MnModel

Landform Sediment Assemblage (LfSA)

How to Construct a MnModel Landscape Suitability Model

This is the original instruction manual, prepared in 2000. Since then, research methods have advanced to make better use of GIS technology. The updated procedures are documented in the Landform Sediment Assemblages in the Upper Mississippi Valley, St. Cloud to St. Paul, for Support of Cultural Resource Investigations. However, the LfSA Map Unit Field Code Key Table (Appendix A of this report) has been updated and links within this report are to the most recent version.

Prepared for:
Minnesota Department of Transportation

Prepared by:
Edwin R. Hajic, Philip E. Paradies, and Curtis M. Hudak
Foth & Van Dyke and Associates Inc.

December 2000

Table of Contents

• 1. Introduction
  
  • 1.1 Objectives
  • 1.2 Flow Chart and Report Structure
  • 1.3 Data Proofing and Quality Control
  • 1.4 Recommended Personnel
• 2. Minimum Requirements
  • 2.1 Assumptions
  • 2.2 GIS Field Code Key
• 3. Source Data Collection
  • 3.1 Assemble Relevant Literature and Source Data
  • 3.2 Interpreting Source Data
• 4. Evaluate Usefulness of Source Data
  • 4.1 Weigh Digital Versus Non-digital Data
  • 4.2 Compare Scale of Potential Source Map Information to the 1:24,000 Scale of Mapping
  • 4.3 Learn the Limitations of the Source Data
  • 4.4 Select Source Data to be Used and Create an ArcInfo Coverage
• 5. LfSA Mapping and Labeling
  • 5.1 Determine LfSA Map Units
  • 5.1.1 Establish General Quaternary Geologic Setting
  • 5.1.2 Estimate a General Suitable Polygon Size Range for Mapping
  • 5.1.3 Determine the LfSA Map Units
  • 5.1.4 Establish Criteria for Placement of Lines that Delineate LfSA Polygon
  • 5.2 Delineate and Label Polygons
  • 5.2.1 Draw Polygons
  • 5.2.2 Establish a System of Labeling Polygons that Reflects Criteria for Mapping
  • 5.2.3 Evaluate Validity of LfSA Map Units and Map Unit Criteria
  • 5.2.4 Monitor for Continuity in Application of Criteria
  • 5.2.5 Label Each Polygon
  • 5.2.6 Review Map Line Work
  • 5.2.7 Review Polygons for the Presence and Accuracy of Map Symbols
  • 5.2.8 Digitize Hardcopy LfSA Maps
• 6. Digitizing Procedures and Standards
  • 6.1 Bring Digital Source Data into Conformance with Project Standards
  • 6.2 Register Hardcopy Source Data
  • 6.3 Register Digital Images
  • 6.4 Maintain an Archive File
  • 6.5 Develop Minimum Digitizing Standards
  • 6.6 Digitize Data
  • 6.7 Check Whether Digitized Lines Match Source Data Lines
  • 6.8 Build and Verify Topology
  • 6.9 Tag Digital Polygons with Labels
  • 6.10 Check for Edge Matching of Lines and Polygons
  • 6.11 Join Map Sheets
  • 6.12 Attach Source or New Data Attributes to GIS Coverage Attribute Table 21
  • 6.13 Proof the GIS Coverage Attribute Table 22
  • 6.14 Digitally Check Each LfSA Map Unit for Labeling
  • 6.15 Digitally Check Each LfSA Map Unit for Location and Geologic Continuity
  • 6.16 Make Necessary Adjustments
  • 6.17 Geologist Reviews Adjustments
  • 6.18 Geologist 2 Cross-Checks
  • 6.19 Update and Maintain Code Key Table
• 7. Assemble the Project LfSA Database
  • 7.1 Code the Map Symbol Field
  • 7.2 Code the Geomorphology Set of Attributes
  • 7.3 Code the Material/Material Sequence Set of Attributes
  • 7.4 Code the Temporal Set of Attributes
  • 7.5 Code the Summary Code Strings
  • 7.6 Code the Geologic Age Set of LSR Attributes
  • 7.7 Code the Depositional/Post-Depositional Environment Set of LSR Atributes
  • 7.8 Calculate and Enter LSR's
  • 7.9 Proof Project LfSA Database
• 8. Model Testing
  • 8.1 Landform Sediment Assemblage Testing
  • 8.1.1 Conduct a Field Review of Landform Sediment Assemblage Mapping
  • 8.1.2 Confirm the Location and Verify the Composition of Landform Sediment Assemblages
  • 8.1.3 Date Specific Landform Sediment Assemblages
  • 8.2 Archeological Database Testing
  • 8.2.1 Cross-check Location of Archaeological Sites Against LfSA Polygons
  • 8.2.2 Cross-check Time Span of Possible Human Occupation Against LfSA Polygons
• 9. Data Management and Presentation
  • 9.1 Create Geologic Cross-sections Using MnModel Standards
  • 9.2 Create a Shapefile of Core Locations and Code Attributes to Allow Future Hot-linking to the Core Logs
  • 9.3 Creat a Shapefile of Cross-sections and Code Attributes to Allow Future Hot-linking to the CADD Cross-section Drawings
• 10. Final Model Quality Control
  • 10.1 Document Model Discrepancies After Model Testing
  • 10.2 Make Adjustments to Address Model Discrepancies After Model Testing
• 11. Prepare a Report that Documents the Landscape Suitability Model
  • 11.1 Introduction
  • 11.2 Objectives
  • 11.3 Overview of the Project Area
  • 11.4 Methods
  • 11.5 Landscapes and Landforms
  • 11.6 Project Area Geology and Landscape Suitability Rankings
  • 11.7 Conclusions
  • 11.8 Recommendations
• 12. References Cited
• 13. Acknowledgements

Tables

• Table 1 Suggested Sources for Geologic and Archaeological Reference Literature to Aid in Mapping
• Table 2 Suggested Sources for Map Information
• Table 3 Suggested Sources for Aerial Imagery
• Table 4 Key Digital Spatial Data Files
• Table 5 Sources of Subsurface Records
• Table 6 Geomorphic Hierarchy Used in MnModel Mapping

Figures

• Figure 1 Flow Chart for Constructing a MnModel Landscape Suitability Model
• Figure 2 Example of Acceptable and Unacceptable Digitizing
• Figure 3 Landform Sediment Assemblage - Map Symbol and Identifier Set
• Figure 4 Landform Sediment Assemblage - Geomorphology Set
• Figure 5 Landform Sediment Assemblage - Material/Material Sequence Set
• Figure 6 Landform Sediment Assemblage - Temporal Set
• Figure 7 Landform Sediment Assemblage - Age and Depositional/Post-Depositional Set

Appendix

App. A MnModel Geomorphology Field Code Key Table

1. INTRODUCTION

1.1 Objectives

This manual was prepared as an educational reference on how to construct a landscape suitability model (LSM) at the 1:24,000 scale. The target audience consists of professional earth scientists, GIS specialists, and database managers. LSM's are powerful planning tools that were developed for support of cultural resource management, specifically to assess the suitability of landform sediment assemblages (LfSA's) to host pre-Euroamerican cultural deposits ("archaeological sites") buried by geologic processes. The resulting LSM's are also suitable for a wide range of other applications.

Standard MnModel LSM's consist of three parts: a Geographic Information System (GIS) digital map of LfSA's; an accompanying GIS-based geomorphic attribute database; and a written report. The GIS attribute table includes landscape suitability rankings (LSR's) for each LfSA mapped. LSR's provide expert system estimates that can be used by resource managers in conjunction with other data to plan for cultural resources within a project area. The ability to plan for cultural deposits, especially those that may be deeply buried and beyond standard one meter deep archaeology shovel tests, allows for the chance to alter designs to avoid potential archaeological resources, or to at least budget time and money for surveys and potential construction delays.

The manual, if followed, will assist in maintaining compatibility between new LSM's and those previously constructed for MnDOT's MnModel. Eventually, MnDOT will edge match and join all the models together; hence minimum standards help ensure compatibility.

1.2 Flow Chart and Report Structure
The tasks to construct an LSM are summarized in a flow chart (Figure 1). Tasks in the flow chart are cross-referenced with task or chapter numbers in the manual so that the reader can locate how to either perform a process or consider factors in making a decision. The flow chart and manual will guide the reader/user through data collection, data processing, digitizing processes and standards, map delineation and labeling, expert system database assignments, data management and presentation, report writing, and data proofing and quality control measures. Outlines of digitizing processes and standards applicable to both hardcopy source data and LfSA map entry into the GIS are outlined. These steps are not greatly elaborated upon, however, because it is one of the assumptions of this report that the end user will have GIS expertise.

The method presented in this manual was developed and used during the first five years of MnModel's life. It was designed to be readily available to, and affordable for, most consulting firms and government agencies. The manual is intended as a guide, not as a "cook book." It is recognized that many of the tasks have multiple ways of being accomplished, and it is expected that new ways of accomplishing them will be developed in the future. Technology is rapidly improving. New techniques and technologies should be considered when starting an LSM project. For example, more efficient and accurate methods of entering data into the GIS are now available compared to when MnModel was first conceived. GIS capabilities will undoubtedly advance with time-leading to better methods of capturing, maintaining, visualizing and analyzing data.

1.3 Data Proofing and Quality Control
A series of quality control tasks must be performed at a number of steps in the process regardless of the method chosen to enter data into the GIS. Models of average size can include thousands of polygons. A MnModel GIS coverage attribute database can contain tens of thousands of entries. Many different types of errors can occur, so it is essential to have a rigorous quality control process. The many data proofing and quality control steps are interwoven throughout the manual and follow the tasks they are designed to check.

1.4 Recommended Personnel
Several people should be involved in the project to achieve maximum quality assurance through multiple independent quality control reviews. Two geologists are recommended. Geologist 1 is responsible for mapping, model development, and report preparation. Geologist 1 performs the initial quality control checks, as well as some subsequent checks. Geologist 2 provides an independent crosscheck and maintains updates on key records. A GIS technician performs technical tasks and performs the initial and subsequent proofs of the GIS spatial data and attribute table. An administrative assistant can perform some of the visual checks.

Return to the Top

2. MINIMUM REQUIREMENTS

2.1 Assumptions
The manual is written with the assumptions that the reader/user will have:

• Ready access to the minimum technology (e.g., hardware, software, etc.) required;
• Working knowledge of all software, hardware, and relevant terminology;
• Considerable experience and knowledge in geologic (emphasis on geomorphic, stratigraphic, and sedimentologic) and pedologic (soil) principles, processes and terminology;
• Considerable experience and knowledge in GIS spatial and attribute database construction and management principles;
• Some basic geologic or geomorphic mapping experience at a 1:24,000 or larger scale;
• Familiarity with the needs and goals of cultural resource managers; and
• Familiarity with MnModel's LSM's (see Hudak and Hajic, 1999).

2.2 GIS Field Code Key
The end user must become familiar with the structure, content, and use of MnModel's Field Code Key Table (Appendix A). The Code Key Table is an integral guide to the construction of the digital landform sediment assemblage (LfSA) map, accompanying project LfSA database, and ultimately the GIS attribute table.

The LfSA is the fundamental map unit used in MnModel to which attributes, including LSR's, are assigned. An LfSA consists of a two-dimensional landform surface and the genetically-related material underlying the landform. As described later in the manual, LfSA polygons are mapped, digitized, and coded with 45 attributes, 33 of which are assigned based on the Code Key Table (see Appendix A). The remaining 12 attributes consist of LSR's and values used to calculate the LSR's. These values are based on information provided directly by, or as an interpretation of, the 33 Code Key attributes.

"Code Number and Title" identify attributes in the Code Key Table. Each "Code Number and Title" has a corresponding field in the project LfSA database and GIS attribute table that is part of each geomorphic model. These attributes describe the physical and temporal characteristics of an LfSA map unit and all polygons that are assigned to that map unit, and include characteristics particularly relevant to establishing and interpreting the contexts of pre-Euroamerican cultural deposits. For additional information, see Hudak and Hajic (1999).

Attributes in the Code Key Table are categorized into three sets: Geomorphology, Material/Material Sequence, and Temporal. The Geomorphology set is used to describe the landform, its characteristics, and its regional physiographic setting. The Material/Material Sequence set is used to describe the sediments and stratigraphy of an LfSA. The Temporal set is used to record various relative, absolute, and time-stratigraphic age assignments. Descriptions of the different attributes are provided in the "Comments" column of the Code Key Table. Additional information is provided in Hudak and Hajic (1999).

The Code Key Table also lists "Values" for each "Code Number and Title." These lists may be amended as additional valid values are needed to map new regions. The "Values" are sometimes defined and discussed and may include relevant references in the "Comments" column. The "GIS Code Symbol" in the Code Key Table provides the abbreviation for each valid "Value." The "GIS Code Symbol" is the actual value entered into the appropriate field in the project LfSA database and ultimately the GIS attribute table. The "Map or Code String Symbol" in the Code Key Table is a one or two alphanumeric abbreviation for each valid "Value." This abbreviation is utilized in developing map symbols for LfSA map units, as well as in constructing summary code strings used in the project LfSA database and GIS attribute table.

Return to the Top

3. SOURCE DATA COLLECTION

3.1 Assemble Relevant Literature and Source Data
Assemble information used to develop an LSM at the beginning of a project, before mapping commences. Look for material that will provide data on, or contribute to, interpreting project area geomorphology, soil-geomorphic relationships, late Quaternary stratigraphy and sedimentology, landform and sediment ages, and landscape evolution history. Sources of information typically include available geologic and archaeological literature, maps, imagery, digital data files, subsurface records, and temporal data. Project-specific data generated through fieldwork also contribute important information. Assembling and maintaining a library of information mentioned in this section is highly recommended. This fosters efficiency in the long run, because much of the information will be applicable directly or indirectly to multiple mapping projects.

A thorough search of the geologic and archaeological literature would include, at a minimum, the sources listed in Table 1. A number of these sources can be identified from searching digital databases, such as GeoRef. Consult abstracts from the national meetings of the Geological Society of America (GSA), Society for American Archaeology (SAA), and other national earth science and archaeological societies; regional meetings of the GSA and archaeological organizations; and local or specialized meetings. Review dissertations from Minnesota universities and colleges, as well as other schools in surrounding states. Dissertations International provides the most complete catalog of dissertations and theses. There is potentially a vast amount of data in the gray literature - unpublished reports, usually completed to fulfill some type of contractual or compliance obligation. Geologic reports might include geotechnical or geoarchaeological investigations completed for various governmental agencies. Reports of archaeological surveys, evaluations, and data recoveries are completed for a wide range of private and governmental bodies. These reports are not widely circulated, but the State Historic Preservation Office (SHPO) maintains a library of report copies, as well as a site file directory. Additional pertinent information might be found in the soils and environmental literature. The Internet is a valuable source for identifying both published and gray literature on all of these topics.

Table 1 Suggested Sources for Geologic and Archaeological Reference Literature to Aid in Mapping

* G = geologic; A = archaeological

Potential sources of modern and historic maps are listed in Table 2. Most are obtainable from state or federal sources. Some are available in digital format. Since MnModel's LSM's are developed at the 1:24,000 scale, U.S. Geological Survey 7.5-minute topographic maps are essential. These maps are also available as USGS 7.5-minute Digital Elevation Models (DEMs) and Digital Raster Graphics (DRGs). Maps not belonging to a formal series, and/or executed for a specific purpose, may be included in other literature sources listed in Table 1 or elsewhere. Key maps include those that provide as detailed information as possible on topography, geomorphology, late Quaternary geology, material thickness, and, for some parts of the state, bedrock geology. Other than USGS 7.5-minute topographic maps, USDA-NRCS county soil maps may be the best source of detailed information, particularly in forested areas where the utility of aerial photography is limited. The basic soil map unit is the soil series and is defined by a modal profile and set of characteristics. NRCS mapping is based on NRCS aerial photography and, more recently, USGS National Aerial Photography Program (NAPP) aerial photography, with a regimen of field checking. One limitation of soil maps is that they give the false impression of rigid boundaries, when in reality, most soil series boundaries are gradations. Also, NRCS field checking, although structured, may be limited. This can be verified by contacting the NRCS regarding the survey in question. Soil series are grouped into associations and related to different landscape positions, but often only in general or vague terms. Although county soil reports contain a potential wealth of information, they are produced for a specific purpose other than geomorphic mapping. Soil polygons should not be assumed to correspond closely to discrete, unambiguous soil units or landscape positions on the ground. Digital soil surveys are now available for 57 Minnesota counties; however, these are of varying quality. Only 14 counties have the highest quality SSURGO data from NRCS.

Some of the most recent Minnesota Geological Survey maps are also available in GIS format.

Table 2 Suggested Sources for Map Information

Sources of aerial photography and other remotely sensed imagery are listed in Table 3. Important qualities sought in aerial photography and imagery include strong tonal contrasts that reflect soil colors, soil water content, and erosional conditions; high resolution; and stereo coverage. A full range of imagery products is available from the USGS EROS Data Center, and they can provide a listing for your study area based on a search of their database. Most modern imagery at EROS is gathered with standardized procedures as part of federal programs. The most useful aerial photography is produced by the U.S. Geological Survey as part of the NAPP. The USGS also is a source for digital orthophotoquads (DOQs). Aerial photography became a common practice in the 1930's, and there is little, if any, available historic imagery before this time. The USDA NRCS has commercially available black-and-white photography, often with multiple years of coverage. However, this imagery is typically collected at a time of the year when conditions are less than ideal for examining tonal contrasts.

Table 3 Suggested Sources for Aerial Imagery

Data in digital format are preferred for use in MnModel mapping because the final geomorphic models must be ArcInfo coverages. Using digital data as sources will reduce mapping errors. Several readily available digital spatial databases are currently suitable for the degree of detail sought in MnModel mapping. It is anticipated that even more high resolution spatial data will become available in the future. Key sources of digital data files are listed in Table 4. Although federal and state agencies are the original sources for many of these files, most of these data are widely available through various Internet sites, either for free or for a fee. Most federal and state natural resource data for Minnesota can be downloaded from the DNR GIS Data Deli. The Minnesota Geographic Data Clearinghouse (mngeo.state.mn.us/chouse), managed by the Land Management Information Center, is another good source. When using a combination of digital spatial data, be well aware of the source scales, methods of conversion, and limitations of each data set. When acquiring digital spatial data, also be sure to acquire its metadata as well.

Table 4 Key Digital Spatial Data Files

*not yet available for all of Minnesota

Subsurface information is available from a variety of sources (Table 5) besides the published and unpublished literature (Table 1), but the utility of this information varies with both the source and the analyst's familiarity with local stratigraphy and sediments. Most subsurface records consist of bore or well logs taken for geotechnical purposes. Such logs usually do not contain the detail required to make fine distinctions within deposits of Holocene age. Subtle buried soils that might be significant to landscape suitability modeling, but insignificant to engineering concerns, usually are not recorded. However, major texture breaks are recorded in geotechnical logs, and sometimes these have regional stratigraphic and/or temporal significance. Ultimately, there is no substitute for the quality of subsurface information that can be generated during project-specific field investigations.

Table 5 Sources of Subsurface Records

Published, comprehensive, state-wide databases of late Wisconsin and Holocene radiometric ages for Minnesota do not exist at this time. Although analysis and interpretation of primary contextual data will always be necessary, currently, radiometric information on the ages of LfSA's and the history of landscape evolution has to be gleaned from the published and gray archaeological and geologic literature, or generated in the field phase of a mapping program. Some relative ages can be derived from the literature, but most come from original assessment of geomorphic and stratigraphic relationships, and associations among prehistoric cultural deposits and landform sediment assemblages. Some degree of familiarity with the geology of the project area will assist in interpreting geologic contexts of relevant archaeological radiocarbon ages gleaned from the literature.

Fieldwork that generates original subsurface data is the most important source of information for developing LSM's. Fieldwork takes place at a later phase of the project, but can yield invaluable data because specific issues can be targeted for investigation. Normally, fieldwork will follow geomorphic mapping so that the mapping can be field-checked. Specific geologic, geomorphic and geoarchaeologic relationships can be examined. LfSA's can be sampled to recover material suitable for radiometric dating.

3.2 Interpreting Source Data

Interpretation of topographic maps (either paper quadrangles or digital raster graphics [DRGs]) in combination with aerial photography and digital orthophotoquads (DOQs) forms the basis for the geomorphic mapping. An overview of generalized late Quaternary stratigraphy and sedimentology for most parts of Minnesota can be obtained from published and unpublished reports and from the DNR GIS geomorphology coverage. Bore logs might provide clues to generalized stratigraphy. NRCS county soil survey publications may provide detailed information on the sediment textures and soils of the uppermost five or six feet of LfSA's. However, the applicability of this information to LfSA mapping is only as good as the accompanying soil maps (and their potential limitations have already been mentioned above). Fieldwork is the primary source of reliable subsurface information. Sometimes subsurface information found in the literature, bore logs, or soil reports can be "calibrated" based on results of project-specific subsurface fieldwork. General soil-geomorphology can be interpreted from analyzing the relationships among NRCS mapped soil series and different LfSA's.

Both absolute and relative temporal information are critical for constructing histories of landscape evolution and interpreting LSR's for LfSA's. Intact prehistoric cultural deposits will occur within deposits that are neither too old nor too young to contain them. Minnesota archaeologists mostly use 12,500 B.P. as the earliest age for occupation. Therefore, it is possible to establish a "basement" below which the potential for intact prehistoric cultural deposits can be excluded. This "basement" in Minnesota is either bedrock or all but the latest Wisconsin glacial deposits. An indication of the age and distribution of different "basement" materials can be obtained from the literature and Minnesota Geological Survey maps and publications. At the other end of the spectrum, anything younger than 200 B.P. is considered "post-Euroamerican Settlement" in age. LfSA's that are younger than this age are interpreted in part from historic maps and surveys, Government Land Office surveys, and early aerial photography. LfSA's that are post-Euroamerican settlement in age also tend to have distinct sediment and soil characteristics in the field. Yet they are too young to contain intact prehistoric cultural deposits.

Relative ages of LfSA's can be determined by interpretation of stratigraphic superposition and sedimentary and geomorphic cross-cutting relationships. Elevational geomorphic relationships for some types of landforms also are used to determine relative age. These types of relationships are established by documenting topographic positions on aerial photographs and topographic maps, and in the field in outcrop exposures, as well as on the basis of geomorphic positions relative to one another.

Absolute temporal information is mostly in the form of radiocarbon ages determined on organic material. Radiocarbon ages may be found in the published and unpublished literature, particularly in the archaeological literature. However, the full context of the dated material is not always documented, so that some data are limited in their utility, while others may require some interpretation in the context of other available or interpreted information. Caution should be employed when using existing radiocarbon data for additional reasons. For example, the available technology and state-of-the-art protocols used at the time of analysis may differ greatly from those employed today. Also, the type of organic material, if recorded, may also influence the weight accorded the age of the sample. Preferably, radiocarbon ages are generated as a result of strategically targeted field collections for the current LSM project.

Return to the Top

4. EVALUATE USEFULNESS OF SOURCE DATA

Evaluate the usefulness of source data for constructing the LfSA GIS map and database. There are several concerns that make some data more useful than others. Chief among these are whether: 1) the available data are in, or can be translated into, digital format; 2) the source data scale approaches the 1:24,000 scale of the model; 3) the source data resolution is adequate; and 4) the source data limitations are acceptable. All must be weighed prior to mapping.

4.1 Weigh Digital Versus Non-digital Data
Given a choice, reliable digital data will always be the preferred format. The GIS relies on digital data layers, and the more reliable layers that are in the database, potentially the more powerful the modeling capabilities. There are few digital databases available at an appropriate scale. However, the ones that are available are quite useful for our purposes. DRG's of USGS 7.5-minute topographic maps provide a base for mapping and the land surface configuration. DOQ's at the same scale provide digital aerial photographic coverage. DLG's provide hypsometric information, but the resolution and internal database problems limit their usefulness at this time.

The DNR GIS geomorphology coverage is useful for determining the geomorphic region and subregion of a project. However, the boundaries were mapped at a scale of 1:100,000 or 1:250,000 and are not accurate at the 1:24,000 scale.

Some USDA-NRCS county soil surveys are available in digital format, but as of this writing, they are not yet available for all of Minnesota. Of those that are available digitally, some are available in vector format (ArcInfo coverages) while others are available only in raster format (EPPL7). County soil surveys can be digitized for a project area, but the effort might be hard to justify. Whether digital or hard copy, some soil surveys are modern while others were developed using out-dated standards. Also, some are mapped on rectified, orthogonalized aerial photographs and others are not. The user must be aware of the quality of the soil survey data and its implications for spatial and attribute data accuracy. The information soil surveys provide is significant, but there is a large amount of "play" for most soil series boundaries with adjacent soil series. An approximated boundary line will likely be as useful as the actual published line.

USGS NAPP aerial photography is not yet available in digital format. However, it may be economically feasible to scan, rectify, orthogonalize, and register photos to construct a digital layer for use with the GIS data.

4.2 Compare Scale of Potential Source Map Information to the 1:24,000 Scale of Mapping
The standard scale of MnModel LfSA mapping is 1:24,000. There is no particular maximum size limit for polygons mapping a single landform. For example, a uniform outwash plain or glaciolacustrine plain may cover square miles, but the uniformity of the plain dictates it be represented by a single large polygon. However, conceptually it is possible to map polygons that are "too large." In this context, a polygon that is "too large" encompasses a number of discrete landforms representing a wide range of landscape variability. At a minimum, the potential significance to modeling the archaeological site - LfSA relationships will be diluted. On the other end of the spectrum, if mapped polygons are "too small," then the resolution may result in polygons that are miniscule, cumulatively cumbersome, not time- nor cost-effective, and that may snap together at the 1:24,000 scale. At a scale of 1:24,000, the minimum discernible object is about 40 feet by 40 feet (12 meters by 12 meters). However, at 1:24,000, an object this size would appear as a point. The smallest detectable polygon would be about 80 feet by 80 feet (24 meters by 24 meters). This would represent the size of the minimum mapping unit. However, it may be more likely that a project's actual timetable and budget would dictate a minimum polygon size that is slightly larger than the minimum mapping unit.

Compare the scale of information likely to be utilized in mapping on digital and non-digital map sources to the 1:24,000 scale for an appropriate fit. Be aware that the five-fold geomorphic hierarchy (Table 6) recognized in the Field Code Key Table implies the potential utility of databases at scales other than 1:24,000. For example, the hierarchy allows for landform subdivisions to be included (i.e., more detailed polygons; see Code No.9 in the Code Key Table). Also, there may be future need for a more detailed map (larger scale) of a particular area. However, if it is important to map smaller polygons than can be supported at 1:24,000 scale, it is absolutely necessary to develop or acquire base map data and aerial photography at a scale appropriate for such mapping.

Table 6 Geomorphic Hierarchy Used in MnModel Mapping

4.3 Learn the Limitations of the Source Data
Each source data item was developed for a particular purpose. None were developed specifically for LfSA mapping. Understanding why, where, and when the source data were developed will help assure this information will not be over-interpreted or otherwise misused. Such information should be included in the metadata for the digital data. If metadata are not provided with the digital data you acquire, you must request it from the data provider. Although there are state and federal standards for metadata, not all data providers maintain metadata according to these standards. Consequently, the quality of metadata varies widely. Some detective work may be required to find out everything you need to know about your source data.

4.4 Select Source Data to be Used and Create an ArcInfo Coverage
After the above factors have been considered, select the digital and/or non-digital source data that will be used in LfSA mapping. If the existing information is already in a digital form, but not in an ESRI-compatible format, it must be converted using the utilities within ArcInfo or ArcToolBox. Non-digital source data will have to be digitized using ArcEdit to create an ArcInfo coverage. The coverage may also need to be put into a common coordinate system using the TRANSFORM and/or PROJECT command. If non-digital source data are going to be used, proceed to Chapter 6, Sections 6.1-6.13.

Return to the Top

5. LfSA MAPPING AND LABELING

5.1 Determine LfSA Map Units

5.1.1 Establish General Quaternary Geologic Setting
The first step in LfSA mapping is to establish the general Quaternary geologic setting. The general setting will help determine Code Nos. 1-6 (Appendix A) and ultimately will provide some help in determining appropriate polygon size. Review relevant maps and the most recent literature for current interpretations and names. Determine the geomorphic region (Code No. 1), geomorphic subregion (Code No. 3), landscape (Code No. 5), and associated geographic or informal names (Code Nos. 2, 4, 6). The Minnesota Department of Natural Resources GIS Geomorphology coverage (1998), the Quaternary Map of Minnesota (Hobbs and Goebel, 1982) and the volume of papers by Patterson and Wright (1998) currently are some of the most useful references for this task.

5.1.2 Estimate a General Suitable Polygon Size Range for Mapping
A number of factors will help determine what polygon size is appropriate, reasonable, or possible. (Also see Section 4.2 of this report.) The project objective and limitations will dictate to a large degree what can be mapped successfully. Map as much detail as possible at this scale for geoarchaeological purposes. As a result, polygon sizes generally will be smaller than mapping for most other geologic purposes. Polygon size must be greater than the minimum mapping unit. However, it is most likely that budgets, schedules, source data scale, and final product scale will be the determining factors that limit the degree of detail. The regional geomorphic and landscape settings dictate what landforms are present and will provide some indication about the scale and topographic complexity of those landforms. Some LfSA's, such as in stagnant ice landscapes, will tend to be smaller and more complex compared to LfSA's in active ice landscapes. In such cases, a number of LfSA's may need to be lumped into a single LfSA, particularly if individual polygons would be smaller than the size of the minimum mapping unit. The quality of available imagery from which to map and for cases where tree cover is extensive, the degree of detail in the soil maps might also limit to some degree the size ranges of polygons. Become familiar with the USGS 7.5-minute topographic maps of the project area. Begin to formulate what landforms are present and what distinguishes them. Weigh the aforementioned factors. Those with limited mapping experience may want to practice mapping a portion of a quadrangle to see if they have selected a reasonable range of polygon size.

5.1.3 Determine the LfSA Map Units
This is the first estimate of actual units that will be mapped. Study the topographic maps, available imagery, and soil maps to determine what landforms are present and what landforms will be differentiated. Code No. 7 of the Field Code Key Table contains a list of mutually exclusive landforms commonly encountered as elements of Minnesota landscapes. Comments regarding definition and usage are included for some of them for additional clarity. The more common and easily recognized landforms have less additional information. This list is not intended to be complete, nor are the definitions of particular landforms immutable. The list was designed for easy amendment. However, any landform additions to Code No. 7's list, or alterations in landform definitions, must be careful not to infringe on past or potential definitions in other parts of the state and MnModel. (Caution: you may be changing another mapper's definition.) Review the geomorphic, material/material sequence, and temporal attribute sets in the Field Code Key Table for items that might be used to distinguish additional meaningful LfSA's. Note that the LfSA (Code No. 7) is not, with two exceptions, defined on the basis of temporal attributes. The exceptions, embodied in Code Nos. 25 and 26, are where geomorphic relationships allow relative age relationships to be established, such as with multiple valley terraces.

5.1.4 Establish Criteria for Placement of Lines that Delineate LfSA Polygons
Each landform will be delineated based on a series of criteria, whether formally recognized or not. Regardless of what the criteria are, it is important that they be applied consistently, both within a project, and to the degree possible, among different projects. A landform will mostly be delineated on morphologic criteria expressed by topographic lines. For all map units (except hillslopes proper), when an LfSA boundary is defined by a slope, the boundary line is placed at the base of the slope rather than midway down a slope or at the shoulder of a slope. Landform boundaries sometimes may correspond with tonal contrasts exhibited on aerial photography or with USDA-mapped soil series boundaries. For example, severely eroded areas (Code No.13) may require distinction. Soil maps might also be used to distinguish one or more sediment sequences beneath a given landform (Code No. 15). Interpretation of soil maps might suggest that a significant overlying deposit (Code No.18) is present and needs to be distinguished. Distinguishing different surface characteristics or modifications (Code No. 11) within a single landform type may be desirable.

5.2 Delineate and Label Polygons

5.2.1 Draw Polygons
There is no prescribed method for physically outlining the polygons, and the method will vary with the means by which the map is entered into the digital realm. If the map is to be digitized on a digitizing board, polygons are outlined on a USGS 7.5-minute topographic map using a thin, soft lead pencil. Criteria from other non-digital map sources will have to be scaled and transferred to the map. This may be done, for example, by drawing polygons on a piece of mylar overlain on the source image and scaling the mylar on a copy machine to 1:24,000. The data is transferred in pencil to the quadrangle map by placing the quadrangle over the mylar on a light table and tracing the lines. These steps introduce a number of errors, and it is difficult to transfer information from hard copy sources accurately. The preferred method would be to map directly into the GIS.

If "heads-up" digitizing will be done, that is, mapping by Geologist 1 directly into the GIS on the computer screen, then polygons are outlined using an input device such as a mouse. Follow Sections 6.4 through 6.19 only after referring to Sections 5.2.2 through 5.2.4 below.

5.2.2 Establish a System of Labeling Polygons that Reflects Criteria for Mapping
Base the labeling system on the Map or Code String Symbols in the Code Key Table. Although there is no set formula for constructing map symbols, systematically basing them on the Code Key Table will ensure a large degree of compatibility among projects. The map symbol consists of a string of alphanumeric characters that reflects key geomorphic and other Code Key attributes. Ideally, the map symbol is intended not only to identify polygons, but also to convey as much information about the polygon in as compact a form as possible. To accomplish this, alphanumeric characters are ordered to correspond to the attribute order in the Code Key. Minimally, the map symbol consists of a landform symbol (Code No. 7), such as "T" for terrace. A landscape symbol (Code No. 5) will precede the landform symbol. A fluvial (Valley) terrace landscape is "V" (as opposed to a glaciofluvial, "O," or lake, "L," terrace); hence the map symbol would be "VT." If the project area encompasses more than one geomorphic region or subregion, symbols representing these attributes (Code Nos. 2 and 4) will precede the landscape symbol. A reach of a valley terrace draining the St. Croix end moraine ("ST," Code No. 4) will have a map symbol of "STVT." A single landform may be subdivided into one or more LfSA's, if there is some distinguishing characteristic, such as sediment sequence or relative geomorphic position. In these cases, the map symbol for the distinguishing attribute or attributes will follow the landform symbol. For example, a reach of the previous example terrace, where it is underlain by sand ("SA," Code No. 15), can be differentiated from where it is underlain by clay ("CY," Code No. 15) by using the map symbols "STVTSA" and "STVTCY," respectively. Note that if the area being mapped abuts and joins a pre-existing MnModel map, then care should be used to make the Map Symbols consistent. Review map symbols from adjacent or nearby MnModel projects.

5.2.3 Evaluate Validity of LfSA Map Units and Map Unit Criteria
Situations will invariably arise where it will be necessary to evaluate the initial selection of landforms and their defining criteria. Additional LfSA map units may be required, or two map units should more appropriately be lumped together. Initial criteria used to define map units may not accommodate all subsequently encountered scenarios. Reassessment of map units may be required; for example, if the polygon size is so small that map presentation problems may occur. If problems occur, then return to Section 5.2.1. If new map units are created, the Geologist 2, or other designated person, must update the Code Key Table and deliver the updated version with the final product (See Sections 6.19 and 11.5).

5.2.4 Monitor for Continuity in Application of Criteria
Make certain the criteria used to define LfSA map units are applied consistently from the start of mapping to the completion. If not consistent, then return to Section 5.2.1 above. Document these criteria, both for yourself and for MnDOT. Make adjustments and corrections to previously mapped areas to account for any changes in the defining criteria. Determine whether these changes will affect areas mapped for previous projects. Any alterations to previously mapped project areas must be accompanied by a report explaining the alterations made and the reasons for those alterations.

5.2.5 Label Each Polygon
If maps are constructed on USGS 7.5-minute topographic maps, place a pencil label in each polygon for the digitizer to reference. Each polygon label is very important because it will be used to link the project LfSA database and GIS attribute database to the mapped polygon. Each LfSA map unit has a unique map symbol, and each polygon is labeled or tagged with one of these map symbols. Again, you may wish to check other nearby LSM's to make sure that the same map symbols are used for the same LfSA's.

5.2.6 Review Map Line Work
Compare mapped polygons with map data upon which they are based. Check for consistency in the application of mapping criteria. Verify that all polygons are closed. Make certain there are no dangling lines. Make certain line-work edge-matches where a polygon crosses onto more than one map. Also make certain line-work edge-matches along adjoining project boundaries with areas previously mapped for MnModel. Add or remove polygons to fulfill mapping criteria and assure correspondence with map data upon which mapping is based.

5.2.7 Review Polygons for the Presence and Accuracy of Map Symbols
Make certain each polygon has a map label by systematically examining each polygon. Confirm that each map label is accurately listed with the correct symbol and is displayed in the correct alphanumeric order. Verify that there is one and only one label in each polygon. Correct any map label errors or omissions.

5.2.8 Digitize Hardcopy LfSA Maps
Submit the hardcopy LfSA maps to the GIS Technician for digitizing (Chapter 6). As maps are being digitized, Geologist 1 can proceed with assembling the project LfSA Database (Chapter 7).

Return to the Top

6. DIGITIZING PROCEDURES AND STANDARDS

Useful non-digital source data and project LfSA maps (see below and Chapter 7) will have to be digitized. Standards and general instructions for digitizing are found in the following sections. Prior to digitizing LfSA maps, follow the steps below and in Chapter 7 regarding LfSA mapping, labeling and attribute assignment. Digitizing can be accomplished on either a digitizing board or directly on screen through "heads-up" digitizing. In either case, it is strongly urged that digitizing be supervised by a GIS professional. Although digitizing can be done in CADD software and converted to GIS format, the CADD operator and digitizing supervisor must be well aware of the requirements and standards for digitizing for GIS.

6.1 Bring Digital Source Data into Conformance with Project Standards
So that all digital source data can be used together in a project, they must be brought into conformance with the MnDOT GIS standards. Non-digital source data and LfSA maps, once digitized, also must conform with the standards. Data must be in the Universal Transverse Mercator (UTM) projection/coordinate system, Zone 15 extended, horizontal units as meters, NAD83, GRS80, no y-shift. Any variation from this will cause inaccurate registration of different sources. To project your data to these standards, you must first know what projection, coordinates, datum, spheroid, and units you received them in. This information should be contained in the metadata. It may also be in the projection file of an ArcInfo coverage. If not, you must ask the data provider.

6.2 Register Hardcopy Source Data
Identify points on the hard copy map or aerial photograph that can be used to register the map to an existing digital file in the UTM coordinate system (see Section 6.1 for specifications). These points may be the corners of a USGS quadrangle, township, range, or section corners, or even road intersections. Whatever features you choose, you must be able to create an accurate master tic file of these features from an existing digital data source or from a file of x,y coordinates known to be associated with these features. A minimum of four tics are needed, but more are preferable. Tics should be spaced as evenly as possible and encompass the entire extent of the area to be digitized.

If your source map is paper, it is preferable to trace the data to a stable (mylar) sheet before digitizing. Under no circumstances should you digitize from folded or wrinkled maps. Tape the hard copy map to a digitizing board and register the map using the master tic file. The map must be registered each and every time it is mounted on the digitizing board. The GIS software will compare the coordinates that are entered by the digitizer with those in the master tic file to determine the amount of root mean square (RMS) error. This process may have to be repeated several times in order to achieve an acceptable RMS error. The RMS should not exceed 0.004 inches, or its equivalent measurement in the coordinate system being used.

6.3 Register Digital Images
If you are performing heads-up digitizing, you must first register your digital images or photographs to a known coordinate system. This involves finding three or more points on the source photograph or image with three or more corresponding georeferenced points on a digital basemap that has known coordinates. The more points you use, the more accurate your results. Moreover, like the master tic file described above, the points used for registration should be as evenly spaced as possible and fully cover the extent of the image.

The image or photograph is then rectified or "rubber-sheeted" using these sets of points to fit the known coordinate system using a polynomial equation. The spatial accuracy of your results will be a function of the accuracy of the points used and the inherent accuracy of the original digital image. "Rubber sheeting" a digital orthophoto using this method should result in little distortion. "Rubber sheeting" a digital image that is neither rectified nor orthogonalized may result in greater distortion, and there could be an irregular level of accuracy. This process should be performed with commercial software designed for use with remotely sensed data and GIS.

6.4 Maintain an Archive File
If you are jumping here from Step 5.2.1, then proceed to the fourth paragraph below.

Make two black-and-white copies of the original non-digital source data. For hardcopy LfSA maps, make two copies of the quadrangle map(s) with penciled polygons and map labels. Keep one copy as an unmarked backup and use the other in the digitizing and tagging process. Archive the source quadrangle or analog file as an original in case the copies are accidentally lost or destroyed.

Create new versions of the digital file in the normal process of digitizing polygons, building topology, tagging labels, and joining and dissolving map sheets. This is part of the default requirements of many Arc commands, as a built-in safety feature. It is wise to archive copies after each major step in case the need to backtrack arises.
If you are coming from Step 5.2.1 and performing heads-up digitizing, then make and maintain a copy of the digital source file through an automatic archival program. Copies of the digital geomorphology files should be made on a periodic basis, and in turn maintained through an automatic archival program.

6.5 Develop Minimum Digitizing Standards
Digitized lines should be smooth enough so that they create curvilinear forms at the scales they will be viewed and used. Figure 2 compares acceptable and unacceptable digitizing standards at 1:20,000-scale. Note the jagged line-work in the unacceptable digitizing example caused by fewer digitizing points. However, too many shape points will create very large GIS files. Lines need to be more or less smooth at a minimum scale of 1:12,000 to 1:15,000-scale for 1:24,000-scale source data It is preferable to give this assignment to an experienced GIS Technician.

6.6 Digitize Data
Digitize polygons from paper quadrangles, other paper-source originals, or DRGs (if doing heads-up digitizing) one map-sheet at a time. Digitize in sections within each map-sheet. This allows for an easier assessment of completed and uncompleted areas. A Public Land Survey System (PLSS) section on a USGS quadrangle, for example, is a typical feature reference that may be used for this process. Polygons on paper quadrangles or other analog sources are traced using a puck on a digitizing board, whereas polygons on DRG's or other digital sources are traced directly on-screen using a mouse or other similar input device. To do on-screen or "heads-up" digitizing, zoom into the area that you are to digitize at a scale equal to, or greater than, the scale of the source data you are using. For example, if you are using a USGS digital quadrangle (DRG), an appropriate scale may be between 1:12,000 and 1:24,0000, depending on the complexity of the shape being digitized and the quality of the digital image base. As polygons are entered into the GIS, keep track by marking off digitized polygons on one of the paper map copies.

Polygons will represent LfSA's when mapping two-dimensionally. It may be desirable to digitize lines representing geomorphic trends that appear on aerial photography within or across LfSA polygon boundaries. Polygon and line features should be digitized on separate layers in the GIS. Similarly, if at some time point source data is to be entered into the GIS coverage, it will require a separate layer. For other non-digital source data, polygons may represent one of any number of entities.

6.7 Check Whether Digitized Lines Match Source Data Lines
GIS Technician and Geologist 1 check line work for correspondence between the original source data, whether it is the LfSA map or some other non-digital source map, and the digitized version of the map. Prepare a hardcopy transparency of the digitized version of the map, at the scale of the original source map. Overlay it on the original map on a light table. Inspect for divergence or omission of polygon lines between the two maps. Reverse the order of the two versions and repeat the inspection. If a paper quadrangle is used, each polygon may be checked or highlighted as it is digitized. If a digital quadrangle is used, a draft map can be plotted to help with this process.

6.8 Build and Verify Topology
When digitizing is complete, execute the CLEAN command in ArcInfo or ArcToolbox to build topology and snap vertices to close polygons. Be sure that tolerances for the CLEAN command are small enough to prevent inadvertent collapse of arcs comprising small polygons. After topology has been built, check for node errors (dangling arcs). These will help you identify unclosed polygons or overshoot arcs. Additional edits may be necessary to assure that all polygons are closed and overshoots are removed. Each polygon must also have a label point onto which an attribute will later be "tagged" or coded. These label points can be added during digitizing or by using the CREATELABELS command after topology has been verified. After creating labels, BUILD the coverage and check for label errors. There should be only one polygon with no label point and no polygons with multiple label points. If any errors are reported, they must be corrected by hand editing.

6.9 Tag Digital Polygons with Labels
Select and tag polygons with the appropriate map code (i.e., map labels on a paper copy equal a map code in a GIS database). Make black-and-white copies of the original source map or a plot from the heads-up digitizing. As polygons are tagged and given a code in ArcEdit or ArcMap, they are checked off on a corresponding black-and-white copy or draft plot. The digitizer reviews all polygons to ensure they have been tagged with a code.

6.10 Check for Edge Matching of Lines and Polygons
Lines must edge-match between different maps (digital files) and polygons must be closed. The GIS Technician uses the EDGEMATCH command to check areas along all adjacent map sheet edges (digital files) for edge-matching errors and should force arcs on adjacent sheets to move into the snap tolerance.

6.11 Join Map Sheets
When multiple map sheets are involved, join the map sheets into one by using a MAPJOIN command for all of the source data map sheets. Check areas along former map sheets for edge-matching errors and correct if necessary. Verify that polygons crossing map boundaries have the same attribute code on both sides of the boundary. Issue a DISSOLVE command to remove any remnant or extraneous boundaries. Visually review the results to be certain that all lines that should have dissolved did so and those lines that should have remained were retained. Be aware that the DISSOLVE operates on the entire coverage and may remove lines between identically coded polygons away from the boundaries as well. This could result in the merger of miscoded polygons with adjacent polygons. Review the results of the DISSOLVE command. Verify that all lines along the map boundary were removed. Check again for node and label errors.

Only visual inspection by someone familiar with the source material will identify such errors. For LfSA maps, the Geologist confirms that the DISSOLVE command did not introduce any edge errors. Examine the joined digital map along former map edges with the individual maps. Also, if adjacent areas were mapped previously for MnModel, make certain that polygon lines match adjacent model polygons. Also look for lines that did not dissolve. If these are found, check the labels in the two halves of the polygon to determine if they contain the same code. It is likely they will not, so determine which code is correct before directing the GIS Technician to remove the line and the label point with the incorrect code.

If the LfSA map is the digitizing project, then proceed to Chapter 7; otherwise, continue to Section 6.12.

6.12 Attach Source or New Data Attributes to GIS Coverage Attribute Table
The last step is to attach the source or new data attributes to the newly digitized source or LfSA data coverage or shape file. For the specific case of LfSA maps, attach the project LfSA database (see Chapter 7) to the GIS coverage to create the GIS coverage attribute table. The common link (join item or primary key) between the GIS coverage and the project LfSA database is the map symbol or alphanumeric code that was tagged to each polygon as the key field.

To export the LfSA database spreadsheet from MS-Excel, first check that every column matches the MnDOT field definition standards. To do this, right-click on each column heading and then first choose the "Format Cells" option, followed by the "Column Width" menu option.

Next, choose "Save As" under "File," and then select DBF 4 (dBASE IV) (*.dbf) option under the "Save as type" choice. Now the file may be used in ArcView if needed, but is best exported to an INFO format. To do this, in ArcView choose the "Export" option under "File," and then pick INFO as the "Export Format" in the "Export Table" menu. Click OK.

To verify that the original table has been exported to the proper format, check the item definitions by issuing the Arc command ITEMS. Next, to verify that the codes (values) have maintained their integrity, review the final INFO table using ArcView, INFO, or ArcMap. Once this quality control step has been done, it may then be joined to the geomorphology polygon attribute table using the Arc JOINITEM command with the MAP_LABEL field as the key field or relate item.

6.13 Proof the GIS Coverage Attribute Table
The GIS Technician checks that all fields have an entry for each Map Symbol. Confirm that the Map Symbol and intended attribute values were entered correctly in each intended field. Verify, if pertinent, that:

• all GIS code symbols were entered accurately;
• LSR's and attributes upon which they were based were entered accurately; and
• all LSR rankings were calculated correctly.

The most effective and efficient way to verify these values is to utilize the summarize functions and query capabilities of ArcView or ArcInfo. Using these functions, determine if all codes are valid and correctly entered, if combinations of codes are logical, and if calculated fields were correctly calculated.

For LfSA coverage, after proofing and correcting any errors, the entire GIS coverage file (attribute table and spatial data) is sent to Geologist 1.

If additional non-digital source data has to be digitized, repeat Sections 6.1. through 6.13. If no additional non-digital source data needs to be entered into the GIS coverage, proceed to Section 5.1.1. If "heads-up" digitizing of the LfSA map is in progress, then proceed to Section 6.14.

6.14 Digitally Check Each LfSA Map Unit for Labeling
The digitizer makes certain each polygon was tagged with a map label. Uncoded polygons can be identified by searching the coverage's attribute table for missing values. Confirm that the map label is accurately entered in the attribute table. For LfSA coverage, Geologist 1 confirms that the map label was entered in the GIS attribute table with the correct map symbol displaying the correct alphanumeric order. Geologist 1 cross-checks and confirms the integrity of the project LfSA database table, and that the digital list of LfSA map units provided by the GIS technician to Geologist 1 in Section 5.13 is the intended list of map units originally provided by Geologist 1 to the same GIS technician in step 7.8. If polygons were tagged with an incorrect map label, or "misspelled" map label, they will appear in the GIS coverage attribute table, usually with only one or a limited number of polygons represented on the map. They should also be missing values in other fields.

6.15 Digitally Check Each LfSA Map Unit for Location and Geologic Continuity
Geologist 1 performs successive queries for each unique Map Symbol, viewing the highlighted selection set on the digital map. Confirm that the location of each polygon makes geologic sense in relation to both other polygons of the same LfSA map unit and polygons of adjacent LfSA map units. If a polygon does not fit its context, confirm that it was not mislabeled. If there are only one or two polygons of a given LfSA map unit, determine whether that LfSA map unit truly warrants a separate map unit or can be combined with another LfSA map unit. Finally, confirm that all polygons and LfSA map units express local, regional, and MnModel geologic continuity. Geologic continuity in this sense refers to a situation where all geologic relationships are plausible and likely and all LfSA polygons are labeled such that they are congruent with the current model of landscape evolution for the project, as well as regional models.

6.16 Make Necessary Adjustments
Compile a list of required adjustments and corrections based on the results of quality control Sections 6.11, 6.13, 6.14 and 6.15, and forward it to the GIS Technician. Geologist discusses changes with the GIS Technician who will be making the changes to the GIS spatial coverage and attribute database. The GIS Technician resolves questions raised by the Geologist using solutions provided by the Geologist quality control procedures. Digitize omitted lines, remove extraneous lines, and smooth excessively choppy lines so that the digitized and original versions of the map coincide. Close any open polygons and resolve any dangling lines. Match any mismatched lines after multiple maps are dissolved into one map. Correct any map label errors or omissions.

6.17 Geologist Reviews Adjustments
Adjustments made in 6.16 are reviewed and confirmed by the Geologist who provided the GIS Technician with the list of changes. If additional questions arise as a result of Section 6.16 regarding map content or other matters that can not be resolved by the GIS Technician, they are referred back to the Geologist. Highlight problem areas on a copy of the original map. Geologist 1 returns answers or solutions to questions or comments that the GIS Technician raised, and the GIS Technician makes the necessary changes. Any changes are reviewed and confirmed by the Geologist.

6.18 Geologist 2 Cross-Checks
Geologist 2 serves as a cross-check on the work of Geologist 1. After Section 6.17, the GIS Technician transfers the GIS coverage file to Geologist 2. Geologist 2 repeats quality control Sections 6.11, 6.13, 6.14, and 6.15. Geologist 2 compiles a list of any required adjustments and corrections, discusses changes with Geologist 1, and forwards the list to the GIS Technician. Sections 6.16 and 6.17 are repeated.

6.19 Update and Maintain Code Key Table
Geologist 2 enters any new values resulting from the project into to the Field Code Key Table under the appropriate Code Number. Enter corresponding GIS Code Symbol, Map or Code String Symbol, and any germane Comments for each new value. Assign a new version number to the Code Key Table, and keep a record of the specific changes and date of those changes. The updated Code Key Table is in a PDF format and is included as an appendix in the final report. Model testing may proceed (see Section 8.1).

Return to the Top

7. ASSEMBLE THE PROJECT LFSA DATABASE

LfSA attributes for a project, including LSR's, are recorded in a project LfSA database. The project LfSA database is presented as a table in the final project report. It serves as a support tool while the mapping is being accomplished and an easy hard copy reference for a summary of LfSA's recorded in a project. The project LfSA database is developed in a spreadsheet program, such as MS-Excel. Once finalized, the project LfSA database is linked to the GIS database. MnDOT maintains a master LfSA database that is augmented by LfSA's newly defined on a project-by-project basis. This master LfSA database should be consulted at the beginning of a new project; and applicable, previously defined LfSA's should be used where appropriate.

LfSA attributes, including LSR's, are recorded in the project LfSA database. The table consists of records, with each record corresponding to an LfSA map unit. Each record consists of a map symbol and three code strings or identifiers summarizing the LfSA attribute, a series of fields corresponding to the LfSA attributes listed in the Code Key Table, and a series of LSR attributes.

7.1 Code the Map Symbol Field
This field consists of the LfSA map symbols used to label or tag polygons (Figure 3). Enter these new map symbols into the spreadsheet. The spreadsheet allows for easier initial editing by the geologists. Although there is no prescribed order in which the map symbols are listed among themselves, experience indicates the most useful is to list them by landscape, in the order that the landscapes are ordered in the Code Key Table, Code No. 5 (Appendix A). In general terms, this order is from oldest to youngest-the usual order for discussing evolution of a landscape. For example, Active Ice LfSA's would be listed in the database before Valley Terrace LfSA's.

7.2 Code the Geomorphology Set of Attributes
Thirteen fields describe properties of the land surface (Figure 4). The landform, which corresponds with the fundamental LfSA map unit, is one level in the geomorphic hierarchy used to define an LfSA and establish its regional context (Table 6). The geomorphic hierarchy is based on observations of the land surface at different scales. All but the most detailed level have a corresponding attribute consisting of an "identifier," or common name, where one conventionally has been used for that particular feature (Table 6).

For each LfSA, determine the most appropriate value for each field based on the lists of valid values under each Code No. 1 to 13 in the Code Key Table (Appendix A). If an appropriate value for a Code No. is not listed, define and add a new value under the appropriate Code No. in the Code Key Table. If an appropriate value for any field can not be determined at the time of mapping, use "No distinction made." The project LfSA database can always be modified at a later date. Enter the GIS Code Symbol for each attribute in the appropriate field of the database.

7.3 Code the Material/Material Sequence Set of Attributes
Determine the most appropriate value for the fields for the Material/Material Sequence set of attributes, Code Nos.14 to 21 in the Code Key Table (Appendix A). If an appropriate value is not listed, define and add a new value under the appropriate Code No. in the Code Key Table. If an appropriate entry can not be determined at the time of mapping, use "No distinction made." Enter the GIS Code Symbol for the selected attribute in the appropriate field of the database (Figure 5).

7.4 Code the Temporal Set of Attributes
Determine the most appropriate values based on the analysis of all available temporal data. If an appropriate value is not listed, which is likely to be the case since some of these values will be unique, define and add a new value under the appropriate Code No. in the Code Key Table. If an appropriate value can not be determined at the time of mapping, use "No distinction made." Enter the GIS Code Symbol for the selected attribute in the appropriate field of the spreadsheet (Figure 6).

7.5 Code the Summary Code Strings
For each of the three sets of the LfSA attributes, there is a Summary Code String or Identifier, consisting of a string of the Map or Code String Symbols separated by periods (Figure 3). Each of the three strings is arranged in the order corresponding to the order of attributes in a record. The summary code strings provide a quick way of looking at the characteristics of a particular LfSA map unit and comparing with other LfSA map units. They also serve as cross-checks on the entry of GIS Code Symbols. Use the Map or Code String Symbol to construct the three Summary Code Strings. Note that absolute ages, Code Nos. 28 to 33, are recorded in the Summary Code Strings without the "tens" column. Enter the Summary Code Strings in the appropriate field of the database.

7.6 Code the Geologic Age Set of LSR attributes
Based on the Temporal set of LfSA attributes, determine whether the age of the LfSA falls within or without the recognized time span that Pre-Contact peoples lived in Minnesota. LSR's can vary based on differing stratigraphic conditions within a single LfSA. Hence, LSR's are determined for surface conditions (0 m) and three depth intervals: 0-1 m, 1-2 m, and 2-5 m (Figure 7). For an explanation of the significance of the particular depth intervals, see Hudak and Hajic (1999). If the LfSA age for a particular surface or depth interval falls outside of the valid time span (i.e., 12,500-200 B.P.), then enter a "0" in the appropriate field. If the LfSA age falls within, then enter a "1" in the appropriate field of the database (Figure 7).

7.7 Code the Depositional/Post-Depositional Environment Set of LSR Attributes
Estimate the degree to which the energy conditions, drainage conditions and other factors would have affected landscape suitability for occupation and preservation of prehistoric cultural deposits. Use the same land surface and three depth intervals mentioned in Section 7.6 above (Figure 7). Base the estimate on an integrated interpretation of sediments, soils, stratigraphy, and geomorphology, including but not limited to the Geomorphology and Material/Material Sequence sets of LfSA attributes. Determine whether the LfSA land surface has been disturbed such that in situ prehistoric cultural deposits would or would not be preserved. At this point, we do not include plowed surfaces in this consideration, although they are clearly disturbed. At some point, incorporation of digital coverage of cultivated ground (past and present) within an LfSA model would be desirable. If the land surface is disturbed, enter a "0" in the appropriate field. If it is undisturbed, enter a "1" in the appropriate field. For the three depth intervals, enter in the appropriate field a number between "0" and "3" that corresponds to the estimated conditions. A "0" means conditions were unsuitable to contain and/or preserve prehistoric cultural deposits. A "3" means conditions were the most suitable to contain and preserve prehistoric cultural deposits intact (Figure 7). Refer to Hudak and Hajic (1999) for a more detailed discussion of the meaning and use of the numerical assignments.

7.8 Calculate and Enter LSR's
The LSR is defined as the product of the Geologic Age attribute and the Depositional/Post-Depositional Environment attribute. Multiply these attributes together for each respective land surface and depth interval to arrive at the corresponding LSR. For each LfSA map unit, enter the LSR's for the land surface and three depth intervals in the appropriate fields in the database (Figure 7). Although this can be done by hand, it is more efficient to use the spreadsheet to calculate these values for each column based on the values of the Geologic Age and Depositional/Post-Depositional Environment sets of LSR attributes previously entered.

7.9 Proof Project LfSA Database
The Geologist 1 checks that all fields have an entry for each Map Symbol. Confirm that the Map Symbol and intended attribute values were entered correctly in each intended field. Verify that:

• all GIS code symbols were entered accurately;
• LSR's and attributes upon which they were based were entered accurately; and
• all LSR rankings were calculated correctly.

Confirm that the summary code strings were entered correctly and that they correspond with the GIS code symbols in the corresponding record. The aforementioned checks have been performed visually in the past. A more effective and efficient way to verify these values would be to utilize the query and macro abilities of MS-Excel. Another alternative is to export the MS-Excel spreadsheet to dBase format (Section 6.12) and to do the query and summarization functions to detect errors.

Following the check by Geologist 1, the project LfSA database is cross-checked by both the Administrative Assistant and Geologist 2, utilizing the same methods. Corrections are made, and the proofed project LfSA database is turned over to the GIS Technician.

Proceed to Section 6.12.

Return to the Top

8. MODEL TESTING

MnModel geomorphic models and the Code Key Table were designed so that they can be refined through archaeological and geological testing. Construction of the initial LfSA models (Hudak and Hajic, 1999) intentionally were done without knowledge of the ages and locations on the landscape of prehistoric cultural deposits ("archaeological sites"). Rather, site location and temporal data were used as a test against initial LfSA mapping and age assignments. Additional testing is provided by targeted geologic investigation for stratigraphic, sedimentologic, and temporal data for a given LfSA model area. Each new Cultural Resource survey and geomorphic investigation generates information that can be used to test and refine existing models. Both types of investigations can yield information that is applicable beyond the limits of the project for which they are conducted. For example, fieldwork in a project area can result in establishing the age of one or more terraces in a river valley. The terraces can be traced up- and downstream, perhaps to other project areas. The terraces may correlate geomorphically with terraces in other river valleys, thus extending the utility of the original temporal data. See Hudak and Hajic (1999) for more details on testing LSM's.

8.1 Landform Sediment Assemblage Testing

8.1.1 Conduct a Field Review of Landform Sediment Assemblage Mapping
Drive all accessible roads through the project area. Compare the landscape as it exists in the field with the various mapped LfSA's. Give special attention to LfSA's that were problematic, or where there was some question regarding LfSA boundary position.

8.1.2 Confirm the Location and Verify the Composition of Landform Sediment Assemblages
Design and conduct a strategic subsurface coring program to address specific issues or questions that arise during mapping. Questionable stratigraphic relationships can be explored at this time. Variability within and between LfSA's can be documented. The program may involve gathering and documenting cores, examination of stream cutbanks and other outcrops, or a combination of the two. Any exposures should be described and, if necessary, sampled using standard soil techniques and terminology (Soil Survey Staff, 1992). All described cores, borings or sections should have an accompanying graphic sediment soil log constructed using MnModel standards.

8.1.3 Date Specific Landform Sediment Assemblages
The subsurface program can and should be designed to collect datable material from as many relevant LfSA's as possible. Direct dating provides a check on the ages of LfSA's generated from existing sources while at the same time enhancing the model. At a minimum, radiocarbon samples must be cleaned, weighed, and taxonomically identified prior to submission. Submit radiocarbon samples to an established, reputable laboratory.

8.2 Archeological Database Testing

8.2.1 Cross-check Location of Archaeological Sites Against LfSA Polygons
Digitally overlay known archaeological site locations on the LfSA model. Examine contradictions, if present, in age and location of archaeology sites versus the expectations of the model. Determine whether contradictions are due to problems with the site data or LfSA model database. Review site files of aberrant sites at the SHPO and State Archaeologists Office (SAO) for temporally diagnostic artifacts, approximated site age, and geomorphic location. Analyze site data contained in reports and publications. For those sites identified as aberrant, verify the location of the site centroid by comparing 1) the SHPO GIS database, 2) the GIS LfSA spatial coverage, 3) the DRGs, and 4) if applicable, the pencil-lined topographic quadrangles on which the geomorphic polygons were originally recorded. Archaeology sites were recorded in the digital state site file by centroid, rather than actual site boundary, and sometimes the centroid fell outside of the actual site limits. If contradictions are not explained by issues related to the record of the site centroid, consider these possible sources of contradiction:

• Vague site location data (exaggerated site boundaries, imprecise or unknown site location etc.);
• Skeptical site location data (site either not in situ, or incorrectly mapped site location);
• Scale issues (map size; degree of smoothness of digitized line; unresolvable LfSA differentiation);
• Incorrectly calculated UTM's for site centroids;
• Misinterpreted LfSA age or polygon boundary;
• Digitizing errors; and
• Unknown reasons (reason that can not be resolved without field checking).

See Hudak and Hajic (1999) for more details on this type of testing.

8.2.2 Cross-check Time Span of Possible Human Occupation Against LfSA Polygons
Review available age data for in situ archaeology sites within the project area from SHPO, SAO, and the literature. Create a confidence ranking for the sites (high, moderate, and low) based on the amount of available temporal information, degree of site integrity, and quality of site investigation and documentation. Compare the absolute and relative ages of the sites that are assigned high and moderate levels of confidence to the independently determined geologic age of the LfSA within which each site is located. Since intact archaeological sites can not be older than the geomorphic surfaces upon which they rest, occurrences of old sites on young surfaces should signal a potential problem in the age interpretation of either the LfSA or the archaeology site.

Return to the Top

9. DATA MANAGEMENT AND PRESENTATION

9.1 Create Geologic Cross-sections Using MnModel Standards
Geologic fieldwork often results in a series of described cores and/or geologic sections, preferably arranged in meaningful transects. Draw geologic cross-sections utilizing graphic sediment soil logs and correlation lines using Microstation CADD or compatible software and MnDOT-specific cross-section templates provided by MnDOT. Attempt to match (as closely as possible) the formatting and fonts of all graphics and labels to the currently approved MnModel standard. MnModel's first suitable landscape projects used a Simplex-type font on mostly 11x17-inch or 8.5x11-inch drawings.

9.2 Create a Shapefile of Core Locations and Code Attributes to Allowing Future Hot-linking to the Core Logs
Core locations are derived from UTM (Universal Transverse Mercator) coordinates collected during fieldwork. Create an ArcView point shapefile from an initial event theme generated using these coordinates. Add a field to the point shapefile attribute table and populate it with pathnames and filenames that reference core logs in textfile format. These path and filenames will allow the user to hot-link the core locations to the core log text files. Consult with the MnDOT GIS Technical Lead for current path names.

9.3 Create a Shapefile of Cross-sections and Code Attributes to Allow Future Hot-Linking to the CADD Cross-section Drawings
Cross-sections are determined after the ArcView point shapefile is created for the soil core locations. Create a line shapefile of cross-sections using the core locations and cross-section drawings as a guide. Add a field to the line shapefile attribute table and populate it with pathnames and filenames that reference cross-section drawings in Microstation CADD format. These path and filenames will allow the user to hot-link to the mapped cross-section lines to the cross-section CADD files.

Return to the Top

10. FINAL MODEL QUALITY CONTROL

10.1 Document Model Discrepancies After Model Testing
Record and discuss in the final report (see Section 11.6 below) any model discrepancies revealed by model testing. Offer any possible explanations that account for the discrepancies, and how existing and possible future discrepancies might be addressed or considered in future LfSA mapping, either in the same area or a different area.

10.2 Make Adjustments to Address Model Discrepancies After Model Testing
Make appropriate adjustments to the model structure, model results, or modeling procedures for those discrepancies caused by errors in LfSA mapping. Bring archaeological database errors to the attention of those responsible for maintaining that part of the archaeological database.

Return to the Top

11. PREPARE A REPORT THAT DOCUMENTS THE LANDSCAPE SUITABILITY MODEL

Each landscape suitability model requires a report be prepared to document the model. The report is considered to be a preliminary geomorphology report of a project area. It is "preliminary" because it generally is written with limited, if any, subsurface fieldwork. Also, it often is written prior to any archaeological survey work.

11.1 Introduction
Include the purpose of the report, the project goal, and any relevant background information.

11.2 Objectives
List the key objectives of the project. For MnDOT, this will include mapping and describing the geomorphology and LfSA's, assessing the age, stratigraphy and depositional environments of LfSA's, and estimating landscape suitability rankings for intact surface and buried prehistoric cultural deposits.

11.3 Overview of the Project Area
Review relevant literature. Describe the regional geologic and geomorphic setting with reference to figures of project location and boundaries, Quaternary geologic setting, and area(s) mapped for LfSA's. Review the general late Quaternary geologic history for the region and project area. Provide a general description of the project area and note any particular significant features.

11.4 Methods
Summarize mapping, field, GIS, and quality control methods. Refer to a figure that illustrates any core and cross-section locations and other relevant information. If descriptions of cores or sections were generated, include them in, and refer to, an appendix.

11.5 Landscapes and Landforms
Describe the results of the LfSA mapping in the project area. Refer to the project LfSA database and include it as an appendix. Describe and discuss each landform in relation to its definition, geographic distribution, morphologic, stratigraphic, and sedimentologic characteristics, variability of LfSA, soil-geomorphology, other LfSA's, age, and any known archaeological sites. Refer to the project GIS spatial data coverage and other relevant figures (geomorphology, stratigraphy) to illustrate the aforementioned relationships. Any new attributes that are introduced as a result of the project need to be identified, defined and added to the Code Key Table. Include the updated Code Key Table as an appendix. Any new LfSA's need to be identified for addition to the master LfSA database.

11.6 Project Area Geology and Landscape Suitability Rankings
Describe the specific LfSA's found in the project area. Include the results of any subsurface investigation, radiocarbon assays, and other analyses. If radiocarbon ages were generated, present in table format the results of the assay(s), laboratory number, sample identification, content, weight, and LfSA context. Refer to figures of graphic sediment soil logs, cross-sections, and other relevant information. This segment may involve detailed discussion of landscape evolution with an accompanying figure portraying the landscape at different times in the past. Landscape suitability rankings are presented and discussed. Particularly significant quality control measures and results also are presented and discussed in this section. In particular, concordance and discrepancies among ages and locations of archaeological sites and LfSA's are documented and discussed.

11.7 Conclusions
Summarize results of the investigation. Where appropriate, include summary LfSA or stratigraphic diagrams.

11.8 Recommendations
Recommend additional geoarchaeologic or geomorphic investigations or suggestions that can guide or improve the results of subsequent archaeological fieldwork and interpretations.

Return to the Top

12. REFERENCES CITED

Hobbs, H.C. and J.E. Goebel
1982 Geological Map of Minnesota; Quaternary Geology, University of Minnesota, Map S-1.

Hudak and Hajic
1999 Landscape Suitability Models for Geologically Buried Precontact Cultural Resources. In A High Probability Predictive Model of Precontact Archaeological Site Location for the State of Minnesota, pp. 12-1 - 12-283 + Appendix E.
Minnesota Department of Transportation CD-ROM report and GIS ArcView database.
Minnesota Department of Natural Resources

1998 Working copy of Statewide Geomorphology coverage. GIS ArcView coverage at 1:100,000 scale. Unpublished.

Patterson, C.J. and H.E. Wright, Jr., editors
1998 Contributions to Quaternary Studies in Minnesota, University of Minnesota, Reports of Investigation 49.

Soil Survey Staff
1992 Keys to Soil Taxonomy, Sixth Edition, 1994.

13. ACKNOWLEDGEMENTS

The authors wish to thank Dr. Elizabeth Hobbs and Mr. Craig Johnson at the Minnesota Department of Transportation for their constructive review comments of an earlier draft.

Appendix A

MnModel Geomorphology Field Code Key Table