The Project Location Data Model to be standardized is a combination of:
- location-specific,
- project-specific and
- geospatial attributes in the form of coordinates or administrative unit boundaries
that jointly describe a project location, as defined here.

Current Draft of Location-level Attributes for the Standard¶

(as per Excel Template and JSON Scheme below):

• Field ID
• Template Version - this could also be part of the project-level attributes below
• Donor Project-No. - this is also be part of the project-level attributes below
• Abbreviation of project name (project acronym) -- this should rather be part of the project-level attributes below
• Data Provider (Institution Name) ? - this could also be part of the project-level attributes below
• Publishing restrictions due to security reasons ? - this could also be part of the project-level attributes below
• Date of data collection or latest update ?
• Related Community / Village / Neighborhood ?
• Location name ?
• Location Activity Status ?
• Planned or actual start date of activity at the location ?
• Planned or actual end date of activity at the location ?
• Activity Description ?
• Location Type Theme ?
• Location Type Name
• Geographic Exactness
• Geospatial Attributes: coordinates (point, line, polygon) OR admin unit polygon from an administrative (unit) repository OR other sector-specifi polyong repository like the IUCN WDPA Protected Areas repository

Current Draft of Project-level Attributes for the Standard¶

(as per JSON Scheme below):

• Donor Project-No.
• Project Title (which language(s)?)
• Abbreviation of project name (project acronym) ?
• IATI Sector Vocabulary No 1: DAC5-Subsector Code (=CRS-Code) and DAC5-Subsector Title (IATI languages) [DAC5 Digit Sector][https://iatistandard.org/en/iati-standard/203/codelists/sector/)
• Country / Region / Supra-National Institution: IATI code list OR ISO Code list? In IATI, there are no supranational institutions and no regions
• Project Status: existing list of IATI categories in activity type codelist?
• Financing Instrument / Type: IATI codelist finance type?
• Donor / Client (of the project, upstream): IATI codelist? -> non longer maintained. Alternative?
• Project Executing / Implementing Agency(ies) (of the project, downstream): IATI codelist? -> non longer maintained. Alternative?
• ESG category of the project? Vocabulary?
• Project team internally responsible for the data (quality assurance): internal logic

We recommend not to use the following IATI standard elements for our proposed new standard core scheme - all to be discussed:
- Geographic Location Class because it conflicts with the proposed location type scheme without adding value.
- Geographic Location Reach: this distinction is mostly relevant for immaterial location types which already incorporated it. TBD
- Geographical Precision: these categories contain overlaps with other categories and its most important elements are already covered by the new proposed categories of exactness together with the location types schema.

Data Schemas and templates for operationalizing the IATI standard regarding project location data¶

Location types / Investment Types / Asset Types List as a mark down table
This list proposes changes to the following IATI codelist: IATI location type codelist

Location types / Investment Types / Asset Types List as an Excel Sheet
This list proposes changes to the following IATI codelist: IATI location type codelist

Complete Excel Template for collecting location-specific data - English version
In addition to the above-mentioned changes to the IATI location type list, this template uses the IATI activity status codelist and proposes changes to the following IATI codelists: activity date type and geographic exactness
Once it becomes part of the IATI standard, it should use its respective version

JSON Scheme of the location-specific part of the data model containing all elements of the above Excel template

To the above schemes, we still have to add the geographic vocabulary/ies = administrative (unit) boundaries repository/ies below (ideally recommending only one) and potentially add it to the IATI geographic vocabulary list

Administrative Boundaries¶

Requirements for Administrative (Unit) Boundaries

Administrative boundaries are crucial for analysis and visualization. However, these boundaries are subject to frequent changes due to modifications at various levels, such as the merging of districts or the division of municipalities. Maintaining such a dataset internally is impractical for KFW. Therefore, it is essential to identify a reliable existing dataset that meets all analytical and visualization needs while accurately reflecting political boundaries.
In global datasets, administrative levels are typically categorized from level 0 to level 5. Level 0 represents the country level, while levels 1 through 5 correspond to progressively lower administrative divisions (e.g., states, districts, municipalities, regions, etc.). This standardized nomenclature helps avoid discrepancies in naming conventions across countries. For mapping and analysis purposes, KFW requires data up to at least the third administrative level.
Given the political sensitivities surrounding boundaries in certain countries, the dataset must allow for flexible representation depending on the audience. For instance, it should enable users to depict Western Sahara as a separate entity or combine it with Morocco, as needed.

Available Administrative Boundaries Datasets

Based on the defined requirements and the strengths and weaknesses of the various data sources, HDX - OCHA Global Subnational Admin Boundaries emerges as the most suitable option. FieldMaps.io remains a very strong alternative.

FieldMaps.io

Key Advantages of FieldMaps.io: - 1. Comprehensive Coverage: The dataset includes administrative boundaries up to level 5 in some countries, enabling precise analysis. - 2. High Detail: The dataset is highly detailed, making it suitable for use at all zoom levels. - 3. Disputed Boundaries: Disputed areas, as defined by the United Nations, are included and can be easily selected using specific ISO3 codes. This allows users to merge the geometry of disputed areas with neighboring countries based on the intended audience. - 4. Accessibility: The data is freely available with no usage restrictions. - 5. Consistency: The dataset has been consistently maintained over the past five years. - 6. Accuracy: When compared to more official datasets like FAO’s, there are minimal discrepancies in the names and geometries of administrative areas (only two countries show major differences). -7. Flexible Download Options: Users can download the dataset at the global level and by administrative level. Country-level data are also available by original data source; however, these country-specific files are provided without edge-matching across international boundaries.

Limitations of FieldMaps.io: -1. Data Size: The dataset’s extreme level of detail makes it heavy, which can slow down analysis and make it challenging to manage when working with entire regions. -2. Maintenance Risks: The dataset is maintained by a single individual, meaning its quality and timeliness depend on the time they can dedicate to the project. Additionally, there is no guarantee of its long-term availability due to potential hosting and maintenance costs.

HDX – OCHA Global Subnational Administrative Boundaries

Key Advantages of HDX -1. Optimized Base Layer for Performance: To address the limitations of FieldMaps.io—where the first layer is extremely detailed and therefore heavy, slowing analysis and complicating management at regional scale—the HDX dataset uses UN Geodata 1:1M as base layer for level 0. It significantly improves performance while maintaining sufficient spatial accuracy. -2. Global Aggregation: Administrative boundaries for 110 countries are aggregated into a single, unified layer, facilitating regional and global analyses without the need to manage multiple country-specific files. -3. Multiple Boundary Variants: The dataset is available in three complementary versions, allowing users to choose the most appropriate option for their use case: o Edge-Matched (Recommended): Subnational boundaries are aligned to UN Geodata 1:1M international boundaries. This process removes gaps and overlaps at international borders, with only minor simplification at borders. Internal subnational geometries remain unchanged. o Original: Boundaries are preserved exactly as provided by their original sources. Gaps and overlaps may exist at international borders. This version is recommended when maintaining full source integrity is critical. o Extended: A specialized output designed for users who wish to perform their own edge-matching, particularly when not using UN Geodata 1:1M international boundaries. -4. Disputed Areas Handling: Disputed boundaries, as defined by the United Nations, are included and can be easily identified using specific ISO3 codes. This allows users to flexibly merge disputed areas with neighboring countries depending on the intended audience or analytical purpose. -5. Historic Administrative Boundaries: A version of the dataset includes historic boundaries for 40 countries, enabling longitudinal and retrospective analyses. -6. Comprehensive Administrative Coverage: The dataset includes administrative units up to level 4 in some countries, supporting detailed and fine-grained spatial analysis. -7. High Spatial Resolution: The boundaries are highly detailed (level 1 to level 4) and suitable for use across all zoom levels, from national overviews to local-scale mapping. -8. Open and Free Access: The dataset is freely available and distributed with no usage restrictions, supporting broad adoption and reuse. -9. Consistency and Maintenance: The dataset has been consistently maintained for over five years, ensuring stability and reliability for operational and analytical use. -10. High Accuracy: Comparisons with other datasets (e.g., FAO administrative boundaries) show minimal discrepancies in names and geometries. -11. Dataset updates: Users are notified whenever the dataset is updated, which is particularly valuable for workflows requiring up-to-date administrative boundaries. A data quality control dashboard has been implemented to control everything single entry that is added/updated. -12. Schema Standardization: A standardized schema is used for country-level metadata, covering both current and historic administrative boundary versions. Boundary geometries at each administrative level also follow a standardized data schema, ensuring consistency across countries and versions. A quality control dashboard validates all incoming and updated data. -13. Formats available: The dataset is currently available in multiple formats, including File Geodatabase, Shapefile, GeoJSON, and Excel. Additional formats are planned and will be made available through the upcoming Geo Data API.
-14. Update Notification Feature: Users can opt in to receive notifications whenever the dataset is updated. -15. Multiple Access Levels: The dataset is distributed as a single global layer. However, individual country datasets can also be downloaded separately, with the same quality control procedures applied.

Limitations of HDX -1. Incomplete Global Coverage: Currently, the dataset includes administrative boundaries for approximately 110 countries only. As a result, it is less suitable for global-scale analyses or worldwide map visualizations, where comprehensive country coverage is required. Additional country datasets are expected in the coming months. United Nations Office for the Coordination of Humanitarian Affairs (OCHA) aims to achieve full country coverage by incorporating boundary data from other UN agencies. This is an active and evolving area of work, with participating agencies meeting weekly to coordinate efforts. Pragmatic progress toward expanded coverage is anticipated in the near term

Conclusion

While FieldMaps.io meets many of KfW’s operational requirements, its limitations—particularly the large file size and reliance on a single maintainer—warrant careful consideration, especially for long-term and large-scale use. In contrast, HDX – OCHA Global Subnational Administrative Boundaries provides a more robust and scalable alternative, offering aggregated multi-country coverage, multiple boundary variants, consistent long-term maintenance, and the option to subscribe to update notifications when revisions occur. That said, both datasets should be considered secondary sources and used primarily when official country-level administrative boundary data provided by national authorities are unavailable, inaccessible, or outdated. Where authoritative and up-to-date national datasets exist, these should remain the preferred reference. Overall, based on the defined requirements and the comparative assessment of strengths and limitations, HDX emerges as the most suitable primary option for KfW’s needs. FieldMaps.io remains a valuable complementary or alternative dataset in specific contexts, depending on operational constraints, performance requirements, and use-case specificity.