An integrated, feature-based Geographic Information System for Venus

Introduction: Seventy years into spaceborne planetary investigations, the community still lacks comprehensive planetary GISs that integrate feature data within a unified or easy-to-compile system. Such a platform or service would be essential not only for setting up individual research projects, but also for enabling faster, and more effective mission planning. Although similar projects have been proposed for decades (Hargitai 2016), they could never be fully realized, primarily because they would not generate new data, but instead rely on the reuse and restructuring of existing information. As a result, each reserarcher and mission team need to build their own system from scratch. This paper presents what may be the first attempt to establish a public, multipurpose, systematic, feature-based geospatial dataset for Venus, supported by ESA. Such system enables users to explore, query, overlay,  and quantitatively analyze dozens of complementary, vector-based data layers enriched with detailed attribute tables.

Background: The concept of a digital planetary cartographic database goes back to 1981 when R. M. Batson proposed such project to include photo imagery and DTMs, noting that “future additions ... may include information from geologic maps” (Batson 1981). Perhaps the earliest spacecraft-based feature catalog, that of Mars, was created by W. K. Hartmann in 1972 titled “Catalog of all craters on Mars larger than 64 km in diameter”, sent in a letter to the Mariner 9 TV team.

Feature-based mapping, as opposed to complex and region-specific geologic unit mapping, proved to be especially valuable for Venus. Magellan radar data was used for the initial global mapping of volcanic, impact, structural and tectonic features, culminating in several publications in 1992. The Magellan Venus mission also pioneered the creation of a comprehensive multilayer GIS: Price and Suppe (1995) developed the first Venus feature GIS compiled directly on Magellan C1-MIDR data and digitizing existing maps (Tanaka et al. 1997). In their paper, they already foresaw the potential future use of such GIS. That dataset has since been converted to ArcGIS format and made available on the ArcGIS platform in 2024 (https://www.arcgis.com/home/item.html?id=962dcfd6b5b64b21a922bc9b6c94ad78).

One of the first digital, tabulated feature catalogs was published in the Venus II book’s CD-ROM supplement in 1997, where original GIS data were incuded only as low-resolution GIF images. A year earlier USGS Flagstaff published Schaber’s impact crater database online, one of the earliest examples of planetary feature data made available digitally on the web. By 2025, however, this database survives only as a printed, then photocopied and scanned document from that era archived as a pdf file by author (https://pubs.usgs.gov/of/1996/0688/), and on the 1996-captured USGS webpage via the Wayback Machine. This illustrates how early digital databases risk being lost, if they are not actively curated, for example, through incorporation into continuously maintained GIS services. Yet, even government-maintained USGS resources can disappear over time. This underscors the importance of selecting robust, long-term data hosting platforms and preservation strategies to preserve the results (or, often, byproducts) of previous planetary geologic or geomorphic studies.

Venus GIS: In the 30 years since the dawn of digital Venus mapping, more than 70 different feature catalogs had been published in the literature in various forms, ranging from tables to images to illustrator or GIS program made raster or vector files (Fig. 1). The number of published catalogs re-analyzing existing Magellan data has increased significantly in recent years (since 2023), resulting from for planning the missions of the 2030. These provide the pieces of data for a still-unrealized comprehensive Venus GIS.

The feature mapping of Venus by 2025 yielded in approximately 100,000 feature data entities - points, polylines and polygons. These need to be cross-referenced to filter out duplicates across different catalogs and in multiple versions of the same catalogs published over the years. After correlation, each data table is converted into centerpoint vector files with all available attributes, and polyline/polygon files if such data is available. Raster image and non-georeferenced vector data, or outdated digital map formats (e.g., gmt) is converted to GIS formats. Wherever possible, centerpoint data is replaced by outlines mapped directly from the Venus Magellan SAR FMAP mosaics. Geologic units would be added from the original mapping files of formal geologic mapping projects. The features and units would be grouped according to geologic process categories. The resulting GIS would integrate updated data obtained directly from the authors of existing catalogs. Each feature would be linked to its original source and publication reference, ensuring traceability across the GIS.

Creating such GIS requires a coprehensive search and processing phase that extends well beyond the scope of a typical astrogeologic study. It requires a dedicated, multi-year effort focusing solely on GIS development and data integration. This paper reports on the results from the first year of that undertaking.

Cataloguing and creating training data of Venusian arachnoids and novae: The cataloging and visualization of volcano-tectonic structures on the surface of Venus is an important part in the creation of the Venus GIS. The main objectives of this sub-taks were to merge the databases of arachnoids and novaes; to take representative samples of these formations, and then to determine the regions with the highest density and variability of volcanotectonic features (Fig. 2) to provide training data and test/control area for machine learning based mapping.

Using all the available databases and filtering out duplicates, we created attribute tables for both types of structures. For the novae, we connected the endpoints of the trenches one by one, while for the arachnoids, we followed the protruding “legs,” or, if these were absent, the circular outline. In both cases, we used a map magnification level that ensured the accuracy of the drawings.

Acknowledgment: The project is supported by the VERATAC PRODEX project of ESA (4000144802)