| Introduction | Review of Research | Geologic Mapping | Conclusions | References |
Within a specific county mapping project the review of existing maps and related data is used to identify areas needing more detailed investigation. For Butler County, two previously published geologic maps cover the entire county. The State Highway Commission of Kansas (now Kansas Department of Transportation, KDOT) published a set of maps at a scale of approximately 1:80,000 (Hargadine, 1967) as part of a construction materials inventory. The Kansas Geological Survey published the 1:500,000 scale Geologic Map of Kansas (Ross 1991) entirely from digital geologic databases. The databases for the state map were derived primarily from the previously published, 1:500,000 scale, state geologic map (Kansas Geological Survey, 1964) with additional reference to numerous county maps in east-central Kansas. A bibliography of reports relating to Butler County geology can be accessed through the on-line bibliography of Kansas geology or in the published bibliography (Sorensen et al. 1989, Sorensen 1994). An extensive library of aerial photography is available for review at KGS.
Under ideal circumstances the outcrop patterns of some geologic rock units can be identified and located directly from aerial photographs. More generally, the mapping process requires interpolation of the outcrop pattern between observation points which are frequently a mile or more apart. The outcrop pattern developed by the geologist is essentially the line of intersection between a 3-dimensional model of the geologic contact surface between two mapped rock units and the model of the land surface represented on a topographic map. As stated by Sawin (1996, p.3), "geologic maps are compilations of data and inference. . . . The geologist's job [in making a geologic map] is to visualize the bedrock near the surface without the soil cover and make a map that reflects this image."
Because of the small scale at which previously available geologic maps of Butler County were published, they contain extensive cartographic generalizations (particularly in outcrop patterns). To reduce the degree of cartographic generalization, all new mapping in the KGS county mapping program is compiled on U.S. Geological Survey (USGS) 1:24,000 scale topographic maps. These topographic maps are published in quadrangles which extend 7.5 minutes in longitude and latitude.
Figure 1 shows the location of Butler County in Kansas and the 7.5 minute quadrangles which provide topographic map coverage of the county. To develop the new geologic map of Butler County, geology was mapped on sixteen full quadrangles and 20 partial quadrangles, covering approximately 1300 square miles. The Kansas Geologic Map Database is designed so that geologic information is stored in subsets which correspond to these full or partial quadrangles within individual counties.
On all county mapping projects that involve new field mapping, including the Butler County map project, Automated Cartography incorporates an additional step in the preparation of input documents which greatly facilitates the following phase of data capture. This step provides stable, transparent base maps where geologic features can be transferred from the geologist's maps.
A base map is prepared on mylar for each of the 1:24,000 scale topographic maps covering the county. These mylar base maps are prepared at the same scale and projection as the corresponding topographic map, using digital cartographic data from the Kansas Cartographic Database (KCD). KCD is a collection of digital cartographic databases representing geographic features from each of the USGS 1:24,000 scale topographic maps in the state of Kansas. Data from KCD is currently distributed by the Kansas Data Access and Support Center (DASC) at the KGS. Metadata providing a technical summary of KCD can be accessed on-line or paper copies of the complete KCD metadata report (Ross, 1996b) can be obtained directly from DASC.
The mylar base maps are designed with sufficient reference information to insure accurate local positioning of the mylar overlays on each of the geologist's cleaned (paper) geologic quadrangle maps. The standard location reference items included from KCD are: 1) section corner locations and township and range lines from the Public Land Survey System (PLSS); 2) all significant hydrologic features from the topographic map, and 3) major highways and rail lines. In addition, the location of the four corners of the quadrangle are added for use as control points in the subsequent phase of data capture.
As stated, geologic features are carefully transferred from the cleaned version of the geologic field map to the mylar base map. This is done using colored pencils. A color coding scheme indicates the attribute code to be associated with each linear or point feature in the data capture phase. The color code will distinguish between a line representing a geologic contact and one representing a fault. The color will also distinguish between features which are visible and located with the highest degree of accuracy, not visible but inferred from indirect evidence and fairly certain in location, or concealed and estimated only with considerable uncertainty. Completed, color coded base maps are compared with adjacent quadrangles to check for edge matching of features. When a mismatch between quadrangles or omission of a geologic contact or other feature from any quadrangle is found, the geologic maps are returned to the geologist for adjustment. The previous steps are then repeated as necessary.
While several critical steps occur during the preparation of these input documents, this first phase of work by Automated Cartography is not very time consuming. The result, however, greatly improves the efficiency of subsequent phases in the map production process. Some time and cost savings may be possible by going directly from the clean geologic information drawn on paper topographic maps to the data capture phase. However, digitizing geologic information from more cluttered topographic maps is much more likely to lead to errors of omission or misidentification of features in the data capture phase.
Separate geologic databases are created for each quadrangle or partial quadrangle within a county as staff members begin the digitizing process for that quadrangle. The longitude and latitude of the corners of the quadrangle are entered along with the digitized table coordinates of the corners as control points so that all data locations taken from the quadrangle may ultimately be related to their corresponding geographic coordinates. The operator enters an appropriate feature code before each new line or feature is digitized. As previously indicated, the color code on the source document will cue the operator to change feature codes when a geologic outcrop changes from visible to inferred or concealed. Because the mylar base map is a relatively uncluttered document it is possible to place other important notes on the map to direct the operator to take other necessary steps in the data capture process. As in the first phase, the data capture phase is not extremely time consuming, requiring only a few hours (at most) per quadrangle.
This procedure has been selected over alternative methods commonly used for data capture. In stream-mode digitizing a cursor is moved along a feature with automatic collection of location coordinates every time the cursor moves beyond a selected threshold. To insure that an adequate number of points are captured to describe complex features, this procedure results in collection of excessive numbers of points in less complex areas. With scanning techniques the source document is first run through a raster scanning device. The resulting image is then processed to extract vector features (or an operator digitizes features from a screen display of the image), and an operator then codes the resulting features. Unless the source document is clear of all but the desired features, scanning procedures may require considerable effort dedicated to cleaning unwanted features from the raw data.
Automated Cartography's technique for data capture is probably the least sophisticated of the alternative methods. Recent comparisons with results obtained by other state geological surveys using high resolution scanning equipment and sophisticated data processing systems indicate that the methods used by the KGS Automated Cartography Group are significantly more efficient and cost effective, in both labor and equipment, while producing a highly versatile end product (Ross, 1996a).
The data capture phase is completed for each quadrangle as the original digitizing table coordinates (measured when the button is pressed on the digitizing cursor) are converted by the GIMMAP system into a binary database with X,Y coordinates in the map projection and scale (1:24,000) of the original topographic map. In this process the X,Y coordinates of the southwest corner of the quadrangle are given the values [1,1] so that all coordinates in each database will be positive.
At this point in the database conversion process, the GIMMAP system provides an automated test for edge matching errors between quadrangles. Using a workstation, the operator checks errors identified by the system and corrects minor problems in an interactive mode--see Figure 3. Where the error is significant, the project geologist will be consulted. If necessary, one or both quadrangles may be returned to the geologist for revisions. Preceding steps will be repeated for the changed features.
Once the identified errors are eliminated, the GIMMAP system builds polygon topology within the new quadrangle databases. The operator then attributes and labels the polygons in the geology database. Test plots are made of the geologic data to check for consistent and correct attribute codes and labels. An appropriate color scheme for the mapped geologic units is developed at this time. Figure 4 presents a test plot for the Butler County portion of the Rosalia NE quadrangle. The boundaries between different shaded zones correspond to the surface location of geologic contacts mapped in the field. Hydrology has been added to this test plot to facilitate comparison with the aerial photograph of a portion of this quadrangle shown in Figure 5.
Each interval of rock units mapped in the quadrangle has a unique color or shade associated with it on the plot. The selection of colors requires consideration of both the artistic and information science objectives of the finished map. Colors for different mapped units must be distinct enough to ensure easy differentiation of the units on the map. At the same time, colors should not be so uncoordinated and dramatically different as to create distraction or unintended emphasis on a particular unit. The Automated Cartography Group generally maintains similarities between colors used on county geologic maps and the colors which occur within the same county on the 1:500,000 scale state geologic map. The ability to do this is limited by the fact that the larger scale county maps will generally include a larger number of mapped intervals, providing greater resolution between individual units, than occurs within the same county on the smaller scale state map.
The digital layout database includes all of the reference information associated with the geologic map as well as basic format items such as map borders and neat lines. Standard reference items included in the layout database for all county geologic maps include: title, author, date of publication, sponsoring agency names and/or logos, map series number, north arrow, credits and acknowledgments, scale bars and/or statement of scale, diagram of township showing section numbering scheme, township and range identifiers around the map, geographic coordinates of the map corners, identification of bordering counties location index map, key to symbols, key to mapped geologic units, descriptive text, references and disclaimers. If appropriate, the layout will also include revision dates. Some authors also include geologic cross sections in the layout.
The base map data include all of the physical and cultural reference features which will appear as an overlay on the geologic map. These features include hydrology, reference lines (section, township and range) from the PLSS, highways and secondary roads, railroad lines, cities, and a variety of associated labels.
The Automated Cartography Group currently uses two types of color plotters (one based on electrostatic technology the other using ink-jet technology) to produce maps for distribution. Figure 6 shows the Butler County map as it is being plotted on an electrostatic plotter. Plotter technology has become a substitute for traditional press-run printing of published maps. The KGS operates on what has been referred to as 'on demand' production of maps for distribution. The objective is to minimize inventory while delivering a timely, high quality product. While electrostatic plotters were initially preferred because of their ability to handle the extremely complex color plots produced by Automated Cartography, improvements in ink-jet technology (which includes lower material costs and more recent availability of large format plotters) have made the latter technology superior to electrostatic plotters.
There are several advantages to plotting on demand. In addition to reduced inventory maintenance costs, this approach avoids the extremely high first copy costs of traditional printing which requires printing a large number of maps to achieve low unit map costs. Many maps produced by the KGS have a specialized and limited audience which makes such large production runs inappropriate. The current method of production avoids those problems. In addition, digital geologic maps produced on demand have the advantage that needed revisions or corrections identified subsequent to initial distribution can be incorporated quickly and at very low cost without the loss of a large stock of unused inventory.
In addition to distribution of the finished paper maps, steps are also taken to facilitate distribution of the geologic databases associated with the county geologic maps. In an automated procedure, the Survey's GIMMAP system converts the spatial and attribute databases into an Arc/Info format. The databases are then transported to Arc/Info where coverages and relational databases are built and linked. The relational databases contain stratigraphic information, automatically derived from the GIMMAP attribute codes, for each of the polygons on the geologic map. Arc/Info is used to export the data in a wide range of formats, allowing the use of the data in Arc-View and other GIS software.
Return to Geospatial Analysis Symposium
Comments to gina_ross@kgs.ukans.edu