By David R. Soller
1996
This report summarizes a technical review of USGS Open-File Report 95-525, "Cartographic and Digital Standard for Geologic Map Information" and OFR 95-526 (diskettes containing digital representations of the standard symbols). If you are considering the purchase or use of those documents, you should read this report first. For some purposes, OFR 95-525 (the printed document) will prove to be an excellent resource. However, technical review identified significant problems with the two documents that will be addressed by various Federal and State committees composed of geologists and cartographers, as noted below. Therefore, the 2-year review period noted in OFR 95-525 is no longer applicable. Until those problems are resolved and formal standards are issued, you may consult the following World-Wide Web (WWW) site which contains information about development of geologic map standards:
This report is preliminary and has not been reviewed for conformity with U.S. Geological Survey editorial standards. Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.
This report summarizes a technical review of USGS Open-file Report 95-525 "Cartographic and digital standard for geologic map information" and OFR 95-526 (diskettes containing digital representations of the standard symbols). If you are considering the purchase or use of those documents, you should read this report first. For some purposes, OFR 95-525 (the printed document) will prove to be an excellent resource. However, technical review identified significant problems with the two documents that will be addressed by various Federal and State committees composed of geologists and cartographers, as noted below. Therefore, the 2-year review period noted in OFR 95-525 is no longer applicable. Until those problems are resolved and formal standards are issued, you may consult the following World-Wide Web (WWW) site which contains information about development of geologic map standards:
URL: http://ncgmp.usgs.gov/ngmdbproject/home.html
If the WWW is not available to you, or if you have a specific question regarding the use of the document, send e-mail (drsoller@usgs.gov) or write to:
Executive Secretary, Geologic Data Subcommittee
USGS
908 National Center
Reston, VA 20192
The practice of geologic mapping, and representation of geologic information in digital databases and on maps, must be supported by standards if we are to witness increased utility for our information in support of societal issues. It is essential that such standards be developed collaboratively, with mutual agreement of the need for, and content of, a standard. Unfortunately, collaboration on USGS Open-file Report 95-525 was minimal. Cooperators on the report were the USGS, the Association of American State Geologists (AASG), and the Federal Geographic Data Committee's (FGDC) Geologic Data Subcommittee (GDS).
The document had been intended for release as a FGDC draft Federal standard, through the GDS. Under Executive Order 12906, which established the National Spatial Data Infrastructure, the FGDC has responsibility for assuring development of standards necessary to promote the dissemination and use of information produced by the Federal government and, with their cooperation, by the states. A rigorous technical review of this document was essential, to determine whether the document was acceptable to its cooperators and to the geologic mapping community in general.
The technical review was conducted in 1996 (see Appendices). Comments were sought from the state geological surveys (20 states provided comments), from mappers and cartographers in the USGS Geologic Division, from the USGS Water Resources Division, and from the FGDC Secretariat. Some private companies also responded. Many of the reviews are excellent, and provide valuable insight to the technical issues of geologic mapping -- for their efforts, I extend my appreciation. I used their guidance to help prepare this report.
The authors of the standards document also certainly deserve appreciation for their efforts. The geoscience community needs various standards for managing and presenting data in digital form and on maps, and this document deserves credit for addressing some of those standards. Many reviews expressed their appreciation to the USGS and the document's authors, and noted the critical need for such standards. However, the most thorough reviews also identified significant shortcomings to the document. Because I am charged through the FGDC with development of Federal standards for geologic information, and because standards must be developed through general consensus among all parties, this report focuses on the need for the document's revision.
The document contains two parts -- a general discussion of geologic mapping procedures and cartographic design ("Part I") and the collection of proposed standard symbols, colors, and patterns ("Parts II and III"). These two parts are separately discussed below. My conclusions are supported by reviewer's quotes. To minimize controversy, I have chosen to omit attribution to the quotes. However, a list of the reviewers is appended.
The title of this document is most welcome -- it suggests a standard for digital map information that is badly needed by the geoscience community. The transition from traditional paper-based map products (destined for application through visual inspection of the map) to digital databases (destined for analysis using statistics and GIS techniques as well as output to paper maps for visual inspection) has been difficult and expensive for geological surveys around the world. Regardless, the opportunity to prepare our information in a format more useable for societal decisionmaking, and the potential analytical power of the technology, have hastened the transition.
Unfortunately, this document does not address a comprehensive digital standard:
"The original title suggests to me that there is a digital standard for geologic information. That is to say, there is a standard for computerized geologic databases, or at least certain kinds of geologic data. This is clearly not the case, as there is no such model even suggested. ... The assumption that the printed (plotted) map is the most important information source is clearly rooted in the past, not the future."
"This report consistently treats digital geologic maps as digital copies of our current paper maps."
As an example of this bias, I refer you to the document's item 1.0.1.2:
"This paragraph permits a point of observation to be moved on the map for the sake of cartographic clarity. I firmly believe that one should never be permitted to move an observation point."
Several reviewers suggested that Part I should be a guidelines document, not a standard:
"...much of the "standard" is not a standard at all, but rather is a style manual for USGS authors and draftsmen. Unless everyone agrees that the only acceptable geological maps are those that are stylistically identical to USGS maps, most of the detail in Part I is irrelevant."
Other reviewers questioned whether they would, or could, adhere to the standards:
"Because we have always viewed USGS cartographic standards such as these as guidelines, some of our comments concern areas where we have found these standards to be impracticable. It should be recognized that [we] and other state geological surveys may have developed (or may need to develop) standards better suited to their specific GIS or digital capabilities."
Similar concern was also expressed by USGS mappers:
"We believe that the draft goes too far in defining standards that are difficult or impossible to apply in all situations or that have little scientific utility in some terranes."
Particularly common were questions about the document's standard for locational accuracy. Because of the rigorous use to which analysts may apply our digital data, resolution of this issue is even more important than in the past. A standard or guideline for locational accuracy must be addressed, collectively, before public release in a document such as this.
"The discussion of locational accuracy in section 1.0.1 is seriously flawed and is not acceptable under the requirements of the ... FIPS standard for exchange of digital cartographic data. ... Federal standards require that the accuracy of features on a digital cartographic product must be specified, along with an explicit statement of how the accuracy has been determined. ... The USGS either has to test the positional accuracy of a wide variety of geologic maps and publish the results, which they can then cite, or they will have to explicitly acknowledge that the positional accuracy of features on their geologic maps are not known but are subjectively believed to be thus-and-so."
The USGS could choose to submit this document as a Federal standard, which by definition need not be applied by the states [even this strategy may be somewhat murky, as some reviewers questioned whether participants in the STATEMAP component of the National Cooperative Geologic Mapping Program must adhere to a Federal standard]. We will instead coordinate the development of standards through broad consensus, among the state, academic, private, and federal geologic mapping community. The following actions will be taken.
Because many issues related to preparation of digital geologic maps and databases are complex and not yet resolved, each issue must be thoroughly discussed and implemented after careful testing, perhaps through pilot studies. The USGS/AASG working groups will develop guidance or standards for technical issues that underlie digital mapping. In certain cases, these efforts are planned to culminate in a proposed FGDC standard; for example, a draft standard for a geologic data model.
"Now that the Geologic Data Subcommittee has developed standardized geologic symbols for making maps and thereby documenting what is known very precisely, the Subcommittee should turn its attention towards the more difficult subject of developing digital standard geologic data. ... It is in this area [development of a geologic data model] that the FGDC might make its most important contribution."
"Several of us who have made maps of relatively uneroded, undeformed terrain have declined to show cross sections on our published maps because the sections would be meaningless."
"Most maps of surficial geology don't have cross sections, therefore, by Table 1, they are not geologic maps."
"In many cases "new" geology is based on reinterpretation and synthesis of the old. Geologic maps are highly interpretive and as such the responsibility for authorship has to fall to the geologist; it can not be legislated in a standard."
Because Part I focuses on geologic mapping and cartography instead of digital data, the proposed guidance document should not address digital issues:
"There is a heavy emphasis on geologic maps as opposed to digital geologic data and discussion of both is mixed together throughout the standard. I think it would be clearer and more useful to separate the cartographic from the digital standards."
This section contains a proposed collection of standard symbols, colors, and patterns, most of which have been used by the USGS for decades. For each symbol, there is a cartographic specification and a unique code based on the Digital Line Graph 3 (DLG-3) standard coding scheme. The symbols are also available in 28 Encapsulated Postscript files as USGS Open-file Report 95-526; however, as noted below, these files are not useable for digital cartography.
There was widespread agreement that Parts II and III, and especially the symbols themselves, will be of great utility to the geoscience community. Reviewer comments on symbols, patterns, and colors will be useful in developing the final standard sets for submission to the FGDC. However, the attribute coding scheme for symbols is perhaps the most contentious issue in the document, and must be addressed before submittal.
"The most significant problem with the proposed standard is expressed in this section [section 2.0.1], where it specifies a "major-minor" code system that requires a minimum of 7 digits to represent a feature, and may require as many as 28 digits. The coding scheme really is an exhaustive, taxonomic-type classification of map symbols, and while theoretically it may be unambiguous in all conceivable circumstances, it is unnecessarily cumbersome and impractical for actual use. It ignores current practice and is incompatible with coding schemes in the most widely used cartographic programs such as ARC/INFO and AutoCAD, as well as the coding procedures in graphic languages such as IGL, HPGL, and PostScript."
It should not be concluded that a coding scheme is unnecessary, only that significant discussion needs to occur among technical experts who must, upon acceptance of such a standard, implement it daily. Implementation is the important consideration -- are the symbols and the attendant coding scheme appropriate to the task of preparing a digital database and map, appropriate to the breadth of hardware and software used by the geoscience community, and appropriate to the standard data model now under development?
"The need for an attribute code standard is without question. Data collected today needs to be useable many years from now. Will software be made available to link these new codes to the ARCDLG command in ARC/INFO? ... Is DLG the best choice, or should a completely new, less cumbersome, format be developed by a consortium of software developers, geologists, digital cartographers? We are going to have to live with this exchange standard for a long time. Therefore, the development of a standard needs to be thought out carefully and with ALL users in mind."
A coding scheme should be designed to most efficiently support database queries and analysis. For example, numeric fields are generally more amenable to sorting, reselection, and query than character fields. The proposed coding scheme does not adequately consider efficiency in the act of encoding a symbol or in the query of a map database.
"Sorting operations appear tedious and time consuming. For example, under this coding structure multiple reselection iterations are required in order to sort for all faults. [Our] coding scheme uses an approach that is the reverse of the proposed USGS system. All faults are tagged by a primary variable; sorting of specific fault types is then conducted by reselection from the general fault population."
A further consideration is that some feature attribution, for example expressions of spatial uncertainty, may be scale-dependent:
"Section 2.0.1 "Attribute Coding for Earth Science Features" has what I regard as a fundamental flaw. This is, the 1:24,000 and 1:100,000 data are lumped together with the same coding (as far as I can determine). That is, I can see no way of distinguishing data with different uncertainty other than "definite, approximate, and inferred." However, what is definite at 1:100,000 scale may be only inferred at 1:24,000 scale."
Although standard symbology is important for map display, utility of the geologic database and map requires that we first specify standard definitions (a "content" standard) for the geologic features that we would symbolize.
"[In section 2.0.1.0.4] the standard comes somewhat close to articulating what is required for digital content standards by saying: "Feature points are represented by a point having attributes in a digital data base." It's the features and their attributes that need to be defined and standardized. In doing so you are defining what information is important; the map is just a means used to represent that information."
Some reviewers questioned whether they had the resources needed to implement the coding scheme. Although a standard need not specifically address issues of implementation, they must be explicitly considered as the standard is being developed.
"It is not clear how the metadata and properties identification is expected to be accomplished. If every item on every map we produce needs to be coded for attributes with both major and minor codes, then we'll need to hire additional staff or expect maps to take orders of magnitude longer to produce. The expectations of how this system is to be applied needs to be realistic and needs to be defined more clearly."
Because of the general lack of digital standards and the stature of organizations cited as collaborators on this document, testing and implementation of the standard has already begun in some agencies.
"I have found in the past that ... attributing, as recommended in ... OFR 95-525, is extremely cumbersome to use in a map database. It is designed (apparently) to be able to lump any sort of data into a single database file, mostly aimed at storing cartographic information, not necessarily geologic information. Given the variety of data that needs to be included in a useful geologic database, it makes a lot more sense to me to put the data in a number of coverages and files, following the Arc/Info logic of putting data in layers that make sense. I'm including a rough draft of the data structure design we're presently using. The attribute coding follows that laid out in OFR 95-525 to the extent that I thought that worked. As much of the useless overhead as possible was removed to make data storage more efficient and coding easier to figure out."
Implementation of the symbol standards in a digital environment, with output to plotters of various resolution and design, to film writers, and to display terminals, requires extensive forethought and development costs. Unfortunately, the digital symbols in OFR 95-526 will not serve as the basis for a digital implementation of the standard.
"[It] simply contains a series of Adobe Illustrator files, each one corresponding to a page in Sec. 2.1. It is possible to ungroup an illustration, select a particular symbol, and enlarge it to any desired size. However, you then discover that is consists only of strokes, identical in nature to the line-drawn symbols in the old Bureau of Standards Hershey fonts or to IGL symbols. Since the symbols are not drawn as outlines, their graphic representations cannot be used to define PostScript or other scalable symbols, and the precision with which they can be printed is dependent upon the original drawing and not on the resolution of the output device. ... The specifications for symbols are adequate only for creating stroked symbols, not scalable outlines, and so are not appropriate for use with modern high-resolution graphics devices."
The draft standard will be prepared electronically and produced as a paper document for FGDC review. It will also become available in electronic format. Preparation of the draft standard will proceed as quickly as possible with available resources. Until the draft is available for comment through the FGDC review process, the symbols, colors, and patterns in OFR 95-525 should serve as a valuable reference.
List of reviewers. In parentheses, I note all reviewers I could identify from the correspondence received. In some cases, the respondent noted that various unnamed persons had reviewed the document. I thank all who took the time to provide comments.
Excerpts from a letter to Thomas Berg (State Geologist of Ohio), asking for the state geological surveys to conduct a formal review of USGS Open-file Report 95-525:
__________________________________
December 5, 1995
Thomas Berg
Chair, AASG Data and Standards Committee
Tom --
Based on our telephone conversations with you and with Walter Schmidt, the Federal Geographic Data Committee's (FGDC) Geologic Data Subcommittee (GDS) requests that the AASG conduct a technical review of USGS Open-file Report 95-525, entitled "Cartographic and digital standard for geologic map information." Symbology represented in that report are released in digital form in USGS Open-file Report 95-526; therefore, a review of that document is also requested. Because the AASG was listed as a cooperating entity in 95-525, it is appropriate that they have the opportunity for a technical review. Likewise, the USGS will be conducting a technical review.
The GDS does not wish to constrain your review because there may be technical issues or difficulties with implementing this standard that cannot be foreseen prior to technical review. We do offer some issues for your consideration, partly based on preliminary reviews of the document:
We do not expect each person to review the entire document with a consistent level of attention throughout, but rather to concentrate on those aspects of the document which are most pertinent to their expertise. Because this document can be subdivided into discrete topics (e.g., symbology, coding scheme, or geologic map content), we do not require that all aspects of the document be readied for FGDC release at the same time. If the entire document is acceptable, then it can be released in its present form. If not, we can submit for FGDC approval those parts of the document that we all agree are worthy and non-controversial; if the review identifies any controversial sections, those could be held back for further discussion without delaying the remainder of the document.
We are mailing you at least 50 copies of the document, and one copy of the floppy disk containing Report 95-526. However, for reviewers who want to examine the digital version of the symbols (separately released as USGS Open-file Report 95-526, but not widely available), it may be easier for them to obtain a copy from Soller's workstation. There you will find twenty-eight EPS graphics files and a WP5.1 (export) version of the documentation. You will also find a compressed tar file. Take your pick. The graphics were developed in Adobe Illustrator. All graphics files are uncompressed and DOS-compatible. On the Mac, the EPS files can be opened in Adobe Illustrator.
If your review identifies specific elements of the standard that would benefit from revision by technical experts and/or managers, the GDS and USGS will work with AASG to resolve those concerns. Thank you for assisting the geoscience community in this endeavor.
John F. Sutter
Chair, FGDC Geologic Data Subcommittee
David R. Soller
Executive Secretary, FGDC Geologic Data Subcommittee