definition requires a store to be licensed as a grocery store, sell at least six different food categories, dedicate either 50% of the store’s total square feet or 6,000 square feet to selling food, and dedicate at least 5% of the selling area to each food category. This definition can be found at https://abca.dc.gov/page/full-service-grocery-store#gsc.tab=0. To distinguish small grocery stores from large grocery stores, we categorized large grocery stores as those 10,000 square feet or more. This analysis was conducted using data from the WDCEP’s Retail and Restaurants webpage (located at https://wdcep.com/dc-industries/retail/) and using ARCGIS Spatial Analysis tools when existing data was not available. Our final numbers differ slightly from existing reports like the DC Hunger Solutions’ Closing the Grocery Store Gap and WDCEP’s Grocery Store Opportunities Map; this difference likely comes from differences in our methodology and our exclusion of stores that have closed.”Staff also conducted a visual analysis of locations and relied on personal experience of visits to locations to determine whether they should be included in the list.

This dataset contains spatial and statistical data related to riparian forests in the region of Vojvodina, Serbia. The data were compiled and processed as part of a research project focusing on the classification, distribution, and temporal changes of riparian forest types along major rivers such as the Danube, Tisa, Tamiš, and Sava.

The dataset includes:

• Raster maps and classified satellite imagery (.tif and .png formats),

• Vector layers and geospatial shapefiles,

• Tabular data with field-based and derived statistics,

• Forest typology maps and vegetation structure analyses,

• Comparative land cover data from 2012 and 2018 based on CORINE Land Cover (CLC),

The visualizations and spatial layers were generated using GIS tools (QGIS, ArcGIS) and remote sensing methods. Public data sources were used, including:

• CORINE Land Cover (© EEA, Copernicus open access),

• EUFORGEN species distribution maps,

• National Forest Inventory of Serbia (NFI).

All other processing, classification, and statistical analysis were conducted by the dataset author. Data are provided in multiple ZIP archives grouped thematically due to file count limitations.

License and Terms:

• CORINE data © European Environment Agency – licensed under Copernicus open access.

• EUFORGEN data © EUFORGEN – provided for research and educational purposes.

• All derived maps, analyses, and statistical tables created by the author are released under Creative Commons Attribution 4.0 International (CC-BY 4.0).

Citation:

If you use this dataset, please cite the original project and the author. Proper citation allows us to maintain and continue sharing open-access datasets.

Contact:

Vladimir Visacki
University of Novi Sad
Institute for Lowland Forestry and Environment (ILFE)
✉️ vladimir.visacki@gmail.com

The data set "20m contour lines Wuppertal 2015" contains contour lines with an equidistance of 20 m for the entire urban area of ​​Wuppertal. There are 5 different specifications regarding data format and coordinate system. The data set is based on the digital terrain model DGM1 provided by Geobasis NRW with the predominant data status of 2015 (statuses 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal city area). The DGM1 is a regular grid of points generated by airborne laser scanning (lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. The contour lines were created with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension through interpolation in DTM1 without considering additional terrain break edges. Beyond the standard behavior of the "Surface - Contour Line" tool available in Spatial Analyst, there was no further explicit smoothing of the curvature of the contour lines. The record is not continued. After an update of the DGM1, the Wuppertal contour lines are re-derived under different product designations. The data set is available under an open data license (CC BY 4.0).

Dunedin City in the South Island of New Zealand has many assets and critical infrastructure sitting on a low-lying coastal plain that is underlain by a largely unseen and relatively poorly understood hazard. Shallow groundwater in this area limits the unsaturated ground available to store rain and runoff, promotes flooding and creates opportunities for infiltration into stormwater and wastewater networks. Groundwater levels are expected to rise as sea level rises, causing greater frequency of flooding and/or direct inundation once it nears the ground surface. This zipped archive contains ArcGIS 10.8 geodatabases and spatial analysis of data gathered from a shallow groundwater monitoring network between 6/3/2019 and 1/5/2023. A series of statistical surfaces represent the present-day (2023) water table elevation and depth to groundwater, the response to rainfall recharge and tidal forcing, the available subsurface storage of rain infiltration. Simple geometric models have also been developed using the present shape and position of the water table, combined with tidal fluctuations, to forecast the future state of groundwater levels at 10 cm increments of sea level rise (up to 1 m). The geometric models are strongly empirical, with many implicit assumptions and caveats – particularly, that they do not account for groundwater flow and possible changes in water-budget mass balance. Although many variables and controlling processes are simplified into a single parameter, the projected groundwater levels highlight how local variations in the water table shape and slope interact locally with the ground elevation or infrastructure networks. They are best considered as a worst-case analysis of groundwater-related contribution to hazard and how this will evolve over time. Data are licensed under Creative Commons Attribution 4.0 (CC-BY-4.0) license without warranty. Further description of these data, and implications from the analysis, can be found in the GNS Science metadata catalogue https://data.gns.cri.nz/metadata/srv/eng/catalog.search#/metadata/06a86338-a0fa-436f-9c80-2d8da1dcd64f oe Cox et al. (2023) GNS Science Report 2023/43 doi:10.21420/5799-N894.

(ACASIAN) is an academic and applied research institution specializing in;SIIASA and SIIRCEASA consortia, and licenses its spatial data sets for academic research and commercial purposes.;Geographical Information System (GIS) databases for Asia and the former Soviet Union. ACASIAN fulfills Griffith University's leading role in the;The Australian Centre of the Asian Spatial Information and Analysis Network

The compressed package (Study_code.zip) contains the code files implemented by an under review paper ("What you see is what you get: Delineating urban jobs-housing spatial distribution at a parcel scale by using street view imagery based on deep learning technique").The compressed package (input_land_parcel_with_attributes.zip) is the sampled mixed "jobs-housing" attributes data of the study area with multiple probability attributes (Only working, Only living, working and living) at the land parcel scale.The compressed package (input_street_view_images.zip) is the surrounding street view data near sampled land parcels (input_land_parcel_with_attributes.zip) with the pixel size of 240*160 obtained from Tencent map (https://map.qq.com/).The compressed package (output_results.zip) contains the result vector files (Jobs-housing pattern distribution and error distribution) and file description (Readme.txt).This project uses some Python open source libraries (Numpy, Pandas, Selenium, Gdal, Pytorch and sklearn). This project complies with the GPL license.Numpy (https://numpy.org/) is an open source numerical calculation tool developed by Travis Oliphant. Used in this project for matrix operation. This library complies with the BSD license.Pandas (https://pandas.pydata.org/) is an open source library, providing high-performance, easy-to-use data structures and data analysis tools. This library complies with the BSD license.Selenium(https://www.selenium.dev/) is a suite of tools for automating web browsers.Used in this project for getting street view images.This library complies with the BSD license.Gdal(https://gdal.org/) is a translator library for raster and vector geospatial data formats.Used in this project for processing geospatial data.This library complies with the BSD license.Pytorch(https://pytorch.org/) is an open source machine learning framework that accelerates the path from research prototyping to production deployment.Used in this project for deep learning.This library complies with the BSD license.sklearn(https://scikit-learn.org/) is an open source machine learning tool for python.Used in this project for comparing precision metrics.This library complies with the BSD license.

The data series "Contours Wuppertal 2015" includes 6 contour data sets in the German main height network 2016 (DHHN2016) with different equidistances (1 m, 5 m, 10 m, 20 m, 50 m and 100 m), each in 2 versions with and without exemption on the buildings of the real estate cadastre. Each of these 12 datasets is available in 5 different versions in terms of data format and coordinate system. All datasets are based on the digital terrain model DGM1 provided by Geobasis NRW with the predominant data status 2015 (statuses 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal city area). The DGM1 is a regular grid of points generated by airborne laser scanning (lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. All contour lines were generated with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension through interpolation in DTM1 without considering additional terrain break edges. Beyond the standard behavior of the "Surface - Contour Line" tool available in Spatial Analyst, there was no further explicit smoothing of the curvature of the contour lines. The data sets from this series will not be continued. After an update of the DGM1, the Wuppertal contour lines are re-derived under different product designations. All datasets in the series are available under an open data license (CC BY 4.0).

Then, we reviewed the Office of Planning’s Food System Assessment (https://dcfoodpolicycouncilorg.files.wordpress.com/2019/06/2018-food-system-assessment-final-6.13.pdf) list in Appendix D, comparing that to the created from the ABCA definition, which led to the addition of a few more examples that meet or come very close to the full-service grocery store criteria. Here’s the explanation from OP regarding how they came to create their list:“To determine the number of grocery stores in the District, we analyzed existing business licenses in the Department of Consumer and Regulatory Affairs (2018) Business License Verification system (located at https://eservices.dcra.dc.gov/BBLV/Default.aspx). To distinguish grocery stores from convenience stores, we applied the Alcohol Beverage and Cannabis Administration’s (ABCA) definition of a full-service grocery store. This definition requires a store to be licensed as a grocery store, sell at least six different food categories, dedicate either 50% of the store’s total square feet or 6,000 square feet to selling food, and dedicate at least 5% of the selling area to each food category. This definition can be found at https://abca.dc.gov/page/full-service-grocery-store#gsc.tab=0. To distinguish small grocery stores from large grocery stores, we categorized large grocery stores as those 10,000 square feet or more. This analysis was conducted using data from the WDCEP’s Retail and Restaurants webpage (located at https://wdcep.com/dc-industries/retail/) and using ARCGIS Spatial Analysis tools when existing data was not available. Our final numbers differ slightly from existing reports like the DC Hunger Solutions’ Closing the Grocery Store Gap and WDCEP’s Grocery Store Opportunities Map; this difference likely comes from differences in our methodology and our exclusion of stores that have closed.”We also conducted a visual analysis of locations and relied on personal experience of visits to locations to determine whether they should be included in the list.

The data set "20m contour lines Wuppertal 2015 (with exemption)" contains contour lines with an equidistance of 20 m and exemption on the buildings from the official real estate cadastre information system ALKIS for the entire urban area of ​​Wuppertal. There are 5 different specifications regarding data format and coordinate system. The data set is based on the digital terrain model DGM1 provided by Geobasis NRW with the predominant data status of 2015 (statuses 2012 and 2017 in some northern and eastern peripheral areas of the Wuppertal city area). The DGM1 is a regular grid of points generated by airborne laser scanning (lidar) with a mesh size of 1 m and a height accuracy of a single point of +/- 2 dm. The contour lines were created with ArcGIS 10.6 and the ArcGIS Spatial Analyst extension through interpolation in DTM1 without considering additional terrain break edges. Beyond the standard behavior of the "Surface - Contour Line" tool available in Spatial Analyst, there was no further explicit smoothing of the curvature of the contour lines. The record is not continued. After an update of the DGM1, the Wuppertal contour lines are re-derived under different product designations. The data set is available under an open data license (CC BY 4.0).