Skip to end of metadata
Go to start of metadata

Mapping Standards Document DRAFT 17 April 2018 0 - updated on wiki since then


David Rothery & Matt Balme, The Open University  initial edits.

List of Acronyms

AcronymExplanation


FGDCFederal Geographic Data Committee
PGMPlanetary Geology Mapping (Protocol)
USGSUnited States Geological Survey
GISGeographic Information System


1. Introduction

writing task update by matteo.massironi

Globally accepted standards of planetary mapping have been developed over several decades (since 1961) by the USGS in consultation with the global community. In order to avoid unnecessary divergence from these, we will conform as closely as possible to USGS standards and practice. 
However, as we progress into an era of online digital and multi-layered products there may be situations in which innovation or change is desirable and situations in which is unavoidable. For this reason we have grouped the PLANMAP products into USGS standard-like maps and thematic maps which can not strictly follow USGS standards (PLANMAP non-standard mapping products). Any departures from or elaboration of USGS practice should be avoided wherever possible in USGS-standard like products, and in any case be discussed within the PLANMAP team before we commit to them.

1.1 Guide to this document

This document provides links to and a brief commentary on the USGS standards. These standards are inherited from an era when the aim was usually to publish a single geological map of a given area, whereas PLANMAP aims to provide users with multi-layered digital maps. Section 2 states that we will adhere to conventional symbology and conventional units, but perform our mapping in ways that make it easy to present different versions or different layers in an informative manner. Section 3 elaborates some examples of types of contact, crater classification, plains classification, superficial deposits, and spatially small features of interest. Although the standards in Section 3 are intended to be generally applicable, the specific examples referred to are for Mercury; this section therefore concludes with remarks on how to apply the standards to the Moon and Mars. Section 4 is dedicated to PLANMAP thematic maps which, although based on USGS standards, must depart from these to allow their specific thematic objectives (i.e. structural geology, spectral-stratigraphic mapping, landing site characterization, in situ mapping and geometric 3D modelling). Finally, Section 5 describes the publication agreement about authorship of PLANMAP products.

2. Existing USGS standards in relation with PLANMAP activities

Existing protocols, guidelines, and symbology were recently updated and summarized in USGS reports Mapping Protocol (2018) and FGDC Planetary Symbology, both of which are downloadable as PDFs from https://planetarymapping.wr.usgs.gov/Page/view/Guidelines
The symbols outlined in the symbology guide are downloadable as ArcGIS templates via the same URL.
The Mapping Protocol (2018) has a useful glossary of definitions (section 0). The Introduction (Section 1) stresses that planetary geologic mapping is continuously evolving (Hansen, 2000; Nass et al. 2011, van Gasselt and Nass 2011, Nass et al. 2017). As such, PLANMAP will play a role in steering the evolution of innovative new map products using the rich and abundant data collected during recent planetary missions. Section 2 contains a disclaimer that it does not seek to 'explain the intricacies of planetary geologic mapping' and cites some external references for those needing guidance regarding the mapping process itself (e.g. Chapter 2 of Wilhelms 1987, http://ser.sese.asu.edu/GHM/). It also re-stresses the need to be 'flexible and adaptable' when producing new maps and related products. Section 3 (USGS Maps) describes some USGS procedures for NASA-funded maps. These do not directly concern PLANMAP, as our 3-year funded timescale forces us towards publishing products at a faster timescale and with a less formal process. However, we are still be guided by some of the bullet points, especially:

  • The map will not contain excessive interpretive detail that is not required to establish and convey the geologic framework (e.g., excessive figures or hypothetical discussions, etc.).
  • The map will be an accurate, concise, and clear representation of the geology of the selected region at the Publication Map Scale, regardless of the resolution afforded by the Primary Map Base. An accurate depiction of the geology at full data resolution has the potential to result in linework that is far too dense to be legible at the selected Publication Map Scale. As such, it is critical for the Author to understand that a USGS Standardized Map is a representation of the geology that is discernible and representable at the selected Publication Map Scale.

Section 3 (USGS Maps), also defines the difference between non-standardized maps and USGS standardized products - the first described as "flexible, exploratory products that can more easily respond to data influxes and innovation", whereas the second are "rigid, framework products that place exploratory results in a larger, more comprehensive structure".  This distinction is pivotal for PLANMAP consortium, because some of the PLANMAP products aimed at innovation will unavoidably diverge from USGS standards. However, PLANMAP map-derived products aim to be not only flexible and exploratory, but also scientifically objective, and thus must closely follow the USGS standards. This document indeed aims to define the general constraints on non-standard products to ensure their rigorous production and cartographic normalization.

Section 4 (Roles and Expectations) is a procedural discussion that does not apply to us directly. However, PLANMAP as a whole has come to an agreement about the 'authorship' of each product which is explained in section 5 of this document. 
Section 5 (Map Package Components) begins by referring authors to recent examples of similar maps. It names GIS Files (Section 5.0) and the Geologic Map (Section 5.1) as items to be published, and mentions procedures for nomenclature. We recommend that PLANMAP maps should not normally show names that do not have IAU approval. Exceptions may be needed (but should be used sparingly) in the case of large scale landing site maps (following precedents from the Apollo program – such as Shorty crater, which at the time of the Apollo 17 was a purely informal name). The rest of Section 5 describes what information needs to accompany a map. Some of this (at least the key to units and symbols) must go on the map sheet itself. Other items (such as a summary of the geologic history) may go more comfortably in what is referred to as 'an accompanying map pamphlet' – in our case this could be an accompanying paper(s) in Journal of Maps or other peer-reviewed journals, or as a PDF document made available online.
Section 6 (Mapping Process) is concerned mostly with the USGS formalities, few of which are directly relevant to PLANMAP. However, please note Section 6.3 (Digital Mapping), which suggests digitising linework at a scale that is 4 times larger than the intended publication scale, to avoid overly detailed linework (which can be a waste of time to draw as well as sometimes obscuring the nature of contacts). PLANMAP mapping activities will follow this general rule as well as the Tobler (1987) one, which is understood to mean that the largest permissible mapping scale is 1:(R x 2000), where R is the raster resolution (pixel size) in metres of the dataset used for photogeological mapping. Note also the USGS Technical Review Guidelines (Section 6.6.1), which sets standards by which we should be guided in preparing maps and their descriptions (Section 5).
Section 7 is references, Section 8 lists useful websites, and Section 9 lists USGS support personnel and facilities.

3. PLANMAP USGS-like standard maps

PLANMAP is operating in an era of multi-layered digital products. There is no longer a constraint to produce a single, definitive, printed map to portray the geology of the chosen region.
It is the responsibility of WP7 to collate the products from earlier WPs and to turn them into science-friendly and out-reach friendly products. The job of the mappers is to provide products that enable WP7 to offer the end-user some (a small number of) helpful choices.
Here follow recommendations for PLANMAP, noting the circumstances where it is likely that we may diverge from basic USGS practices.

3.1 Basic principles

Using any GIS system, map elements are composed of attributed symbols. For planetary mapping, these symbols will usually be multi-vertex lines, points, or polygonal areas. Commonly, these are referred to generically as 'features', 'objects', or 'shapefiles' – a collection of computer files containing many individual objects of a certain type. Each of these objects can have many attributes stored within the associated shapefile – usually as a string of text or a number (floating point, integer etc). Thus map objects such as lines can be given attributes that contain names, short acronyms, lengths, latitudes, elevations, etc. Attributes are generally visible to the user as a table within the shapefile, formatted as one row per object and one attribute per column. It is common for contacts between units to be digitised as line objects, and units themselves to be digitised (usually semi-automatically in the GIS software) from these contacts. Contacts are most often stored with few attributes (maybe just an ID number and contact type string), whereas units are often given multiple attributes (e.g. ID number, unit name, acronym, area, formation, crater density, etc.).
In a GIS it is easy to add an extra column in an attributes table at an early stage in mapping. It is therefore simple to enact a later decision to use common, identical, symbology to display features (lines or areas) that were identified during mapping with different attributes. This choice can be made at a very late stage. Conversely, the reverse (subdividing features with identical attributes into two or more display types) cannot be done without re-examining and re-defining each example.
Therefore, please plan for flexibility.

Section 3.2 draws particular attention to Mercury, but the principles there apply to ALL bodies. Sections 3.3 and 3.4 note additional items (or exceptions) relevant to the Moon and Mars.

3.1.1 Visible background (transparent foreground)

Since the early days, lunar maps (e.g., Wilhelms & McCauley, 1972) and the USGS Mariner 10 geologic maps of Mercury (referenced in Section 3.2), were printed on top of an airbush mosaic background, so that the topography could be discerned behind the foreground colours. MESSENGER-era geologic maps of Mercury (Galluzzi et al., 2016; Mancinelli et al., 2016; Guzzetta et al., 2017) take a similar approach except that they use a seamless monochrome image mosaic as the background. This is common practice in other planetary maps too. It is usually helpful for the user to be able to see the topographic texture.

PLANMAP recommendation

  • Extensive units mapped in colour should usually be presented with a small degree of transparency, allowing a suitably-chosen 'background layer' (such as a seamless monochrome mosaic, or 'hill shaded' digital elevation model) to show through. However, line work and small symbols should be solid (i.e., non-transparent). In the case of a map presented digitally (i.e., not printed) it should be possible for the user to manipulate the degree of transparency (or even to choose a different background layer).

3.2  USGS-like standard maps -  general standards, exemplified mainly by Mercury

Compared to the Moon and Mars, Mercury currently has no complete global geologic maps, due to the limited imaging coverage provided by the Mariner 10 mission (e.g., https://astrogeology.usgs.gov/search/map/Mercury/Geology/Mercury_5M_GIS_conversion_v2). New geological maps are currently being generated using data from the recent NASA Messenger mission, and are particularly relevant for ESA's BepiColombo mission, with a planned launch in fall of 2018.

Based on the ongoing experience of PLANMAP team members on the Mercury mapping effort, here we describe cases where PLANMAP mapping may diverge from USGS standards. Some features, such as faults, are relatively non-contentious and should be shown with the obvious USGS symbology, and conform to the style of previous recent maps of the body in question. Discussion here focuses on example situations where there is a stronger need for guidance.

3.2.1 Contacts, including 'gradational contacts'

Section 5.2.2 of the Mapping Protocol (2018) gives the following guidance regarding contact types:

The quality of contacts varies considerably on most maps. Contact types need to be defined and used as consistently as possible for a given map. There are technical and philosophical issues that arise from consistently using different kinds of geologic contacts within a single geologic map, particularly when mapping is based solely on remotely-sensed data sets. In general, (1) a "certain" contact is used when a contact is known to exist and is confidently located; (2) an "approximate" contact indicates that the contact is presumed to exist but its location is not confidently identified at Digital Mapping Scale (perhaps due to data quality and (or) obscuration by an overlying surficial unit); (3) a "concealed" contact indicates that surficial material buries the contact but morphologically the contact is still traceable, although subdued; and (4) an "inferred" contact, which may be used to delineate map units where the validity of the map unit or distinction between the units is hypothetical. "Gradational" contacts have been used in past maps, though this contact should be reserved for cases where other contact types are insufficient and, even in such cases, should be applied sparingly (if at all). Beware also that there are places where a wavy or tortuous contact as mapped at large scale appears gradational at small scale. As long as the Author is explicit in defining the contact types and consistent in applying such definitions throughout the map, the Author has flexibility to employ a reasonable geologic contact scheme, though such schemes will be subject to review.

PLANMAP may encounter situations in which it makes sense to mark a contact as gradational. An example would be a boundary between two plains units defined by colour (spectral properties), such as 'red plains' and 'blue plains' where the distinction between the two is arbitrary and the mapper believes that 'red' transitions into 'blue' gradually (see section 4.1). The meaning of gradational contact is subtly different to approximate contact. An approximate contact is where the mapper is certain that there is, in nature, a definite limit (such as the outer edge of continuous ejecta around a crater) but the mapper cannot be certain exactly where this line occurs. Neither of these cases is quite the same as an inferred contact or a concealed contact.

PLANMAP recommendation.

  • Distinguish between approximate and gradational contacts when mapping (remember, they can always be grouped together at a later stage).

The recommended linework for showing a gradational contact is FDGC symbol 1.1.21 (Figure 1), from https://ngmdb.usgs.gov/fgdc_gds/geolsymstd/download.php (this is of course easy to change, if we decide to)



Figure 1: Gradational contact linework (use 1.1.21) from https://ngmdb.usgs.gov/fgdc_gds/geolsymstd/download.php


For convenience of comparison, Figure 2 shows the relevant linework for other types of contact.



Figure 2: Contact linework (25.1-25.4) from FDGC Planetary Symbology via https://planetarymapping.wr.usgs.gov/Page/view/Guidelines

3.2.2 Crater classification

Degradation state

At some scale, and on some bodies, it is common practice to distinguish craters (and their ejecta) by degradation state. In particular, Mercury quadrangle mappers in the Mariner 10 era found it useful to distinguish five such types (for craters > 30 km), ranging from freshest (youngest) to most degraded (oldest) (McCauley et al., 1981). However, the mappers of the first quadrangle (H02) to be mapped using MESSENGER data (Galluzzi et al., 2016) found when applying this scheme to craters >20 km that there were sometimes contradictions between the implied relative age of adjacent craters based on degradation and the relative age shown by cross-cutting or superposition relationships. Because their intention was to make a morpho-stratigraphic map, they reduced the number of crater classes to 3, which removed the contradictions. Maps of H03 and H04 did the same.
However, the group of ex-MESSENGER team members working on a 1:15M global map of Mercury (Kinczyk et al., 2016; Prockter et al., 2016), classified craters >40 km into essentially the original 5 classes (though numbered 1-5, young-old, which is the reverse of the original scheme).
Mapping of quadrangles H05, H10 and H14, underway at the Open University, is proceeding using BOTH a 5-class and a 3-class scheme (this simply requires an extra column in the ArcGIS attributes table).


Planmap recommendation.

  • On all newly started Mercury maps, craters should be classified according to BOTH a 5-class and a 3-class scheme. This keeps options open for WP7 to offer a stratigraphic map or a version that shows crater degradation state. The end-user therefore has a choice.
 Crater chains

Note that FGDC indicates how to show chains of craters or catenae (25.110) or a pit-crater chain (25.105), by means of simple linework drawn to scale (Figure 3).
 


Figure 3: Recommended symbology for chains of crater (25.110 and 25.105) from FDGC Planetary Symbology via https://planetarymapping.wr.usgs.gov/Page/view/Guidelines. There is essentially no difference between these two.

Planmap recommendation.

  • Since on Mars and the Moon collapsed lava tubes are clearly distinguishable from crater chains they can be classified with the same symbol but with different colors.

3.2.3 Plains types (and crater-floor deposits)

 By age

Mercury has previously been mapped with 3 age classes of plains (smooth plains, intermediate plains, intercrater plains). Mappers will be aware of a suggestion from some members of the MESSENGER team that the intermediate plains unit should be dropped (Whitten et al., 2014). Intermediate plains is absent from the Prockter et al. (2016) 1:15M map, but 1:3M maps published or in preparation retain all three units.

PLANMAP recommendation.

  • On all newly started Mercury maps, plains should be mapped using at least 3 age classes.
  • Where the location of a contact between two plains types is uncertain, it will usually be appropriate to map this as an 'approximate' contact (not a 'gradational' contact).

 By location

It has been the practice of some mappers to subdivide plains by location as well as by morphology/age, notably to distinguish the plains interior to a particular basin from plains of similar age exterior to the same basin. For example Mancinelli et al. (2016) and Guzzetta et al. (2017) each mapped the plains interior to Mercury's Caloris basin as a Caloris smooth plains unit, even though by morphologic (though not necessary spectral) criteria these are the same as the smooth plains exterior to the basin.

By colour (spectral properties)

Plains of identical morphology may be subdivided by colour according to what is indicated in section 4.1.

By context or textural content

In some settings, such as crater-floor deposits, and for some investigations, the mapper may wish to use geomorphic principles to distinguish units of likely different origins, as for example possible impact melt sheets on the floor of the lunar crater Tycho (Krüger et al., 2016).

PLANMAP recommendations.

  • During initial mapping, plains (and crater floors) should not be subdivided merely by location, although in a later stage this can be recommended when attributions to different feeding systems (i.e. major basins on the Moon and Mercury) or different spectral units (see section 4.1) are obvious.
  • Initial mappers may recognise or infer internal contacts within a plains (or crater-floor) unit. In such a case they should certainly map this, and flag any area within a plains unit that could be separately mapped, to aid decision making at a later stage.

3.2.4 Superficial deposits

Some thin deposits overlie older units, and are recognised by albedo or colour, but are too thin to obscure the texture of the older, underlying unit. Crater rays are one example. On Mercury another obvious example is faculae (or 'red spots'), which are usually interpreted as explosive pyroclastic deposits. In order to present as much information on our maps as possible, the chosen ornament for such a unit should be some kind of stipple (or other transparent ornament) that enables the underlying unit to be shown (see Section 3.6 for how to indicate examples that are too small to show their shape at the published scale).
FGDC 25.118 and 25.119 are recommended for dark-coloured and light-coloured ejecta respectively (Figure 4), and should be used for crater rays. FGDC 25.129 is recommended for light coloured mantling material, and should be used for faculae (Figure 5). On maps where it is desired to draw special attention to, or to make apparent, these features it may be useful to use a coloured version of these patterns. Generally use red for faculae, but black for rays. Note that 25.129 and 25.119 look very similar if shown in the same colour.



Figure 4: Recommended symbology for crater rays (25.118 and 25.119) from FDGC Planetary Symbology via https://planetarymapping.wr.usgs.gov/Page/view/Guidelines 



 Figure 5: Recommended symbology for faculae on Mercury (25.129) from FDGC Planetary Symbology via https://planetarymapping.wr.usgs.gov/Page/view/Guidelines 

PLANMAP recommendation.

  • Use a stipple to show units that do not obscure the underlying unit.

3.2.5 Fields of hollows

Some small patches on Mercury are scarred by 'hollows', which are steep sided, flat bottomed depressions, usually surrounded by a high albedo blue halo. Individual hollows are 100s of metres across, but fields of hollows can be 10s of km across. Galluzzi et al. (2016) and Guzzetta et al. (2017), but not Mancinelli et al. (2017), used a blue ornament similar to FGDC 25.135, which we recommend (Figure 6).



Figure 6: Recommended symbology for fields of hollows on Mercury (25.135) from FDGC Planetary Symbology via https://planetarymapping.wr.usgs.gov/Page/view/Guidelines

3.2.6 Spatially small features

Some features (cones, knobs, small faculae, small field of hollows) are too small to show in detail at the published scale. If they are likely to be of interest, they should be mapped as points distinguished by an appropriate entry in the attributes table. Symbols on the final map should be based on FDGC Planetary Symbology, or previous European precedents, whenever possible. 

PLANMAP recommendation.

  • Use points to identify features of interest whose actual shape is too small to show at the publication scale.

3.3 USGS-like standard maps - specific recommendations for the Moon

Due to the Moon's proximity to the Earth, lunar mapping began in earnest with the development of the telescope in the 17th century. However, with the American Apollo and Soviet Luna programs in the 1960s and 1970s, a significant effort was made to produce maps to guide the selection of landing sites, as well as to understand the geological evolution of the Moon (e.g. https://www.lpi.usra.edu/resources/mapcatalog/). Recent USGS efforts have led to an updated 1:5M global lunar map (https://astrogeology.usgs.gov/search/map/Moon/Geology/Lunar_Geologic_GIS_Renovation_March2013). This contains updated linework and map boundaries using new topographic data and image mosaics, which are adjusted to the Unified Lunar Control Network 2005 (ULCN2005,  https://pubs.usgs.gov/of/2006/1367/). This morpho-stratigraphic map, based on photogeological observations using mapping guidelines described in Wilhelms (1987, http://ser.sese.asu.edu/GHM/) and references therein, serves as an important basemap for producing more detailed modern lunar maps using data collected by recent lunar missions. Such photogeological stratigraphic maps are combined with other datasets, such as spectral data, topography, rock abundance, etc. (see Section 4) to allow the separation of additional geological units that can then be investigated using techniques such as crater size-frequency distribution measurements.

For example, Ivanov et al. (2018, in review) present a new 1:500K stratigraphic map of the northeastern portion of the South Pole-Aitken basin, which includes five crater age classes, four plains-forming units, and basin materials associated with Orientale, Apollo, and South Pole-Aitken basins. Each defined unit was analysed using crater size-frequency distribution measurements to provide independent evaluation of the ages of the stratigraphic units. Ivanov et al. (2018, in review) applied the mapping techniques as described in the USGS guidelines and by Wilhelms (1987). Continued mapping activities in the scope of PLANMAP for the Moon will follow these guidelines, unless there are specific cases where exceptions are required, as discussed above for Mercury.

3.4  USGS-like standard maps - specific recommendations for Mars

There are many published Mars maps, many performed as part of the USGS Mars global, regional, and quadrangle-mapping frameworks. As for the Moon, the USGS recently renovated the Scott and Carr (1978) global map for use in GIS (https://astrogeology.usgs.gov/search/map/Mars/Geology/Mars_SIM-1083_Mariner_9_GIS). The poles above and below 60 degrees are mapped at a scale of 1:25M, whereas the rest of the planet is mapped at 1:5M. The USGS has digitised many other martian maps (https://astrogeology.usgs.gov), including combining three 1:15M maps into a single renovation (https://astrogeology.usgs.gov/search/map/Mars/Geology/Mars15MGeologicGISRenovation), and researchers have also self-published regional and local martian maps (e.g., Bernhardt et al., 2016). These follow USGS rules and regulations and provide good examples and inspiration for smaller-scale mapping of Mars. Hence, mapping of areas of Mars at 1:1M scale or smaller, should follow the same general principles as the Moon and Mercury. Always follow USGS precedents, unless there is a good case for diverging.

Non-standard Mars maps can vary in complexity from simple sketch maps provided to give context in published papers, to survey datamaps showing locations of specific landforms (e.g. Ramsdale et al. 2017), to more formal geomorphological maps of non-standard scale, orientation or shape.

In particular, both the weathering regime and the sedimentary processes on Mars as well as the complex interactions among diverse volcanic complexes and lava fields may require special symbols or units that make them different to Mercury or Moon maps. At large scale, the remote sensing data available for Mars allows such a diverse variety of landforms and surface types to be identified that providing generic advice on unit and structure is difficult. Examples such as delta landforms with multiple channel and lobe components, various styles of aeolian cover, landslides, probable glacial landforms and glacially modified terrain are found alongside more 'traditional' planetary features such as volcanic plains and impact craters. Maps can focus on bedrock, cryotic materials, or surficial materials ("drift" maps), for example.

We suggest that larger scale maps may require more significant departures from the FGDC symbology and USGS methods. These should be made with due regard to conforming, where appropriate, to precedents on any earlier comparable maps. What is necessary in all cases, though, is sufficient information to describe the projection, scale, location and orientation of the map, and supporting information about the units and symbols used, and what their interpretations are.

Overall, the level of detail and "formality" should be chosen based on the desired outcome of the study. The changeable nature of the martian surface means that the finest scale mapping efforts may reveal superficial and short-lived morphological features. Transient CO2 ice deposits, dusty slope streaks, "araneiform (spider-like)" landforms and surface areas with high coverage by dust devil 'tracks' all provide challenges to a mapper, and change with the martian seasons (Figure 7). Determining what should and what should not be included in a large scale martian morphological map is just not as simple as "digitising what you see in the images". Fortunately, GIS can again be used to display images or basemaps from different years or seasons (e.g. Erkeling et al. 2016, Heyer et al. 2018; http://muted.wwu.de/), thus providing guidance as to what should be usefully recorded in the context of a given mapping project.


Figure 7: Interpretation of small, possibly seasonal, features on Mars, as seen in MOC, HiRISE, or CTX images (three rightmost panels), require the geological context provided by lower resolution datasets (e.g. HRSC, left)(from Erkeling et al., 2016).

4. PLANMAP non-standard mapping products

PLANMAP aims at providing some innovative geological maps that can depart at various degree form the USGS standards. In this section are listed various PLANMAP non-standard products, their main focusses and their potential divergence with the USGS-protocols.

4.1 Integrated geo-spectral and geo-stratigraphic maps

francesca.altieri  cristian.carli

Compositions or spectral information are not included routinely in planetary geological mapping outputs. Neither spectral properties nor composition are considered in the Planetary Geologic Mapping Protocol 2018 (PGM2018), nor are they integral to the planetary geologic features of the FGDC. In particular, in the latter are shown only symbols for dark-colored (25.118) or light-colored (25.119) ejecta and low albedo smooth materials (25.121). Occasionally, indication of color variations are also included. The Lithologic Patterns panel (FGDC 37.1) shows monochrome patterns for lithologies encountered during terrestrial mapping, but these are suitable mainly for stratigraphic columns rather than maps. These can rarely be applied (at least with any certainty) on Mercury or the Moon, though some of the sedimentary symbology may be useful for Mars.
The use of colors, as symbol, is the only suggestion present in the PGM2018 to introduce some discrimination between different materials “may be used to reflect the group they are in (e.g., warm colors for volcanic materials, cool colors for sedimentary rocks, yellows for crater materials, browns for ancient highland materials)”. So there are very few suggestions from USGS standards for integrating compositional information into the planetary mapping. PLANMAP products will explore how to taken into account lithological/spectral information in planetary geological maps. However,  the integration of compositional information on the maps will be an ongoing PLANMAP activity that should be updated during the project.

At a beginning stage the following different situations can be envisaged (see Giacomini et al. 2012 as partial reference, Figure 8) :

1) A given morpho-stratigraphic unit of the USGS-like standard maps can be clearly associated with a compositional/spectral unit.

PLANMAP recommendation

  • a contact symbol and a specific color have to be added, the compositional information should be defined in the stratigraphic legend.

2) A well defined morpho-stratigraphic units (for instance plains, or wide lava fields) can be characterized by a spectral variability indicative of lithological variations.

PLANMAP recommendation

  • integrate this variation within the geological map using specific unit contacts and diffeent tone of the same colour, the spectral/compositional information must be described within the stratigraphic legend.

3) A given spectral unit can correspond to different contiguous geomorphological units or morphostratigraphic units.

PLANMAP recommendation

  • if this variation can be potentially integrated within the map  with a transparent ornament, the spectral/compositional information must be described within the stratigraphic legend;

4) Outcrops or thin deposits characterized by the same spectral properties of a sourrounding unit.

PLANMAP recommendation

  • "a color saturation”can be used in order to highlight bedrocks from deposits, as suggested in PMG2018 for B/W maps; spectral/compositional information must be described in the stratigraphic legend.

5) Outcrops or thin deposits can be characterized by different spectral properties than a surrounding unit.

PLANMAP recommendation

  • implement this information within the map with a distinctive contact, transparent ornament or dedicated color, spectral/compositional information must be described in the stratigraphic legend.

The above-mentioned variety of examples highlight that, once all the spectral units have been identified for each studied case, a revision and an homogenization of colours and/or symbol will be likely needed in the timeframe of the project.


Figure 8: Excerpt from Giacomini et al., (2012). a) THEMIS Day-IR mosaic of Daedalia Planum lava flows, on the south-western side of Arsia Mons (Mars). b) Geologic map with different lava flow units (D1 to D13) distinguished by their morphology and stratigraphic position. c) Morpho-stratigraphic map superposed to the SAM (Spectral Angle Map) classification based on OMEGA (Mars Express) datasets; spectral differences due to surface morphology, mineralogy and textures inside the lava units are shown. The black/grey portions were not considered in the analysis, not being related to the lava field. Id) Geo-spectral map chachterized by the integration of morpho-stratigraphic and spectral units. 


Geo-spectral maps will be the base for specific geo-stratigraphic maps where the spectral units whose clear stratigraphic position as been inferred from cratering excavation, relative and absolute model ages or exposed walls, are favored with respect to the USGS standard like morpho-stratigraphic units (such as craters of different degradation stages,  and undifferentiated  crater floors and ejecta) (Figure 9). Such maps will be essential for 3D geological models reconstructing the geological volumes underneath the surface.




Figure 9:  Example of different geological maps of a portion of Rembrandt basin. a) Messenger color mosaic of the Rembrandt basin area; b) Morpho-stratigraphic map; c) Geo-stratigraphic map.

4.2 Geo-structural maps  

A geo-structural map is a geological map in which all the features able to explain any deformational event that affected the mapped area are properly highlighted. This means that fault geometries and associated  kinematics must be specified whenever possible and with a proper symbology and fold types (synclines vs anticlines) and geometries (symmetrical, asymmetrical-verging, recumbent, plunging etc.) must be well recognizable by dedicated symbols and an adequate number of strata measurements (when strata are available). In addition a particular attention should be paid to cross-cutting relationships among faults and folds.  Although not always required in USGS standardized geological maps (see PGM 2018), geological cross-sections are mandatory for any geo-structural map and must be almost perpendicular to the major mapped structures in order to enable their subsurface representation. Finally rose diagrams and stereo-plots reporting measurements of faults, folds and strata are highly recommended. This kind of map is rare in planetary geology, but dedicated mapping symbols are provided by the USGS FCDG document and the USGS map of western Candor Colles by Okubo (2014) can be taken as a reference for planetary surfaces (Figure 10).


Figure 10: Excerpt from the geo-structural map of Candor Colles (Mars) from Okubo, (2014). The symbology, showing fold axis, fractures, faults, fault zones, unconformable contacts, and strata orientation, is adopted from FGDC.

4.3 Geo-modelling maps

The production of three-dimensional subsurface models requires the creation of specific structural and stratigraphic mapping products with the specific aim of providing constraints that can be directly used within geological modelling software packages. In planetary geology the data are normally limited to the surface (we lack underground data such as wells or seismic surveys) and can, in most of the case, be represented in the form of maps. The geo-modelling maps must emphasize the structural relationships between geological bodies that are expected to be reconstructed through geological modelling. This can be done by providing two dimensional geometries that will be coupled with a three dimensional DEM surface to produce true three-dimensional representations. The mapped geometry must be coupled with attributes that unambiguously describe the geometrical relationship among geological bodies in a consistent way, by providing unique identifier codes for each type of feature.

Three-dimensional geologic modelling is made by providing all the mapped observations (i.e. geologic contacts between units domains, limits and structures) to the modeling software . On the basis of these constraints the software will provide a first interpretation of the subsurface geometry, that will be refined depending on the robustness of the interpretation, the amount of available data and the specific modeling approach. In general, the mapped features should specify geologic characters that are almost certain. The types of feature that will be object of mapping and can be used to derive constraints for geological models are summarized in the following table:


GEOMETRY TYPEDESCRIPTIONUSED FOR
POINTFeatures that must be consider as point-like for the scale at which the mapping is made

Specify dip and strike of a planar feature (e.g. unit contacts, sedimentary bedding, faults, etc) at a specific location, e.g. derived from DTMs (Digital Terrain Models) and/or rover data analysis and DOMs (Digital Outcrop Models)

Highlight outcrops or other features that cannot be mapped as areal features at the scale of the mapping and establish their pertinence to a given stratigraphic unit

Highlight a location where the thickness of a geo-stratigraphic unit is well constrained for example by cratering excavation or direct exposition.


LINEFeatures characterized by line-like appearance on the mapping medium

Any linear feature derived from the intersection of a geological plane with the surface can be used to obtain attitude observation for the feature of interest, for example:

Bedding-plane traces  can be used to constraint the orientation of the the bedding along the polylines

Mapped faults can be used to constrain the geometry of fault planes in the subsurface

Unit's boundaries provide constraints on the orientation and the thickness of compositional and stratigraphical units

POLYGONAreal features

Specify the pertinence of a large area to a stratigraphic/compositional unit and, where possible, define its thickness from its cratering record (e.g. Hiesinger et al., 2002; Ferrari et al., 2015)

Identify exposed planar features as for example exposed bedding planes (i.e. structural terraces in a layered terrain) that can be used to derive a strike and dip observation (e.g. Massironi et al., 2015; Penasa et al., 2017). This geometry associated with a point measure, where possible, will provide more constraints to the modeler.

Table 1: Vector topology and their applicability to 3D mapping and modelling (see also van Gasselt et al., 2011; Naß et al., 2017)

4.4 Landing site maps (maps for future landing sites and traverses)

Landing site maps will follow the standard defined here above for regional maps. However, they will also include information relevant to the higher resolution of data used for mapping, such as data acquired by in-situ probes when existing. In particular, these differences will include:

  • Information relevant to landing site selection such as lanslide ellipses, slope-gradients, and mapping of boulders, and outcropping and cover materials.
  • Subdivisions of units at a higher level of details than usually considered in regional maps, e.g. members of a unit with a given lithology, but identified to be different based on local texture variations, local facies differences (truncation, cross bedding, etc.) , local composition (orbital or in-situ), on erosional properties, etc.
  • Subdivisions of surface textures related to local observation by high resolution images or by the rover, i.e. variations in regolith type, variations in eolian feature types, variation in rock txture types etc.
  • Local in-situ information from a rover along its traverse enabling the identification of different facies, texture, composition of rocks, etc.. Note that the difference in resolution between ground-based and orbital data will generate an heterogeneity of scale on the whole map, which is unavoidable because it is necessary to provide the most relevant data.
  • Eventual suggestions on potential traverses

For purpose of landing capabilities or rover trafficability, geological maps of landing sites may integrate more information on surface texture, rock distribution, dust index, roughness maps, etc. , than usually provided in geological maps (e.g. Golombek et al., 2012; Ivanov et al, 2015; Potts at al. 2015) (Figure 11, 12). Those will nominally be represented as independent layers in a GIS environment, but may locally be merged with information from traditional geological maps such as in geological units with specific texture or roughness (Pajola et al., 2016).

The list above is not exhaustive and will be refined during the course of the project.


Figure 11: Potential landing sites of interest are evaluated by producing local geomorphological maps (left) and comparing these with other datasets such as slope maps (right). In this case for Boguslawsky crater on the Moon, which was a potential Luna-Glob landing site, boulder counts were performed for the locations marked with the small white boxes in the map to investigate potential hazards to a lander (from Ivanov et al., 2015).


Figure 12. In situ maps would be developed from rover traverse maps made with orbital data, for example for a proposed site in Schrödinger basin on the Moon, where LS = landing site and the numbers represent stations (from Potts et al. 2015). (A) LRO WAC mosaic with NAC image overlays, (B) Geological map of Kramer et al. (2013), (C) M3 data (Kramer et al. 2013), and (D) slope map.

4.5 In-situ maps (maps with integrated ground observations)

In-situ maps are developed at the scale of rover traverse or lander surroundings; i.e. at scales below 1:50,000. These maps are created from the merging of orbital, aerial (when existing) and in-situ data, thus responding to a different logic than maps composed of pure orbital data in which facies cannot be defined as they can in the field (Figure 11, 12). The definition of geological units and members is obtained from visual inspection of images in 2D and 3D, at various scales (including texture from high resolution images at mm-scale) as in field work on Earth. However, a difference between field work and  in-situ rover observations is that the latter has variable capability (i.e., elemental chemistry vs X-Ray Diffraction mineralogy), implying that:

  • Differences in the instruments available on probes (and in the data that have been able to be acquired eventually) will create variations in the definition of geological units from one map to the other, or one mapped region to the other in the same mapped site. These differences are intrinsic to the in-situ measurements and cannot be avoided easily.
  • In-situ data are distinct in scale and types from orbital data and thus may create differences when merging maps from orbital and in-situ data. In this case, in-situ data will be the ground-truth over which the map will be built primarily.
  • The degree of precision of in-situ data enables the definition of more types of units than can be done in orbital data. E.g., more than one type of regolith may exist, sedimentary units may be defined from their facies, incl. texture at mm scale, contacts in a vertical scarp not visible from orbital data, etc.
  • The degree of precision of data will help local structural measurements such as layer dips, fault plunges, etc., enabling the map to be more specific than orbital maps.

The information provided in GIS layers will be included in Virtual Reality environments by tools such as partly transparent layers added over the existing 3D landscape. Tools developed in Virtual Reality will also help to develop the mapping itself, for instance by helping the identification of stratigraphic contact on scarps not visible from a vertical perspective.

The list above is not exhaustive and will be refined during the course of the project.



Figure 13: Landing site maps generated from orbital data can be updated and refined using ground observations and sample information, for example this new geological map for the Apollo 12 landing site (Iqbal et al. 2018). This map uses the LRO WAC mosaic, LRO NAC images, M3 data, and Clementine data to define mare basalt units and count areas for crater size-frequency distribution measurements. The ages derived from the orbital data are comparable to the radiometric ages of the mare basalts sampled by the Apollo astronauts (Iqbal et al. 2018). The "unconfirmed contacts" represent spectral boundaries observed in M3 data.

4.6 3D / Subsurface models

Three-dimensional geological subsurface models are a numerical representation of geological features of interest in a particular area. The three dimensional representation is obtained by means of meshed surfaces of geological structures as fault/fractures surfaces, bedding planes, stratigraphic horizons and contacts of any kind. The meshed surfaces are the result of a modelling process that takes into account all the available geological constraints provided by geomodelling maps, and integrated into the modelling software (Figure 14). 3D geological models in this sense constitute the sum of all geological investigation, summarizing the available knowledge into a numerical representation that can be updated with new constraints whenever they become available. The choice of the modelling approach is driven by the availability of constraints and data for the geological body of interest.


Figure 14: excerpt from Calcagno et al., 2008, showing the process of generating three dimensional subsurface models (b) from geomodlling map constraints (a), that can be used to derive geological cross sections (c).


Modelling methods are often subdivided in either explicit or implicit approaches. Explicit methods require the manual definition of the geological surfaces by interpreting cross sections of the study area that are then hand-digitised and interpolated on the whole domain. On Earth explicit models make use also on borehole intersections with geologic bodies and geophysical constraints such as seismic sections, as an additional subsurface information to the one derived by geological, geo-structural and geo-stratigraphic mapping. On planetary bodies impact craters can be used to constrain the locations, thicknesses and positions at depth of specific units, this acting as substitutes of boreholes.

This kind of approach is possible only when a large amount of good-quality data is available. Implicit methods (Cowen et al., 2011, Calcagno et al., 2008, Cowen et al., 2003) are instead particularly suitable in the context of planetary mapping because they allow three-dimensional models to be produced by using the constraints provided by goemodelling maps, with minimum operator intervention. Being implicit models not based on interpretative geological cross-sections, they need many more constraints on the orientation of planar geological features (i.e. strata, gological contacts, fault planes, fold axial planes etc.)

Figure 15 show the increasing information needed for the different 3D geo-modelling approaches


Figure 15: Products with different degree of landscape geological knowledge needed for  the different 3D geo-modelling approaches. Note that the explicit approach potentially needs more information than the implicit one, although the latter requires more constraints on the orientation of planar geological features. 

5. Publication Policy

PLANMAP and Horizon 2020 grant agreement No 776276 must always be acknowledged in any kind of products derived from the consortium activity. In particular all the geological maps must report the PLANMAP logo and the following acknowledgments to the European funds: "This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 776276-PLANMAP".

The authorship of any PLANMAP publication will respect the normal procedure for scientific publications without particular attention to consortium representatives. This means that neither the PI nor the WP leaders will automatically be present as authors on every publication.

The main authors will appear first (not necessarily in alphabetical order), followed by the secondary ones in alphabetical order. With first authors we intend the ones directly involved in the production of geological maps or giving relevant scientific advice on features interpretation. The secondary authors are the ones that have assisted with their technical expertise or provided basic data-sets or given minor but substantial scientific suggestions. In any case the contribute of the secondary authors must be relevant for the final mapping product. However, when the data products are of high level such as the ones needed for spectrally integrated geological maps, landing site maps or maps from which 3D model can be derived, people providing them must be within the primary authorship. 

In the case of different version of the same map primary and secondary authors can be switched case by case.

The first authors of the geological maps can be different from the first authors of the scientific publication eventually derived from the mapping discoveries. This is because the first authors of the geological maps must be the ones that have directly drifting most of the work of linework, while the first authors of the scientific publications might be the ones that during map production have only provided substantial scientific advice. 

All the maps and related manuscript must undergo to an PLANMAP internal review before any submission.

References

Bernhardt et al. (2016) Photogeologic mapping and the geologic history of the Hellas basin floor, Mars. Icarus 254, 407-442.

Calcagno, P., Chilès, J.P., Courrioux, G., Guillen, A. (2008) Geological modelling from field data and geological knowledge. Physics of the Earth and Planetary Interiors 171, 147–157. https://doi.org/10.1016/j.pepi.2008.06.013

Cowen, E.J., Beatson, R.K., Ross, H.J., Fright, W.R., McLennan, T.J., Evans, T.R., Carr, J.C., Lane, R.G., Bright, D.V., Gillman, A.J., Oshust, P.A., Titley, M. (2003) Practical Implict Modelling. 5th International Mining Geology Conference, Bendigo, Victoria. pp 89 – 99.

Cowen, E.J., Spragg, K.J., Everitt, M.R. (2011) Wireframe-Free Geological Modelling – An Oxymoron or a Value Proposition? 8th International Mining Geology Conference, Queenstown, New Zealand.

Erkeling, G., Luesebrink, D., Hiesinger, H., Reiss, D., Heyer, T., and Jaumann, R., 2016, The Multi-Temporal Database of Planetary Image Data (MUTED): A database to support the identification of surface changes and short-lived surface processes. Planetary and Space Science 125, 43-61.

Ferrari S., Massironi M., Marchi S., Byrne P.K., Kilmczak C., Martellato E., Cremonese C. (2015). Age relationships of the Rembrandt basin and Enterprise Rupes, Mercury. ” In: Platz, T., Massironi, M., Byrne, P. K. & Hiesinger, H. (eds) Volcanism and Tectonism Across the Inner Solar System. Geological Society London, Special Publications, 401, 159-172

Galluzzi, V., Guzzetta, L., Ferranti, L., Di Achille, G., Rothery, D.A. and Palumbo, P., 2016. Geology of the Victoria quadrangle (H02), Mercury. Journal of Maps12(sup1), pp. 227-238.

Guzzetta, L., Galluzzi, V., Ferranti, L. and Palumbo, P., 2017. Geology of the Shakespeare quadrangle (H03), Mercury. Journal of Maps13(2), pp.227-238.

Giacomini, L., Carli, C., Sgavetti, M., Massironi, M., (2012). Spectral analysis and geological mapping of the Daedalia Planum lava field (Mars) using OMEGA data. Icarus, 220, 679- 693

Golombek, M., Grant, J., Kipp, D., Vasavada, A., Kirk, R., Fergason, R., Bellutta, P., Calef, F., Larsen, K., Katayama, Y., Huertas, A., Beyer, R., Chen, A., Parker, T., Pollard, B., Lee, S., Sun, Y., Hoover, R., Sladek, H., Grotzinger, J., Welch, R., Noe Dobrea, E., Michalski, J., Watkins, M., 2012. Selection of the Mars Science Laboratory landing site. Space Sci. Rev. 170, 641–737. https://doi.org/10.1007/s11214-012-9916-y

Hansen, V., (2000). Geologic mapping of tectonic planets. Earth Planet. Sci. Lett. 176, 527–542.

Heyer, T., Hiesinger, H., Reiss, D., Erkeling, G., Bernhardt, H., Luesebrink, D., and Jaumann, R., 2018, The Multi-Temporal Database of Planetary Image Data (MUTED): A web-based tool for studying dynamic Mars. Planetary and Space Science, in press. https://doi.org/10.1016/j.pss.2018.04.015

Hiesinger, H., Head, J. W., Wolf, U., Jaumann, R., Neukum, G. (2002). Lunar mare basalt flow units: thicknesses determined from crater size-frequency distributions. Geophysical Research Letters, 29, 8, http://dx.doi.org/10.1029/2002GL014847

Iqbal, W., Hiesinger, H., and van der Bogert, C. H., 2018, Reinvestigating the geology and CSFDs of the Apollo 12 landing site. Lunar and Planetary Science Conference 49, 1002.

Ivanov M.A.,  Abdrakhimov A.M.,  Basilevsky A.T., Demidov N.E. ,  Guseva E.N., Head J.W. , Hiesinger H. ,  Kohanov A.A., Krasilnikov S.S. (2017, in press). Geological characterization of the three high-priority landing sites for the Luna-Glob mission. Planetary and Space Science

Ivanov M. A., Hiesinger, H., Orgel, C., Pasckert J.H., van der Bogert, C. H., and Head, J. W., 2018, Geology of the northern portion of the SPA basin on the Moon: Evidence for compositional stratification of the ancient lunar crust. Lunar and Planetary Science Conference 49, 1138.

Ivanov M. A., Hiesinger, H., Orgel, C., Pasckert J.H., van der Bogert, C. H., and Head, J. W., in review, Geologic history of the northern portion of the South Pole-Aitken basin on the Moon. Journal of Geophysical Research - Planets.

Kinczyk, M.J., Prockter, L.M., Byrne, P.K., Denevi, B.W., Ostrach, L.R., Head, J.W., Fassett, C.I., Whitten, J.L., Thomas, R.J., Buczkowski, D.L. and Hynek, B.M., 2016, June. Preparing the First Global Geological Map of Mercury. In Annual Planetary Geologic Mappers Meeting (Vol. 1920).

Kramer, G. Y., Kring, D. A., Nahm, A. L., and Pieters, C. M., 2013, Spectral and photogeological mapping of Schrödinger basin and implications for post-South Pole-Aitken impact deep subsurface stratigraphy. Icarus 223, 131-148.

Krüger, T., van Der Bogert, C.H. and Hiesinger, H., 2016. Geomorphologic mapping of the lunar crater Tycho and its impact melt deposits. Icarus273, pp.164-181.

Mancinelli, P., Minelli, F., Pauselli, C. and Federico, C., 2016. Geology of the Raditladi quadrangle, Mercury (H04). Journal of Maps12(sup1), pp.190-202.

Massironi, M., et al. (2015). Two independent and primitive envelopes of the bilobate nucleus of comet 67P. Nature, vol. 526, p. 402-405, ISSN: 0028-0836, doi: 10.1038/nature15511.

McCauley, J.F., Guest, J.E., Schaber, G.G., Trask, N.J. and Greeley, R., 1981. Stratigraphy of the Caloris basin, Mercury. Icarus47(2), pp.184-202.

Nass, A., van Gasselt, S., Jaumann, R., & Asche, H. (2011). Implementation of cartographic symbols for planetary mapping in geographic information systems. Planetary and Space Science, 59(11-12), 1255-1264.

Nass, A., et al. (2017). Planetary Cartography and Mapping: Where we are today, and where we are heading for? In: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 42(W1), 105-112.

Okubo C.H., (2014). Bedrock Geologic and Structural Map Through the Western Candor Colles Region of Mars. USGS Scientific Investigations Map 3309

Pajola, M., Rossato, S., Carter, J., Baratti, E., Pozzobon, R., Erculiani, M.S., Coradini, M., McBride, K., 2016. Eridania Basin : an ancient paleolake floor as the next landing site for the Mars 2020 rover. Icarus 275, 2–8. https://doi.org/10.1016/j.icarus.2016.03.029

Penasa, L., et al. ( 2017). A three-dimensional modelling of the layered structure of comet 67P/Churyumov-Gerasimenko. Monthly Notices of the Royal Astronomical Society 469, pp. S741-S754

Potts, N. J., Gullikson, A. L., Curran, N. M., Dhaliwal, J. K., Leader, M. K., Rege, R. N., Klaus, K. K., and Kring, D. A., 2015, Robotic traverse and sample return strategies for a lunar farside mission to the Schrödinger basin. Advances in Space Research 55, 1241-1254.

Prockter, L.M., Kinczyk, M.J., Byrne, P.K., Denevi, B.W., Ostrach, L.R., Head, J.W., Fassett, C.I., Whitten, J.L., Thomas, R.J., Buczkowski, D.L. and Hynek, B.M., Ostrach, L.R., Blewett, D.T., Ernst, C.M. and the MESSENGER imaging group, 2016, The First Global Geological Map of Mercury. LPSC 47, #1245

Ramsdale, Jason D., Balme, Matthew R., Conway, Susan J., Gallagher, Colman, van Gasselt, Stephan A., Hauber, Ernst, Orgel, Csilla, Séjourné, Antoine, Skinner, James A., Costard, Francois, Johnsson, Andreas, Losiak, Anna, Reiss, Dennis, Swirad, Zuzanna M.,

Kereszturi, Akos, Smith, Isaac B. and Platz, Thomas (2017). Grid-based mapping: A method for rapidly determining the spatial distributions of small features over very large areas. Planetary and Space Science, 140 pp. 49–61.

Tobler, W. 1987. “Measuring Spatial Resolution”, Proceedings, Land Resources Information Systems Conference, Beijing, pp. 12-16.

van Gasselt, S., & Nass, A. (2011) Planetary mapping—The datamodel's perspective and GIS framework. Planetary and Space Science59(11-12), 1231-1242.

Whitten, J.L., Head, J.W., Denevi, B.W. and Solomon, S.C., 2014. Intercrater plains on Mercury: Insights into unit definition, characterization, and origin from MESSENGER datasets. Icarus241, pp.97-113.

Wilhelms, D.E. and McCauley, J.F., 1971. Geologic map of the near side of the Moon. Reston: US Geological Survey. https://www.lpi.usra.edu/resources/mapcatalog/usgs/I703/

Wilhelms, D. E. (1987) The Geological History of the Moon, USGS Professional Paper 1348, 302 pp., https://pubs.er.usgs.gov/publication/pp1348.

  • No labels