Cartographic Metadata for Improving Accessibility and Facilitating Knowledge Extraction and Validation in Planetary Mapping Based on Remote-Sensing Observations

Abstract

The field of planetary mapping and cartography builds almost exclusively on remote-sensing data and can be defined by three distinct concepts: systematic imaging as performed through spacecraft surveying, reference mapping as performed through the compilation of reference maps, i.e., regional to global image and topographic maps, and thematic mapping, which aims at abstracting and contextualizing spatial information to generate complex thematic maps, such as geologic or geomorphologic maps. While thematic mapping represents the highest form of abstraction of information that is provided through systematic mapping, thematic mapping also provides scientific reasoning in support of systematic mapping and exploration through spatially contextualized knowledge. For the development of knowledge, it is paramount to manage and exploit the value of thematic maps as research products, and to design a reliable and transparent development process from the beginning of the mapping phase as there is almost no validation for thematic maps. A key element in accomplishing these objectives is well-designed structures and metadata which are maintained within spatial data infrastructures (SDI) and shared as a coordinated process in research data management through data models. In this contribution, we focus on the need to transfer planetary thematic maps into findable, accessible, interoperable, reusable (FAIR), as well as transparent research data assets to facilitate improved knowledge extraction and also to compensate for limitations caused by the lack of conventional validation options. We review the current status of planetary thematic mapping, and we discuss the principles and roles of mappers and publishers in the process of creating and stewarding digital planetary maps and associated data products. We then present and discuss a set of recommendations that are closely tied to the FAIR concepts in research data management to accomplish such tasks.

1. Introduction

1.1. Background

Planetary mapping comprises various investigation activities related to the exploration of planetary bodies using remote sensing data. It is targeted at the collection and assembly of spatial information to create maps and cartographic representations [1,2,3,4,5,6,7]. Its objective is to advance our understanding of the evolution and characteristics of planetary bodies by contextualizing spatial observations to help answer basic questions or to help investigate sites for future exploration and potential exploitation. In its most general form, planetary mapping may be described as the systematic and comprehensive acquisition, analysis, and representation of spatial data pertaining to the surface of planetary objects, such as planets, moons, and asteroids. The concept of mapping can also be described through methods and the outcome by focusing on instruments, spacecraft, and ground-based observations to create detailed cartographic representations, topographic and image models, as well as compositional analyses for the scientific investigation of planetary surface characteristics. By focusing on mapping methods and output, planetary mapping may be represented through three distinct concepts:

• systematic or reconnaissance mapping, i.e., raw data acquisition and basic processing using remote-sensing platforms mainly;
• reference mapping in the form of compiling reference maps, i.e., mostly image and topographic reference maps;
• thematic mapping, i.e., the abstraction of information to build complex-analytical thematic maps. These can be geological maps, landing-site maps, or any other form of thematic information within a consistent spatial context [8].

These concepts are interdependent, with systematic mapping being the first step leading to reference mapping as the second step and ultimately thematic mapping which also represents the highest form of abstraction of information. Each of these phases contains an inherent scientific value which is increased along further abstraction. They are interdependent because thematic mapping provides scientific context and reasoning for future systematic mapping. Each mapping step includes specific data products (Figure 1 and Figure 2) and with the implementation of each new step, the level of extractable potential knowledge increases. The increase in knowledge and abstraction along the path from systematic to thematic mapping also increases the accessible scientific value of data from systematic to thematic mapping. It can, therefore, be argued that systematic mapping provides the necessary foundational data [9,10,11,12], while thematic mapping provides one aspect of scientific value and the reasoning behind systematic mapping. This then closes the cycle (Figure 1) and establishes the connection between these three mapping concepts. This concept also implies that the quality of a thematic map highly depends on the quality of each input product that contributed to the development of a thematic map [8].

Our point of departure for this investigation is the assumption that digital maps need to have three characteristics to qualify as reusable and sustainable research products, which we will discuss in the contribution:

• digital maps are living assets for the community that can be efficiently located, accessed, and returned by the community;
• digital maps are contextual assets that need to serve as higher-order foundational data on which new research investigations can build and which can be integrated into new mapping investigations;
• digital maps are transparent assets which allow insights into the provenance and lineage of the map as well as its foundational data that were used for the development of the map.

In this investigation, we focus on planetary thematic maps, in particular, planetary geologic maps as, thus far, they form the highest form of cartographic abstraction within the domain (see Figure 2). Among all the thematic maps published in the planetary domain, geologic maps are the most prominent ones in terms of number and usage. These maps highlight the geologic and geomorphic settings within an investigation area and thus constitute a highly efficient way to communicate spatiotemporal relationships and serve as a foundation for building an associated research investigation. Also, planetary “geologic maps are perceived as the most relevant planetary geoscience map type, though geomorphologic, surficial, and compositional maps are also relevant map types”, according to recent survey findings by the United States Geological Survey (USGS) [6].

As of today, the only long-term and stable coordinated program in existence to support planetary thematic mapping is that of the USGS with the earliest maps published in the early 1960s [13,14] and a continuous publication record of maps and associated publications spanning the last 60 years [15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31]. The USGS supports the community with tools and provides a platform for the publication of mapping investigations that undergo a rigorous internal scientific and technical review process [4,6,7]. In addition, published maps and associated data are stored and made accessible through the USGS library and the digital warehouse facility for long-term use and have recently been integrated into an interactive webGIS platform for interactive searching and viewing (usgs.maps.arcgis.com (accessed on 17 February 2024)). Such platforms are further complemented by tools and access points to investigate reference data, such as JMARS [32,33,34].

Besides institutionally coordinated mapping investigations, there are investigations that are either exclusively carried out by research individuals or that are conducted as part of a scaled-up mapping investigation coordinated by a principal investigator for the relatively short duration of a funded project. In both cases, mapping investigations are commonly published in a conventional research outlet and are accompanied by a map or figure depicting a typically downscaled thematic map. The actual map and its associated data remain often unpublished.

Notable recent scaled-up mapping projects include two collaborative mapping investigations conducted in the context of the Dawn mission to Vesta [35,36,37,38,39,40,41,42,43,44] and Ceres [41,45,46,47,48,49,50,51,52,53,54,55]. The results were published in a standard journal as part of a special issue highlighting the geologic findings of these mapping investigations. While geologic mapping was the core objective of these investigations, no map sheets have ever been published alongside the reports. Individual, non-coordinated community mapping investigations, i.e., mappings carried out without any overarching coordination or conducted within a programmatic context, include the majority of investigations published in conventional research journals as well as dedicated map journals [56,57,58,59,60,61,62].

The publication landscape for planetary geologic maps is heterogeneous with essentially no agreed-upon standardization in the community and with widespread outlets that operate differently using their very own metadata specification and infrastructure. This ranges from map-centric warehouses providing public access to maps and rich map data in the case of the USGS, over commercial map-focused journals, to publishers that do not have any specific platform to publish maps. The community has been recognizing the lack of standardization and dissemination platforms for geologic maps for many years [63]. Over the last decade, efforts have been undertaken to discuss and even to develop standardized mapping routines across national boundaries. To a large extent, the Planetary Cartography and Geologic Mapping Working Group (PCGMWG) and its successor, the Mapping and Planetary Spatial Infrastructure Team (MAPSIT) with its Geologic Mapping Subcommittee (GEMS) have been addressing these issues, albeit with a strong focus on the requirements of the US Geological Survey (USGS) mapping requirements. While MAPSIT’s main aim is the development of tools for planetary data infrastructures, GEMS explicitly focuses on geologic mapping and analysis needs. For its active period 2022–2027, GEMS’ aims at discussing standards, guidelines, and workflows for the creation of standardized and non-standardized geologic maps for known bodies and for bodies to which mapping techniques have not been developed and applied yet [64].

On the European side, projects exist that could be summarized as investigations and implementations of prototype infrastructures targeted at improving access, creation, and use of planetary data, maps, and tools. These projects were or are being funded by the European Commission (EC) in the context of the European Union Horizon 2020 framework. A representative of these developments has been the PLANetary MAPping project (PLANMAP [65]) which is aimed at providing data-rich geological maps that are disseminated using webGIS technology [66]. Published geologic mapping investigations conducted within the context of the project include the main focus bodies, i.e., the Moon [67,68,69,70], Mercury [59,71,72] and Mars [73]. The EC project Geologic MApping of Planetary bodies (GMAP) constitutes a continuation of sorts and an extension to PLANMAP by integrating additional projects and by aiming at providing a European infrastructure for planetary data [74,75]. Also, an ongoing China–EU project on Key Technologies and Demonstration of Standardised Planetary Geologic Mapping funded by the Chinese Ministry of Science and Technology (MOST) aims at standardizing planetary mapping approaches.

Despite innovative approaches on the European side, such discussions are not new, and initial ideas date back many decades. The lack of planetary missions on the European side did not make such realizations feasible but with the advent of European spacecraft exploration such as SMART-1 to the Moon (2003–2006), Mars Express (2003–), or Bepi Colombo (2018–) to Mercury led by the European Space Agency (ESA), the topic was revisited. Discussions were led during the Geologic Mapping of Mars workshop in Tuscany, Italy, in 2009 which addressed these matters as a side aspect under the participation of selected European terrestrial geologic surveys as well as the US geological survey (e.g., [76]).

Coordinated geologic mapping and infrastructure initiatives may exist in other countries that pursue activities in programmatic planetary spacecraft exploration, in particular, Russia, India, China, and South Korea. Yet, very little is known and communicated in the international research literature, despite the considerable potential to share and benefit from experiences. All these countries have a strong national history in terrestrial geologic mapping and resource exploration with many decades of experience to build on. It is conceivable that a coordinated and streamlined international approach—given the different infrastructural development levels and national constraints, data policies, and the anticipated future value of such data—may not substantialize anytime soon.

The lack of standardization limits the reuse of produced maps, and any lack of visibility of investigations limits the accessibility to such products, thus limiting the development and knowledge based on previous investigations. On a higher level, solutions require research data infrastructure. On a lower level, they require an awareness and the participation of mappers and research individuals. When taking considerations of large-scale infrastructure out of the equation when thinking about solutions to make the results of mapping investigations available in the community, we need to look at how research individuals can contribute to the process. Top-down approaches might work within national boundaries, or within state unions such as the European Union which implemented the Infrastructure for Spatial Information in Europe (INSPIRE, [77,78]). Community-driven bottom-up approaches are targeted initiatives, but they require stable funding and long-term development until they are eventually realized. It, therefore, seems feasible to consider parallel approaches to how individual researchers can ascertain that their maps, map data, and associated material become living data assets that can be reused and that are transparent concerning their development process. While the overall mapping process might not be a matter of standardization across national boundaries anytime soon, a first step towards the direction of unification and increasing interchangeability and transparency would be to agree on using already established methods to describe data and to make them available to the community.

The aforementioned challenges are not new and many of those aspects have been dealt with in the traditional library sciences [79,80,81,82,83,84,85]. Despite the long-lasting history, the needs of today’s research landscape are different when it comes to the opportunities that modern publishing demands and that dissemination platforms could provide. Until the early 2000s, planetary maps had been used exclusively as reference products as there was no way to integrate data into new mapping investigations due to the lack of digital data and tools. Planetary maps were provided using analog material, map sheets in Portable Document Format (PDF), and, to a limited degree, webGIS visualizations. Maps are still being used as reference maps, but today, digital formats and modern workflows allow us to efficiently integrate data. The Astrogeology Team of the USGS has been experimenting with an online webGIS platform for selected planetary maps for many years during a time that few people used GIS in their workflows [86,87,88,89,90,91].

Today, for many new geologic maps the original map data are provided along with rich metadata (see, e.g., [29]) and thus become findable and accessible. The roles of map and metadata curation and stewardship are of importance to make map knowledge accessible to a wider community. Inserting maps into, and extracting maps from, an active research cycle are essential operations to gradually increase the amount of knowledge over time and build on existing foundations rather than having to redo work over again. To accomplish this, maps and their associated information need to be curated in a way that they can be located and actively used by researchers, i.e., maps need to be maintained and become “living” data assets that can move in and out of repositories. Research data frameworks cover the stewardship and curation of research data and exist in various incarnations (e.g., [92,93,94]).

A second challenge is concerned with interoperability. If digital maps are used for mapping, exploration, and potential resource exploitation rather than as reference documents in the future, they become invaluable research assets that are seamlessly integrated into workflows and derivative products. While a comparably large number of maps are being created, few of them serve as actual building blocks to create new research but serve as general references only, despite the availability of digital assets. Upon combining existing information with new information (additional data, new interpretation methods, or new models), new higher-level map derivatives can be produced. This way, maps can become a special kind of foundational asset, on which new data and knowledge can be built.

The third challenge is concerned with the transparency of the mapping process. Mapping aims at combining information and increasing knowledge gain through contextualization. By reducing subjectivity and bias, this gain will be increased and that would require including as many datasets as possible for an investigation. The reality is that map sources are often pre-selected based on pragmatic reasoning despite available access to a wide range of resources. This is likely due to the large volumes, variety, and complexity of available reference data and foundational datasets that make it less feasible and viable to integrate larger amounts of data despite significant differences in overall quality and internal integrity. It seems, therefore, reasonable to aim for obtaining detailed information about the chosen datasets on which mapping investigations are based, as well as on the datasets if they have been processed beyond standard processing, as there exists no validation for planetary geologic mapping based on remote sensing data.

1.2. Objectives

For this investigation, we will focus on methods, approaches, and implementations for the stewardship of maps, for increased interoperability and data transparency. We approach the aforementioned challenges from the direction of the research individual, i.e., the mapper or the investigator, who follows a set of objectives that include the investigation itself, its publication and dissemination, as well as the long-term use of the investigation as measured through citation feedback. If we assume that widespread visibility and research impact are key goals for a mapper, then the individual can design the research investigation and its dissemination in a way that visibility is increased and improved re-use of data is facilitated. To achieve this, the mapper is confronted with several choices regarding the preparation of data, the focus of the mapping investigations, and the selection of a dissemination platform.

Our first objective is to better contextualize the need for stewardship, interoperability, and the concept of transparency by visiting the FAIR principles [95] and the related indicator framework [96,97]. The concepts of making data assets findable (F), accessible (A), interoperable (I) and reusable (R) are specifically addressed in a community-driven initiative that brought the FAIR principles to wide attention [97,98,99]. This initiative is based on a seminal paper in which the principles of Findability, Accessibility, Interoperability, and Reuse (FAIR) of data and the need to improve data infrastructures are emphasized, in particular, for academic purposes [95]. These principles explicitly address the use of computer systems to help address the aforementioned principles.

To achieve that contextualization, we will need to assess the current level of planetary geologic maps concerning FAIR principles. We aim at developing an understanding of how these principles relate to planetary maps as digital research data assets, and how these principles might affect the value of such maps. We include a discussion on how FAIR principles could be achieved by different groups involved in mapping. This contribution thus explores how planetary maps can retain and multiply their scientific value by becoming objects that can be considered FAIR [95].

For our second objective, we will define and discuss the concept of transparency within the broader context of data provenance and the development process of maps, in order to identify important indicators which support transparency. This objective will include a discussion of how such transparency could be achieved on the level of the individual, as well as on the level of the publisher. Lastly, we will discuss the process of turning maps into FAIR data assets. In this process of FAIRification [98,99] we will discuss how various principles could be implemented. This will ultimately lead to a number of recommendations for the researchers and mappers regarding the transfer of their products to FAIR research data.

1.3. Structure

We will describe our methodological approach in Section 2, where we provide details on which basis we evaluate the current status of planetary maps and how we analyze existing indicators regarding their relevance to planetary thematic maps (objective 1), and in particular, with respect to map data transparency and, by extension, provenance (objective 2). We then provide details on how we approach the process of turning planetary geologic maps into findable, accessible, interoperable, reusable, and transparent data assets (objective 3). Results will be presented for each objective separately in Section 3. The discussion in Section 4 will provide a synthesis of the results and a contextualization of findings. A summary and conclusions, including recommendations, will be presented in Section 5.

2. Methods

The first objective comprises the characterization of FAIR principles [95] and their applicability and relevance for planetary thematic maps as digital assets. The FAIR principles are actively used and widely adopted within the research data community and thus represent not only a mature level of recommendations and guidelines but also provide use cases to build on. The Research Data Alliance’s (RDA) maturity model will constitute a core component in our characterization and discussion. This includes a critical review of the relevance of indicators for each principle as presented in the RDA maturity model report [96,97]. We discuss indicators for two different aspects of each mapping investigation: organization of mapping conduct and map publishing or dissemination. The first approach is labeled as a coordinated approach and is represented by institutionally organized and supported mapping investigations consisting of the coordination and publishing of maps and associated reports according to internal rules and review procedures.

For the planetary domain, the USGS is currently the only representation on an international stage with visible and widely accessible output. Traditional mapping examples are the series of geologic mapping investigations of the Moon from the Geologic Atlas of the Moon published in the 1960s and 1970s as individual map sheets with map scales between 1:5000 and 1:5,000,000. These maps were digitized from an originally analog map sheet format and published online. Despite their different approach at the time of creation, we include them here as they now form digital map assets that are included in data repositories. An example of a mapping investigation is the “Geologic Map of the Maurolycus Quadrangle of the Moon (I-695)” which was published in 1972 with a map scale of 1:1,000,000 [100], and which has been included as a digital asset in the USGS data repository. An example of a contemporary mapping investigation is the “Geologic Map of the Olympus Mons Caldera” published by the USGS in 2021 at a scale of 1:200,000. The map is accompanied by a mapping pamphlet and supplementary material as well as extended metadata [29]. The second approach comprises either individual mapping investigations conducted and published by individuals in ad-hoc mapping investigations, or group mapping investigations coordinated by an investigator for the course of a funded project. The results of both types of mapping investigations are commonly published in conventional journals, which may allow the cross-linking of material from research data repositories. They constitute a distributed form of map publishing with reports and related supplementary data spread over different repositories.

Therefore, two cases can be reviewed here: (1) maps integrated into the primary platform alongside the associated publication, and (2) maps published as supplementary assets cross-linked to a research data repository. Examples of such conventional publications are the DAWN geologic mapping investigations [35,45] which were a group mapping investigation coordinated for the duration of a project and published in a conventional journal. Other examples are mapping investigations that were published in a dedicated map journal in which the maps are cross-linked to a research data repository [56,57,58,59,60,61,62]. It is, furthermore, safe to assume that an organizationally coordinated mapping investigation will likely not be published in a conventional journal. Also, non- or temporally coordinated mapping investigations will likely not be published in an institutional publication series as a main outlet. We highlight the current status regarding FAIR principles and indicators, as summarized within the FAIR maturity model [96], for each of these contexts. As the indicator framework might change over time, we provide the underlying data table in the appendix for reference (

Table A1. FAIR data maturity model indicators; • useful indicator, $••$ important indicator, $•••$ essential indicator [96] (cf. Figure 3).

ID	Indicator	Priority
Findability
F1-01M	Metadata is identified by a persistent identifier	$•••$
F1-01D	Data is identified by a persistent identifier	$•••$
F1-02M	Metadata is identified by a globally unique identifier	$•••$
F1-02D	Data is identified by a globally unique identifier	$•••$
F2-01M	Rich metadata is provided to allow discovery	$•••$
F3-01M	Metadata includes the identifier for the data	$•••$
F4-01M	Metadata is offered in such a way that it can be harvested and indexed	$•••$
Accessibility
A1-01M	Metadata contains information to enable the user to obtain access to the data	$••$
A1-02M	Metadata can be accessed manually	$•••$
A1-02D	Data can be accessed manually	$•••$
A1-03M	Metadata identifier resolves to a metadata record	$•••$
A1-03D	Data identifier resolves to a digital object	$•••$
A1-04M	Metadata is accessed through standardised protocol	$•••$
A1-04D	Data is accessible through standardised protocol	$•••$
A1-05D	Data can be accessed automatically	$••$
A1.1-01M	Metadata is accessible through a free access protocol	$•••$
A1.1-01D	Data is accessible through a free access protocol	$••$
A1.2-02D	Data is accessible through an access protocol that supports authentication and authorisation	•
A2-01M	Metadata is guaranteed to remain available after data is no longer available	$•••$
Interoperability
I1-01M	Metadata uses knowledge representation expressed in standardised format	$••$
I1-01D	Data uses knowledge representation expressed in standardised format	$••$
I1-02M	Metadata uses machine-understandable knowledge representation	$••$
I1-02D	Data uses machine-understandable knowledge representation	$••$
I2-01M	Metadata uses FAIR-compliant vocabularies	$••$
I2-01D	Data uses FAIR-compliant vocabularies	•
I3-01M	Metadata includes references to other metadata	$••$
I3-01D	Data includes references to other data	•
I3-02M	Metadata includes references to other data	•
I3-02D	Data includes qualified references to other data	•
I3-03M	Metadata includes qualified references to other metadata	$••$
I3-04M	Metadata include qualified references to other data	$•••$
Reusability
R1-01M	Plurality of accurate and relevant attributes are provided to allow reuse	$•••$
R1.1-01M	Metadata includes information about the license under which the data can be reused	$•••$
R1.1-02M	Metadata refers to a standard reuse license	$••$
R1.1-03M	Metadata refers to a machine-understandable reuse license	$••$
R1.2-01M	Metadata includes provenance information according to community-specific standards	$••$
R1.2-02M	Metadata includes provenance information according to a cross-community language	•
R1.3-01M	Metadata complies with a community standard	$•••$
R1.3-01D	Data complies with a community standard	$•••$
R1.3-02M	Metadata is expressed in compliance with a machine-understandable community standard	$•••$
R1.3-02D	Data is expressed in compliance with a machine-understandable community standard	$••$

, Appendix A) to represent the status at the time of writing. We decided against using the statistical presentations suggested alongside the maturity model [96] for the reason that these assessments are built on the idea that data holders are actively working on the implementation of FAIR principles. This is likely not the case in the domain we are here reviewing and if it was, we would not be able to know the timeline of developments and the maturity level of implementations.

For the second objective of this investigation, we collected indicators related to what we consider to be representing transparency, i.e., indicators concerning the development process of maps and the provenance and lineage of data [96,97]. Some indicators are explicitly related to the concept of data provenance and describe data transparency, while others are indirectly related to transparency. Indicators and their respective priorities were then used to assess common planetary map assets based on the results in the previous section. This includes a review of current methods of collecting and storing lineage metadata, their potential application to planetary maps, and a discussion about implementations and control.

The third objective covers the process of creating digital map assets that can be considered FAIR and transparent. For the general case of digital data, this process is often referred to as the FAIRification process [97]. In the results section we will summarize activities related to indicators in the RDM maturity model by aggregating them in different activity groups that are related either to mappers or to publishers, or both. By doing so, we generate a pool of activity groups and activities that can be assigned to actors and which would facilitate the realization of FAIR principles for digital map data and associated metadata. In the discussion chapter, we will then contextualize these findings through published recommendations and procedures and adapt them to the case of planetary thematic maps to arrive at recommendations for individuals who conduct mapping investigations. We identify the project phases in which specific principles could be targeted, and discuss how processes can be designed to facilitate the creation of FAIR data and metadata for map products. Published use cases, experiences, and discussions of data FAIRification processes will help to better understand and adjust specific procedures. While the majority of discussions about such processes in the research literature are related to health data, they do share some of the complexity with compound map data [98,99,101,102,103,104].

3. Results

3.1. Status of Planetary Maps

In the first step, we converted the maturity model indicators [96] (see Table A1 for reference) into a relationship graph with an emphasis on the ternary relationship between data, metadata, and users or systems to assess where the data and metadata are co-dependent entities and how access to data and metadata are defined in the maturity model (Figure 3). In this graph, entities are represented by nodes, whereas actions and relationships are indicated as edges. Color and line thickness refer to the principle class and indicator priority, respectively. This depiction provides a more synoptic view of indicator relevance and relationships between (meta) data assets and users and systems and will provide a guide for the discussion of selected principle indicators in the next section. Although principles are characterized by different numbers of indicators and different priorities, we consider each of these four principles as equally important as they are all dependent components to ensure data can become FAIR. In addition to this ternary relationship, we looked into the specific properties of data and metadata assets to identify what are considered relevant properties for each asset group (Figure 3). The priority of properties and relationships are extracted from the same indicator table (Table A1). The reusability and findability of assets are predominantly related to the metadata group, as this also identifies access to data (map) assets. The value of the graph shown in Figure 3 becomes more apparent when comparing it with the use-case graph shown in Figure 4. The graph in Figure 4 was created by extracting activities from indicators that are targeted at making map assets more accessible (see also

Table A2. Topical relationships between activities of mappers/coordinators and publishers/research data platforms. Activities are symbolized as follows: ■ main contribution; partial contribution; □ no contribution.

Activity	ID	Fair	FID	Mapper	Publisher	Link
Provide Identification	1	F1	RDA-F1-01M		■
	2	F1	RDA-F1-01D	□	■
	3	F1	RDA-F1-02M	□	■
	4	F1	RDA-F1-02D	□	■
	5	F3	RDA-F3-01M	□	■
	6	F4	RDA-F4-01M	□	■
	7	A1	RDA-A1-01M	□	■
	8	A1	RDA-A1-03M	□	■
	9	A1	RDA-A1-03D	□	■
Achieve Compliance with Standards	10	R1.3	RDA-R1.3-01M	■	■
	11	R1.3	RDA-R1.3-01D	■	■
	12	R1.3	RDA-R1.3-02M	□	■
	13	R1.3	RDA-R1.3-02D	□	■
	14	R1.1	RDA-R1.1-02M	□	■	48
	15	I2	RDA-I1-01D		■
	16	I2	RDA-I1-01M		■
	17	I1	RDA-I1-02D	□	■
	18	I1	RDA-I1-02M	□	■
	19	I1	RDA-I2-01D	■	■
	20	I1	RDA-I2-01M	■	■
	21	A1	RDA-A1-04M	□	■
	22	A1	RDA-A1-04D	□	■
Provide Cross-References	23	I3	RDA-I3-01D	■	■	41
	24	I3	RDA-I3-01M	■	■	42
	25	I3	RDA-I3-02D	■	■	43
	26	I3	RDA-I3-02M	■	■	44
	27	I3	RDA-I3-03M	■	■	45
	28	I3	RDA-I3-04M	■	■	46
Provide Access Channels	29	A1.1	RDA-A1.1-01M	□	■
	30	A1.1	RDA-A1.1-01D	□	■
	31	A1.2	RDA-A1.2-01M	□	■
	32	A1	RDA-A1-02D	□	■
	33	A1	RDA-A1-02M	□	■
	34	A1	RDA-A1-05D	□	■
	35	A1	RDA-A1-04M	□	■
	36	A1	RDA-A1-04D	□	■
Provide Lineage Information	37	F2	RDA-F2-01M	■
	38	R1	RDA-R1-01M	■
	39	R1.2	RDA-R1.2-01M	■
	40	R1.2	RDA-R1.2-02M	■
	41	I3	RDA-I3-01D	■		23
	42	I3	RDA-I3-01M	■		24
	43	I3	RDA-I3-02D	■		25
	44	I3	RDA-I3-02M	■		26
	45	I3	RDA-I3-03M	■		27
	46	I3	RDA-I3-04M	■		28
Provide License Information	47	R1.1	RDA-R1.1-01M	■	■
	48	R1.1	RDA-R1.1-02M	■	■	14
	49	R1.1	RDA-R1.1-03M	■	■

for a different view of relationships). Nodes refer to actors and objectives, while edges highlight the action to reach these objectives. These actions are related to specific model indicators and are characterized by a FAIR principle, priority, and the actor, i.e., either mapper/coordinators or publishers. It can be seen that the majority of necessary actions are directly related to the publisher or publishing platform to turn data and metadata into FAIR data assets. Furthermore, while a substantial number of indicators require both mappers and publishers to be active, nearly half of all indicators are exclusively connected to the publisher, while there is no single indicator that is connected only to the mapper. These indicators are mostly related to the way data and maps are integrated into the publishing platform and refer to the first two principles: findability and accessibility. When it comes to cross-references, compliance with standards as well as transparency, contributions from the mappers’ side become paramount as these refer to the interoperability domain. This graph will allow us, in the next sections, to identify crucial indicators and activities related to provenance information (Section 3.2), and it will provide information about tasks to create FAIR digital map objects (Section 3.3) from existing, i.e., completed, mapping investigations as well as for new investigations during the planning phase.

Based on the indicators we can identify the following major activities (cf. Figure 4).

• Provision of identification is a group of activities that cover the majority of indicators related to the findability (F) and accessibility (A) principles. The core of this group is of a technical nature and is connected to the publisher and dissemination platform to ensure that data and metadata can be found and accessed.
• Provision of transparent information relates to the need to make data, data sources, and data development transparent, in particular, as no validation exists for planetary thematic maps. Transparency in this context refers not only to the map’s development but also to establishing a connection between mapped features and the data that were used. These can be single images from one sensor, multi-temporal observations from one sensor, but also multi-temporal observations from multiple sensors under different observation and illumination geometries. This group covers aspects of all four principles.
• Compliance to standard formats is a group of tasks that refers to how (meta)data are represented emphasizing the use of community standards, representations, and established vocabularies. This group is mainly covered by interoperability (I) and reusability (R) principles.
• Provision of cross-references is an activity group that comprises cross-referencing data and metadata, including internal cross-references, as well as cross-references to external (meta)data. The former is related to the findability (F) principle mainly, while the latter is related to the reusability (R) principle.
• Provision of access channels refers to various protocols and interfaces enabling access to (meta)data. This group is represented exclusively by the accessibility (A) principle.

These major activity groups are not exclusive and various activities can refer to different groups (cf. reference principles in Figure 4).

After extracting these visual representations we looked into the USGS repository and supplements of journals from publishers that are regularly publishing planetary geologic investigations, i.e., Elsevier via ScienceDirect, John Wiley & Sons, Inc., via the Wiley Online Library, and Taylor & Francis via Taylor & Francis Online. We aimed to obtain an overview of the general status of published geologic maps regarding FAIR principles rather than conducting a representative study of published material, which would have also involved more publishing platforms. For the reviewed published material we checked to which level indicator objectives are met, and if mapping investigations and maps constitute assets that are FAIR. The results are shown in

Table 1. Implementation of FAIR principles in different mapping and map dissemination platforms. Symbol meanings: □ not implemented, partially implemented, and ■ completely implemented or implementable.

Principle	Indicator	Institutional Platform [29]	Conventional Journal Platform [45] (Internal Supplement)	Journal and Repository Platform [105] (External Supplement)	Archival Platform (External Supplement)
F1	RDA-F1-01M	□	□	□
F1	RDA-F1-01D	■	□	■
F1	RDA-F1-02M	□	□	■
F1	RDA-F1-02D	■	□	■
F2	RDA-F2-01M		□	□	■
F3	RDA-F3-01M	□	□	■	■
F4	RDA-F4-01M	■	□		■
A1	RDA-A1-01M		□	■
A1	RDA-A1-02M	■	□	■	■
A1	RDA-A1-02D	■	■	■	■
A1	RDA-A1-03M	□	□	■	□
A1	RDA-A1-03D	■	□	■	□
A1	RDA-A1-04M	■	□	■	■
A1	RDA-A1-04D	■	■	■	■
A1	RDA-A1-05D	■	□	■	■
A1.1	RDA-A1.1-01.M	■	□	■	■
A1.1	RDA-A1.1-01.D	■	■	■	■
A1.2	RDA-A1.2-02.D	■	■	■	■
A2	RDA-A2-01M	■	□	■
I1	RDA-I1-01M	■	□		■
I1	RDA-I1-01D	■	□	■	■
I1	RDA-I1-02M	■	□		■
I1	RDA-I1-02D	■	□	■	■
I2	RDA-I2-01M		□		■
I2	RDA-I2-01D		□		■
I3	RDA-I3-01M		□	□	□
I3	RDA-I3-01D	□	□	□	□
I3	RDA-I3-02M	□	□	□	□
I3	RDA-I3-02D	□	□	□	□
I3	RDA-I3-03M	□	□	□	□
I3	RDA-I3-04M	□	□	□	□
R1	R1 RDA-R1-01M		□	□	■
R1.1	RDA-R1.1-01M	□	□	■	□
R1.1	RDA-R1.1-02M	□	□	■	□
R1.1	RDA-R1.1-03M	□	□	■	□
R1.2	RDA-R1.2-01M	□	□
R1.2	RDA-R1.2-02M	□	□
R1.3	RDA-R1.3-01M		□	□	■
R1.3	RDA-R1.3-01D	■		■	■
R1.3	RDA-R1.3-02M	■	□	■	■
R1.3	RDA-R1.3-02D	■	□	■	■

.

The majority of indicators related to accessibility and findability are implemented for institutional platforms as well as research data repositories that are linked to journal platforms. For conventional journal publishing platforms, maps appear as images in supplements and can neither be found nor accessed due to the lack of metadata information and cross-links. In the interoperability domain, institutional platforms and research repositories perform comparably well. It is noteworthy, however, that all platforms largely fail to establish cross-references to other data and metadata. The reasons for this are unknown and we have to speculate that it might be related to the need to establish semantically valid connections which require deeper insights into the topic. A similar picture is also provided when looking at indicators in the reusability domain. The institutional publishing platform achieves similar scores when compared to journal publishing in combination with data repositories. Publishing in conventional journals by making use of publishing maps and map data through the publisher’s internal dissemination platform constitutes the least complete option.

3.2. Transparency and Data Provenance

Due to the lack of validation in planetary geologic mapping, transparency becomes paramount to communicating findings and mapping decisions in a way that mapping decisions are traceable for the reader. In this investigation, the concept of data transparency refers to the data provenance of a compiled map product as well as to the status and origin of each contributing dataset. While conventionally, data provenance refers mainly to tracing the development stages and modifications of an object, we here include not only the map’s provenance but also a reference to the data that have been used at the time of mapping explicitly. In Figure 4 we highlighted the indicators focusing on aspects of data transparency and provenance and thus the development of data under “provision of lineage information” (see Table A2). We specifically identified interoperability indicators (I3) as well as indicators covering reusability (R) and findability (F2) principles. The majority of these indicators refer to the mapper, while publishing platforms have a contributing role. RDA-F2-01M refers to the provision of rich metadata, which is a necessary criterion for transparency. The interoperability indicators refer to the representation of data and, in particular, to cross-references and qualified references between data assets. As mapping investigations build on various data sources, these sources need to be known and specified, which would be covered by this indicator group. The reusability indicators RDA-R1.2-01M and RDA-R1.2-02M explicitly refer to provenance information according to the standards used in the community and across those boundaries.

However, as there are no standards for provenance information in the domain, this issue requires a more in-depth discussion. In some map publications that we have reviewed, provenance information has been interpreted as references to published literature, which, we feel, does not completely cover the intended purpose. As seen in Table 1, the aforementioned indicators are not covered by any dissemination platform when it comes to interoperability principles covering the cross-referencing of data assets. For provenance-related indicators (RDA-R1.2) only repositories may document versions in their metadata and this only applies to the map object itself, not to the contributing data.

To summarize, the activities related to data provenance in a wider sense comprise defining rich metadata, establishing qualified and non-qualified cross-references, as well as contributing provenance information in the metadata.

3.3. Towards Reuse of Map Data Assets

In our model view, we look at two actor groups that contribute to making data FAIR objects: (1) mappers alongside their associated coordination instance, and (2) publishers and distribution platforms. Both actors provide information attached to digital map data as well as to the associated metadata objects (Figure 5). Based on the FAIR principle indicators, we identified the following major activities which we separated or aggregated according to thematically similar concepts. They are displayed as individual nodes while edges connect the respective actor group with each of the nodes according to their activity domain. This way we can detach these indicators from the more general FAIR structure, and associate activities related to these indicators with actors directly, thus highlighting their potential responsibilities during the mapping investigation.

• The activity group “provision of identification” (Figure 5) refers to adding information to data (maps) and map metadata to make these digital assets findable by computer systems and by standard tools used by the community. The majority of indicators referring to findability fall into this group. Identification is also required for allowing access to data assets and thus partially covers the accessibility principle. These tasks are assigned to the publishing platform due to its technical nature.
• The activity group “compliance with standards” covers interoperability and reusability principles mainly. In particular, when it comes to the provision of information in support of the interoperability principle, both actor groups are required (Figure 5).
• The activity group “provision of cross-references” is entirely covered by interoperability indicators, in particular, for principle I3 which requires such references. For this activity group, both actors are equal contributors as the thematical cross-references need to be provided by mappers, but storage and efficient use must be facilitated by the publisher. These indicators also cover aspects of data provenance which are listed as an additional activity (Figure 5).
• The group “provision of access channels” comprises activities exclusively related to indicators covering the accessibility principle. As such they are in their entirety covered by activities from the publisher’s side as that actor needs to provide the appropriate access channels (Figure 5).
• The group “provision of transparent information” covers aspects related to the storage of information concerning the source and development of the map data product as well as all associated meta-information and data products as summarized in the previous section. This group of activities comprises interoperability indicators that are concerned with cross-referencing data and metadata, as well as reusability and findability indicators. In the following chapter, these elements will be discussed in a wider context (Figure 5).
• Lastly, the group “provision of license information” is related to the reusability of data and is, therefore, equally covered by the mapper who provides the reuse license and the publisher, who facilitates accessing that information and who adds publisher-specific license information (Figure 5).

In the following chapter, we will discuss how and when the aforementioned activities can be implemented within a mapping investigation.

4. Discussion

In this chapter, we will provide a synthesis of our results presented in the previous chapter and expand on the covered concepts. In the first step, we will develop this discussion along a generic higher-level mapping investigation that starts with a (1) conceptual phase in which the cartographic and publishing concepts are defined, continued by (2) implementation or mapping phase and terminated by the (3) dissemination phase during which map data, metadata and associated objects are published and inserted into a research cycle. Along this path, we highlight activity groups as discussed in the previous chapter which contribute to the FAIRification process and compare individual approaches to the published literature. Here, in particular, we focus on the activity group related to the provision of transparent information (Figure 5).

4.1. Mapping Phases

The conceptual phase (Figure 5) is the pre-mapping phase and comprises the definition of mapping-specific aspects, such as map topic, map scale, map extent, map sheets, target audience, procedural definitions, and data use. This phase also involves defining how mapping is conducted during the mapping phase, which data to process and incorporate, and how the final product(s) might be disseminated on a higher level during the dissemination phase. This coordinating activity thus involves definitions of data, as well as related metadata, contents and presentation formats. This stage might also include the selection of publishing outlets, depending on the project and its focus.

Figure 5. Associations between activity groups (gray circles, see Figure 4), associated FAIR principles and mapping phases (black circles). The depicted relationship highlights the required activities as represented by activity groups to ensure the development of FAIR data assets within each mapping phase.

Figure 5. Associations between activity groups (gray circles, see Figure 4), associated FAIR principles and mapping phases (black circles). The depicted relationship highlights the required activities as represented by activity groups to ensure the development of FAIR data assets within each mapping phase.

If data are to be inserted into a research data repository with the aim to be FAIR, this stage requires awareness about community standards and how to comply with standards when it comes to data and formats (Figure 5 (CS)). This might also include re-considering the mapping platform that is used to allow better access options. For GIS-supported mapping, the choices at the time of writing are relatively limited and while some platforms provide tools for rich metadata definitions and compliance with standards, other platforms are more cost-effective and offer availability on different platforms. At this stage, mechanisms need to be established for collecting cross-references for later metadata synthesis (Figure 5 (CR)). This will also contribute to the collection of information about data transparency (Figure 5 (LI)).

For the storage and streamlining of the mapping investigation, a suitable data model could be defined at this stage. This not only allows for providing containers for information and constraints to keep the mapping investigation consistent, but it also allows for efficient versioning and maintaining the integrity of the mapping database. With a targeted definition of such a data model, the consistent maintenance of metadata information and cross-links to external information become more efficient but also more restrictive. During the conceptual phase, the definition of data and information as well as workflows are exclusively covered by the mapping and coordinating site, but information about the publishing platform will be needed to decide about specific formats and requirements.

The implementation and mapping phase (Figure 5) consists of the actual mapping investigation in which various datasets are used to visually determine geologic contacts as well as material units, processes and stratigraphic relationships. Based on a mapped set of contacts and geologic structures, as well as material units, geologic units can be defined according to sets of geometric and topological rules. During this mapping phase, mappers can establish and record the connection between a particular mapping decision and the associated external dataset on which the decision is based. These connections may be covered during the provision of cross-references that have been established during the conceptual phase (Figure 4 (CS)). This association, however, is commonly not enforced in mapping today. While mappers may communicate the exact data basis on which they made each decision in associated documentation, it is commonly not conducted consistently. This, in conclusion, means that any form of indirect validation of mapping results will be missing. In order to improve that situation, we consider it important to include exact references and cross-references to the exact dataset on which a set of decisions was made, and to also include that dataset in an accessible repository with cross-links to the map and investigation. Often, much effort has been invested into creating a consistent data basis for mapping. This includes the selection of image data, but also processing, filtering and mosaicking. Orthorectified image mosaics require considerable effort and while these datasets provide a valuable mapping basis, they are commonly not published due to a lack of options.

Practically, this could be implemented by, e.g., associating a mapped contact of a geologic unit with the dataset identifier on which the contact was observed. If different datasets cover the same contact, or if the contact covers multiple images from multiple sensors, such a relationship can easily be implemented without introducing exceedingly high workloads on the mapper. One potential implementation could consist of a frontend in which datasets are checked using an interactive selection of image labels. On the backend, a relation with image labels and related metadata can be stored in the mapping database and can be made available during the mapping process (Figure 6a,b). During this mapping phase, transparent information will be provided (Figure 4 (TI)) which is composed of rich metadata, along with cross-references and data lineage information (Figure 5).

The dissemination phase (Figure 5) is the most important phase in terms of assembling and providing metadata and preparing data assets for FAIR and transparent use. This phase is predominantly supported by and relies heavily on the publisher once a dissemination platform has been selected by the mapper. Activities, such as providing access channels (AC), providing identification details (ID), licensing (LC), staying compliant with standards (CS) and also the degree to which data and metadata can be provided in the most transparent and traceable way (TI) depend almost exclusively on the publishing platform and its implementations (Figure 5). When it comes to the provision of licensing information and information regarding data transparency, mappers and coordinators are required.

Conventional journal platforms covering the planetary domain do not provide means to update contributions and supplements which might be related to the traditional style of paper publishing where research has been treated as a news article, but not as a developing piece of information. The post-publication versioning concept is today implemented in several data journals and for research data repositories where updates can be introduced whenever feasible. This dichotomy in publishing approaches spawns an entirely different discussion in a highly transformative publishing environment. The reason for us to mention this concept is to increase the awareness of mappers and coordinators that, in order to make research traceable and transparent, but also widely accessible and living research assets, the selection of the right publishing platform becomes a crucial element. Mappers usually have control over the choice of publication outlets but not over the way how data and metadata are implemented on these platforms. By knowing a publisher’s approach towards FAIR principles, mappers can make an informed decision that balances the publication impact and reusability value of their research. It needs to be added, however, that by providing FAIR and more transparent research, it does not necessarily become openly accessible, and it is equally feasible to publish a paywalled mapping investigation while the map is freely accessible via an open-access research repository [106].

4.2. Processing FAIR Map Assets

The process of developing mapping investigations for FAIR and transparency needs to be separated into two possible paths that can be taken. One path includes the entire conceptualization and mapping design with the a priori objective of developing FAIR and transparent map assets, while the other path involves procedures of turning map assets and investigations FAIR after publication. The general workflow can be broken down into several phases and procedures. A detailed breakdown is given by [98] which includes three distinct phases that also involve an assessment after publishing data. An important aspect of this overall procedure is that the process of FAIRification requires expertise from different fields [98,107]. In the same way that cartographic expertise is needed to support the geologic mapping investigation conducted by a team of geologists and geomorphologists, the conceptual modeling and building of a semantic data model and metadata model might require specific expertise in the field of research data management.

During the conceptual phase of the mapping investigation, it is feasible to consider and discuss the FAIRification objectives [98]. This is followed by an analysis of data and metadata which are used for the underlying FAIRification process. The next step includes creating the semantic data and metadata models if they are not readily available and provided as FAIR objects for the specific domain. In the domain of planetary mapping, these models are, to the best of our knowledge, not readily available but current efforts inside the community and from within mapping organizations have developed at least partial implementations. In the next steps, data and metadata need to be transformed and digitally linked with respect to the semantic models, and hosted on a suitable platform.

The situation is less straightforward if a mapping investigation has already been published. Conventional publishing platforms do not allow for updating information in papers except for rare cases where errata are published. Versioning is not a viable concept as it involves considerable effort to maintain and update publications and a different approach to citing bibliographic references. While the written mapping investigation might be a static dataset, associated data stored in data repositories can commonly be versioned. Making digital maps and data available at a later point is possible, although referencing these data might not be possible anymore. While digital data assets stored in research repositories might not be referenceable from within the published research paper anymore, digital data assets can refer to these publications and at least provide a lasting unidirectional link to satisfy basic FAIR principles. Publishing data even though the text-based mapping investigation has been published at an earlier point in time, seems a viable option to support a more transparent and sustainable research-data management.

If maps and data were created initially without FAIR principles in mind, turning them into FAIR data assets might become challenging, and while it might not be possible to satisfy all indicators towards FAIR principles, a considerable level of FAIR data can still be achieved by separating data assets from the mapping investigation report, as long as conventional publishers cannot provide inclusive outlets. The most important tasks towards this objective on the mapper side include (1) choosing a reliable data repository; (2) metadata enrichment to improve the findability of maps and data; (3) data linking for increased transparency and establish a link between publications and data; (4) employing standard data formats to improve interoperability; (5) adding data documentation highlighting data structure and context to improve reusability and (6) the provision of license information. However, the retroactive transfer to FAIR data assets might not be as effective as taking an approach that initially targets FAIR principles within the mapping conceptualization phase.

4.3. Mapping Transparency

In the planetary mapping domain, and in particular, in thematic mapping, mappers and map users need validation. Planetary mapping relies almost exclusively on the visual inspection and interpretation of surface units that were imaged under geometric and observational constraints. Thus, each image is a snapshot of a complex situation that is being observed from only one vantage point. While it seems simple to discard that need as we cannot simply generate validation by going into the field, other ways need to be found to become closer to some sort of validation or increasing the credibility of thematic maps. The costs and benefits may be debatable, after all each thematic map represents only one potential interpretation of the geologic settings of an area and as of now, there is no direct economic value connected to such an investigation. And yet, first of all, geologic mapping investigations are research products that need to adhere to scientific standards. Secondly, the economic value may be indirect as areas of special interest may be selected as future landing sites and interpretations may largely differ depending on personal focus and the data that are used for mapping [69,108,109,110,111,112,113,114]. Past landing-site studies have thus revealed considerable discrepancies in interpretations and the Mars Exploration Rovers have clearly documented the misinterpretation of landing sites through in-situ observations. While for terrestrial geologic mapping, unit delineation, and characterization are commonly documented by photos and field samples, planetary geologic mapping investigations are often constrained to a set of characteristic remotely-sensed images presented in the investigation report. While the inclusion of data from various imaging sensors seems to be a less biased approach and potentially leads to more representative and robust results, mapping decisions become less transparent if geologic units and the underlying data are not linked.

While it seems feasible for the purpose of unbiased mapping to include as many datasets as there are available, it is usually not viable in terms of efficiency and time needed to create accurate datasets and to map larger areas based on different datasets with relatively similar quality characteristics. Therefore, maps need to become "transparent" assets that allow tracing their development and using data basis precisely. Such an approach is not new. While it is common to include supplementary maps detailing certain selection criteria and image characteristics in publications accompanying reference mappings (e.g., [115,116]), that information becomes largely irrelevant for the thematic mapper as the mapping basis has been established already at that point. With that, the data basis and its completeness and integrity are usually not questioned again. While that reference mapping basis may be of high quality and well documented, its creation did not specifically target the needs of thematic mappers.

This leads to the conclusion that while data basis may still be valuable, the full potential has not been exploited. Instead, most of these maps have been assembled based on aesthetic criteria, i.e., exhibiting a similar resolution, similar radiometric properties, and similar observation and illumination geometries. For geologic thematic mapping, this is not necessarily all that is needed. The thematic map is a final product of its contributing components, i.e., image data, and the quality of a mapping investigation consequently depends on the quality of the available input data and thus cascades down to aspects of image acquisition and sensor characteristics. It seems, therefore, straightforward to include information about the mapping base in the final thematic map product. This is commonly conducted as information put on the map sheet or as a note within the accompanying publication, i.e., a research paper or the geologic investigation pamphlet. But for digital map sheets, and in particular, for data provided as downloadable GIS data packages, this information is not consistently made available.

Data provenance, data lineage and trust are closely linked and in research data management, they generally refer to the need to document the origin, versions and development path of a data asset with all its stages [95,107,117,118]. It must be noted that the concept of trust we refer to differs from the RDA TRUST principles framework [119]. We here refer to knowledge about a particular data asset, such as a map, including its underlying data which were used to develop the map. By linking map objects to the underlying data directly, and by making used data available as data products, better traceability is granted and map users may not only better understand mapping decisions, but could also make use of that secondary data. We would, therefore, use the concepts of data provenance and lineage wider and include references to participating data assets including their versions and development stages. These aspects are, to our understanding, at least partially covered in the FAIR principles (Figure 4 and Figure 5 (TI)) and could be incorporated directly into mapping models (Figure 6).

5. Conclusions and Recommendations

5.1. Summary

Planetary maps play an important role as exploration tools as they visually and semantically compile different pieces of spatial and non-spatial information to provide a rich base for knowledge extraction. Geologic maps, in particular, allow for extracting such knowledge about spatial configurations and allow us to relate that information to time and processes. As such, maps require careful stewardship as they serve as valuable input for decision-making and for building future exploration strategies. The scientific value of such maps, however, is hard to assess if a map’s development is not well documented and if the data basis is ambiguous or not well known. It could be shown that thematic maps, after all, are only as good as the quality of their individual inputs. A clear communication of how and on which basis a map was compiled is, therefore, of utmost importance to better assess its validity and accuracy, and to understand the novelty of a new mapping at a later point in time. The value of such maps is also limited if they cannot be accessed and if they cannot be incorporated into future investigations.

With the advent of digital map compilations and GIS data for download, the way thematic maps need to be distributed has been changing significantly. Findability, Accessibility, Interoperability, and Reuse (FAIR) are keywords in the context of transparent research data management. The FAIR framework provides ample resources to allow turning digital thematic maps into FAIR data assets to be used by future generations of researchers and mappers. This investigation also discussed the relevance of planetary maps as FAIR digital assets in order to provide not only findable, accessible, interoperable and reusable data products but also maps whose development process is transparent and traceable at all times. This transparency is needed, especially right now as there exists no validation for thematic maps, except for a few local spots on the Moon and Mars. This situation will change eventually in the future and better ground truth will become available and serve as a validation tool, but validated global coverage is far away. We believe that this justifies the emphasis on data and lineage transparency.

Firstly, maps are largely subjective abstractions built on image representations that were selected on different and not always coherent criteria. Secondly, maps are generalized and simplified views on a topic whose emphasis is decided by the mapper, not by objective criteria. This includes how much simplification the mapper allows, but it also includes which data the mapper selects.

Maps are important tools for the planetary community and for understanding the evolution and characteristics of regions of interest. Beyond the local spatial contextualization of various data sources and interpretations, maps provide a means to build global synopses through the temporal and spatial correlation of spatial units and temporal events. Finding and accessing existing maps and being able to integrate/work and reuse them are crucial steps in mapping processes. Without limiting research freedom, well-designed mapping data models can be fundamental tools to facilitate the creation of FAIR maps and allow for a better review and revision process as they provide more transparency into which data have been used.

5.2. Recommendations

Making digital map assets findable, accessible, interoperable, reusable (FAIR) as well as transparent requires the adjustment of procedures throughout all mapping phases, i.e., (1) conceptualization phase, (2) implementation phase, and (3) dissemination phase (Figure 5,

Table 2. Project phases of mapping investigations and associated activity groups (cf. Figure 5). □ not realized in that phase, partially realized, and ■ completely realized.

ID	TASK	Project Phase
		Conceptual Phase	Mapping Phase	Dissemination Phase
1	Provision of identification	□		■
2	Compliance with standards	■	■	■
3	Provision of cross-references	□	■	■
4	Provision of access channels	□	□	■
5	Provision of transparent information		■
6	Provisions of license information		□	■

). Some information is required from the mapper, in particular, when it comes to data formats, (meta) data links and scientific content. Other information, such as contextual metadata related to the investigation is required from the investigation coordinator if that agent exists. Information related to finding and accessing data is required by the publisher, or, the underlying infrastructure provider. As long as no coordinated and synchronized mapping approaches and infrastructures have been established or have even been defined, the following recommendations might allow to turn existing map and map metadata, as well as new projects, into reusable assets that follow guiding FAIR principles. With the added value of transparency as defined through richer provenance metadata, mapping investigation may then become more accessible and reusable, and also better verifiable.

Definition of Directions: While many tools can be implemented to realize sustainable research data management once researchers and institutions are onboard, the long-term strategic direction needs to be defined first. Once a critical mass of supporters is found, the definition can be developed by the community as a bottom-up approach. Other planetary and terrestrial initiatives related to data management and development of infrastructure were initially defined by the community in the past and a number of technical and research notes have been published on this topic in recent years. In a similar way, once a practical approach has been tested and defined through community interaction and iteration stages, and once the concept has been proven to work, funding sources might become available as agencies attempt to push toward improved research data management to make spending processes more transparent. With these two processes developing in the same direction, we believe it will become much easier to motivate the larger research community to participate and to find additional incentives for their participation. An initial focused definition phase, along with an appropriate testing phase would create a transparent process that can be joined by other community members. The Research Data Alliance (RDA) might be one platform to initiate such a development, but we believe that the existing venues such as the annual Geologic Mapper Meetings and the existing targeted Europlanet initiatives have the required means to initiate such discussions and develop directions.

A large challenge will be the definition of incentives and the communication of the added value for researchers to participate in such a process, as it demands additional time and effort. We do not have any solution at this stage as we imagine that this motivation must originate intrinsically. However, we believe that by providing additional information and by disseminating data through accessible research repositories, the initial work will receive more views, more citations, and wider transparency and distribution, which in the end is an important cornerstone of the publishing process. If that effect can be shown, it might bring the much-needed incentive for members of the community to join.

The following recommendations will focus on potential tools to implement FAIR+T principles on institutional and individual levels.

Transparency and Extended Provenance Information: Well-designed mapping data models will help to keep track of used data and to relate mapping findings to a specific dataset on which certain findings have been located. The definition of a map data model is part of the conceptual phase. Its main application is in the implementation and mapping phase. It will contribute to a higher level of transparency of contributing data origins and data use, as well as improve data integrity when it comes to mapping geometries and topologic relationships. It will directly link to indicators such as RDA-R1.2-01M and RDA-R1.2-01D (provenance information) as well as RDA-I3-01M, RDA-I3-01D, RDA-I3-02M, RDA-I3-02D, RDA-I3-03M and RDA-I3-04M (data references). The USGS has recently published a mapping data model guideline for planetary mapping [7] which could, for example, easily be expanded to accommodate provenance information to a much greater extent. Implementations have also been published by a number of terrestrial surveys over the last decades [120,121], and a discussion of an early planetary data model configuration was provided in [122].

As maps and associated metadata information will largely be available in digital form only in the future, more dynamic visualization and distribution platforms are needed to bring digital maps to the community. This also applies to maps designed and distributed in the planetary domain. In order to facilitate this, a cartographic concept needs to exist that explicitly targets the needs of the anticipated user community. For research use, principles should be discussed starting from the conceptual phase to provide a targeted and reusable map product that can easily be located by computer systems.

Dissemination and Visibility: Choosing a dissemination platform that allows the storage of standardized data and metadata information for maps in accordance with FAIR principles would solve the challenges related to the FAIR principles’ findability and accessibility. While dedicated map journals are a straightforward solution in which reduced map images and associated investigations are directly linked on one platform, it is also conceivable to separate the textual part of the mapping investigation from the digital map and data assets. The latter could be published on a variety of research data platforms and repositories that also allow the linking and updating of data and metadata across different platforms. While less coherent and straightforward, it provides a larger range of options regarding the choice of journal outlets. As long as cross-institutional standardization has not been employed, individual mappers as well as collaborative mappers could benefit from directly adapting principles to their mapping investigation which would not only allow future reuse and potential recompilation of their work but also encourage citation of their work. An important step towards this is making maps findable and accessible which can already be managed with limited effort on the metadata level. This initial step can be made by any researcher at any stage of the research process.

Interoperability and Workflows: Interoperability needs to be considered as an opportunity by mappers when working with digital map data as it allows the integration and recompilation of existing work. No paper map can be made interoperable in the end, and no printed map sheet could be a part of the process of research data reuse and only serve as a potentially citable but otherwise not re-usable reference. This opportunity opened by digital map data requires additional steps to be implemented starting during the conceptual phase. This happens on the metadata level but also on the structural semantic level of data. If carefully designed in the conceptual phase and with the potential user community in mind, collaboration, integration and re-compilation of research insights will become much faster and reliable.

Ready-to-Use Mapping Models: The community needs ready-to-use mapping models as long as there is no centralized and long-term coordination platform. Conceptionally sound mapping data models would not only help to implement FAIR principles more efficiently, but it would also allow to put constraints on the mapping investigation where constraints are needed to maintain a certain level of quality, e.g., by topological integrity checks and domain controls, without limiting the mapper’s freedom in conducting her or his research. Using such constraints would allow for a high degree of comparability between mapping products but it will not make the mapping experience more flexible. Also, version control becomes easier, and testing is more agile. Past projects demonstrated that streamlining and standardizing ad-hoc mapping investigations are challenging, in particular, when project time constraints are putting pressure on the investigation. Extended individual mapping freedom will lead to imbalances in the appearance and consistency of maps and map data, respectively, and the priority for such investigations will shift to documentation, rather than to consistent map production. Such models have been proposed in the past and are currently also developed and employed by the USGS. The adaptation of such approaches during the conceptual phase might increase the conceptual workload but will turn out to be highly efficient during error checking, control and final map (sheet) compilation.