Keywords: USGS Large Scale Topographic Maps, SVG, interactive topographic web maps
Peter's got a master's degree at the Comenius University in Bratislava in 2000 at the Department of Cartography, GIS and Remote Sensing. He also got a master's degree at the University of Nebraska at Omaha. Currently, Peter is completing his Ph.D. studies at the Comenius University in Bratislava. His research focus is oriented towards open-source WebGIS applications (tools for integration and distributed use of geographic information) and modeling of natural phenomena in a GIS environment.
Dr.Peterson is a Professor of Geography and chair of the Maps and the Internet Commission of the International Cartographic Association. He teaches courses in world regional geography, cartography, and geographic information science. Dr. Peterson has served as visiting professor at the University of Washington, Free University of Berlin, University of Hawaii-Manoa, Royal Melbourne Institute of Technology, Technical University of Vienna, University of Applied Technology in Munich, and Carleton University in Canada. He is past-president of the North American Cartographic Information Society (NACIS) and past-editor of Cartographic Perspectives. His books include Interactive and Animated Cartography (Prentice Hall), Multimedia Cartography (Springer), and Maps and the Internet (Elsevier).
United States Geological Survey (USGS) large scale topographic maps at a scale of 1:24,000 have been traditionally distributed in paper form. Scanned versions of these maps are now available through the Internet as raster representations. Instead of common raster format presentation, the solution presented here is based on a vector approach using Scalable Vector Graphics (SVG). SVG provides many advantages compared to the use of a raster-based presentation, such as the quality of the graphical representation, maintenance, actualization, interactivity, and extensibility through other Web programming languages. The purpose of this research is to propose an optimal and logical structure of a SVG document with a minimal file size that would be universally applicable for all USGS large scale topographic maps. This study shows that SVG is a promising technology for delivering high-quality, fully-vector topographic maps via the Internet, both in terms of graphic quality and interactivity.
The approach described here can be applied not only to USGS topographic maps but essentially to any similar types of maps. It also can serve as a basis for the design of a database that would facilitate the distribution of large scale topographic maps in SVG through the Internet.
Link to the project site: http://ptolemy.unomaha.edu/~pavlicko/
2. USGS Large Scale Topographic Maps
2.1 Large Scale Topographic Maps on the Web
3. Online Large Scale Topographic Maps using SVG
3.1 Processing and Conversion of Source Data to SVG
3.2 SVG Document Structure
3.3 Map Symbology
4. Results and Discussion
Large scale topographic maps portray detailed information about the landscape and are used for a wide variety of purposes. United States Geological Survey (USGS) large scale topographic maps have been traditionally distributed in paper form. Scanned versions of these maps are now available through the Internet in raster form. These scanned maps are often of low quality and not very legible. The solution presented here is based on a vector approach. The vector format provides many advantages such as the quality of the graphical representation, maintenance, actualization, interactivity, and extensibility through other Web programming languages. SVG as a new, promising vector format may bring new possibilities for the distribution of topographic of maps through the Web.
This research purposes a method for creating USGS primary series topographic maps at a scale of 1:24,000 using the SVG format. The goal is to create an innovative way for delivering high-quality, fully-vector topographic maps via the Internet, both in terms of graphic quality and interactivity. It will be demonstrated that the Internet can be used to distribute maps in a graphic quality comparable to maps on paper.
However, there are still several limitations in publishing topographic maps over the Internet. The structure of the source USGS Digital Line Graph (DLG) data, in which topographic maps are distributed, is one limiting factor. Source data contains many unneeded properties that “interrupt” graphic quality of the resultant map.
The file size of a SVG document is another important aspect in the distribution of large scale topographic maps over the Internet. Ideally, distribution of whole topographic maps (topo-sheets) would logically represent the best solution. However, the file sizes of such individual topo-sheets may be too big and therefore not appropriate for distribution through the Web at this time. The question is, how large an area depicted on a SVG map would be appropriate in terms of the transmission at this time. This study examines this question. It also tries to define an optimal structure of the SVG document that would be universally applicable to all USGS large scale topographic maps distributed via the Internet in the SVG format.
Interactivity represents an important method that allows users to better extract topographic information from a web map. The number of such interactive features that can be implemented is basically unlimited and can be developed to serve specific applications. Zooming, turning layers on and off, providing coordinate readouts, and displaying attributes through a mouseover function represent basic functions that would allow interactive manipulation with a topographic map.
The approach described here can be applied not only to USGS topographic maps but essentially to any similar type of map. It also can serve as a basis for the design of a database that would facilitate the distribution of a user-defined subset of a large scale topographic map in SVG through the Internet.
Topographic maps portray detailed information about the landscape for both natural and man-made features. They are an indispensable tool for government, science, industry, and leisure for a wide variety of purposes. The USGS (U.S. Geological Survey) provides topographic maps and updates the standard series maps in different scales for the entire United States. The map coverage is completed for large scale maps (1:24,000), intermediate scale maps (1:50,000 and 1:100,000) as well as small scale maps (1:250,000, 1:500,000, and 1:1,000,000). Large scale topographic maps in scale 1:24,000, also known as 7.5-minute quadrangle series, are the most detailed and the best known USGS maps. They represent the only one uniform map series that covers the entire area of the United States in considerable detail. They also are the elementary part of The National Mapping Program (NMP) provided by the USGS. A series of government publications (http://topomaps.usgs.gov) provides all relevant information about topographic maps – where they can be found, including standards, data sources, links, etc.
Figure 1: Example of 7.5-minute quadrangle; portion of Fort Smith, Arkansas [USGS, 2003a] .
Like other topographic maps, large scale topographic maps at a scale of 1:24,000 have been traditionally distributed in paper form. A need to distribute these maps electronically came with the advent of computers and the Internet. The need for data standards became apparent in this process. The USGS, along with other academic, industrial, federal, state, and local government and cartographic agencies, required a standard data format for transferring and exchanging spatial data. The Spatial Data Transfer Standard (SDTS) fulfills this requirement.
SDTS defined by USGS is a "robust way of transferring earth-referenced spatial data between dissimilar computer systems with the potential for no information loss. It is a transfer standard that embraces the philosophy of self-contained transfers, i.e., spatial data, attribute, georeferencing, data quality report, data dictionary, and other supporting metadata all included in the transfer [USGS, 1998] ."
There are two USGS cartographic products in the SDTS standard, vector Digital Line Graph (DLG) and raster Digital Elevation Model (DEM) data. There are several ways/methods to obtain DLG data for 1:24,000-scale topographic maps, either by downloading or ordering. A geodata searching tool EarthExplorer and FTP server provided by USGS represent one of these possibilities.
EarthExplorer - http://earthexplorer.usgs.gov for searching various kinds of digital geodata including raster and vector formats for large scale topographic maps (DLG, DEM, etc.).
ftp://edcftp.cr.usgs.gov/pub/data/DLG/LARGE_SCALE/ - USGS FTP server with DLG files (all currently available layers) for the entire area of the United States. The name of the map needs to be determined through EarthExplorer, Global Mapper, or other source.
Online large scale topographic maps provide many advantages. A user may select a particular area of interest. Such information is available anywhere and anytime if the Internet connection is provided. A user does not have to rely on his/her software that would process and display such information. With implemented interactive features one can even analyze topographic maps online and make quick decisions. There is no question for a practical significance of these solutions. They can be used anywhere where fast access and very detailed spatial information is necessary.
Currently, there are many ways to deliver large scale topographic maps for the Web. Solutions can differ from simple to very sophisticated implementations. Static map images in raster format, even if they are scanned depictions of paper topographic maps or maps exported from GIS to a raster format, represent the simplest form of maps for the Web. They can be embedded either inside an HTML page or disseminated separately as raster images, since a web browser is able to display them. Different raster formats are used for these types of images such as JPEG, GIF, PNG, etc. This solution is, of course, at a lower level in terms of possible interactivity, operational possibilities, dynamic features, etc.
Internet GIS represents an intelligent way to distribute maps via the Internet. Aside from the technologies are used, they can offer sophisticated solutions for dissemination of highly interactive and dynamic maps. A database, where geospatial data is stored, plays a central role in this process. However, these solutions do not have to necessarily represent the best possible way for map dissemination in terms of graphic quality, interactivity, and flexibility. At the present time, as [Neumann, A.] points out, most commercial systems are mainly server-based and use expensive GIS-Mapping-Server/Java-Applet/Plugin combinations. Each interaction with a map requires contacting a database server, which generates answers/feedback for the requests. This process is quite time-consuming and results in undesirable delays when querying or examining objects. Furthermore, the large-amount of client-server traffic burdens map-servers and bandwidth. Solutions based on client-side architecture using open web standards technology like SVG can bring these types of maps to the Web in effective, inexpensive and reliable quality. [Neumann, A.] points out ways in which large scale topographic maps can be delivered using SVG and open-source database technology.
There are several projects for online distribution of USGS large scale topographic maps via the Internet. Topozone.com is perhaps the best known. It uses a server-side architecture and resultant maps are in raster format.
DLG represents the initial source of digital data for USGS large scale topographic maps. SVG is a relatively new format, so such conversion methods have not yet been developed. Because there are no tools available for direct conversion of this file format to SVG, other methods for converting DLG to SVG have to be used.
Most GIS systems import DLG files, including both graphic and attribute data. Once imported, data can be processed, analyzed, and converted to various vector formats. On the other hand, most of current desktop GIS programs also offer export possibilities to SVG. This means that DLG can be easily imported, processed, and exported to SVG using a GIS program.
However, there are still several shortcomings in this process. Geometric properties of input DLG files have inappropriate impact on the resultant SVG graphics. The SVG code produced by GIS programs usually requires modifications and optimizations to the most appropriate document structure. Symbolization of map features, especially for more complex symbols, is also not very well supported. Reasons for using a GIS program for processing source DLG data can be summarized in three points:
Figure 2: Inappropriate properties of DLG objects that have to be eliminated - unneeded number of vertices and breaks (on the left) instead of one solid line (on the right).
The whole process of processing and converting source data to the SVG structure is depicted on the Figure 3 . The figure demonstrates how DLG files, distributed in nine basic thematic layers, have to be imported, processed, and separately exported to the final SVG structure according to their geometric and thematic properties/affiliation. For example, spot heights (point features) and contours (line features) represent the different geometry (point/line) but the same thematic layer - hypsography. The purpose of applying this procedure is to separate individual objects in order to assign their symbology as well as prepare these individual objects for further use. This means creating the logical structure of the final SVG document.
Conversion possibilities for exporting processed data from a GIS program to SVG differs between most common used GIS programs. However, in general they can be divided into three main groups:
“Indirect graphical conversions” are many times inefficient since they loose geometrical properties of the map content during the conversion process. For example, the map could be re-projected to a new unknown projection. Although this might be acceptable for small-scale general reference maps where spatial accuracy is less important, is not appropriate for large scale topographic maps. Moreover, any use of such resultant SVG maps for further applications, such as maintenance of the map content, is practically impossible.
“Indirect conversions” that are based on XML platform using GML and XSL represent a conversion process that could be subject of another investigation in a future. In this study solutions offered by standard GIS software packages were used. Currently, as it was already mentioned, most common GIS programs provide export to the SVG format. Some of them offer more complex solutions for publishing maps on the Web than just a pure format conversion. Using such programs, one can create a fully functional cartographic product. Besides the map content in SVG, other features can be added to the final map such as scale, legend, interactive features, etc.
Although these tools offer good solutions, they still have a lot of limitations. For example, they cannot convert more complex features such as multi-lines (a line that consists of more than one line), and symbols, that is so necessary for the creation of a large scale topographic map according to the standard symbology (see [USGS, 2003c] ).
Once all map objects are encoded in separate SVG files, the process of compacting and creating one solid file can begin. For the creation of the SVG topographic map, the structure of the SVG document has to be defined. The following sub-section discusses the SVG document structure that could serve the purpose of large scale topographic maps. The focus in this process is placed on two aspects, the optimal and logical file structure that reduces the file size to a minimum and also appropriate symbology of large scale topographic maps.
Since the SVG topographic map will be distributed via the Internet, the size of the resultant file should be as small as possible. Therefore, the structure of the code is important as it directly determines the file size. Each topographic map consists of a series of map objects that are represented by symbols. These symbols have to follow the USGS standards. The process of SVG code optimization and minimization is divided into two parts. First, the structure of the SVG document is proposed. The second part discusses the cartographic symbols that have to be created in order to create the final map with the standard symbology.
Many things need to be considered to create a completely vector topographic map in SVG. For example, how map objects will be grouped together, what is the rendering order of the SVG document and how effectively these objects are encoded in SVG need to be given careful consideration. The topographic map contains several thematic layers including of area, line, point, and text features. To avoid unnecessary overlap when objects are rendered, the order of area, line, and point objects among all thematic layers has to be established. According to SVG Specification 1.1, elements in a SVG document fragment have an implicit drawing order with the first elements in the SVG document fragment getting “painted” first. Subsequent elements are painted on top of previously painted elements. Grouping elements such as the 'g' tag have the effect of producing a temporary separate canvas initialized to transparent black onto which child elements are painted [SVG 1.1 Specification]
The proposed SVG structure starts with the assumption that area features have to be rendered first, which means that they also have to be defined first in the document. Line features are then as second, followed by point and text features. Moreover, the area, line, point, as well as text features that relate to the same thematic layer have to be organized in order to provide the proper rendering. The structure of the SVG document has to fulfill certain hierarchical object order (rules) and rendering order to avoid problems with display. It has been proposed as follows:
According to the previous rules, the SVG document would look like the following code:
<g id=”areas”> <g id=”surface_cover_areas”> <g id=”forest”></g> </g> <g id=”hydrography_areas”> <g id=”hydrography_areas_stream></g> <g id=”hydrography_areas_industrial></g> </g> </g> <g id=”lines”> <g id=”hypsography_lines”> <g id=”hypsography_lines_intermediate></g> </g> <g id=”hydrography_lines”> <g id=”hydrography_lines_intermitterstream”></g> </g> </g> <g id=”points”> <g id=”hypsography_points”> <g id=”hypsography_points_spotheights”> </g> </g> <g id=”hydrography_points”> <g id=”streammileage“></g> </g> </g> < g id=”text”> <g id=”hypsography_text”></g> <g id=”hydrography_text”></g> </g>
Figure 4: The proposed general schema of the SVG document
Figure 5 and Figure 6 show what can happen if the rendering and hierarchical object order is disrupted.
Figure 5: Overlapping problem caused by an incorrect hierarchical rendering object order (the river and road in the middle of the figure).
Figure 6: Spot elevation (brown cross) as a point feature is overlapped by line features (section line).
Another aspect that deals with the SVG code optimization is about possibilities for its minimization. Suggestions proposed by [Neumann, A. and Isakowski, I.] leads to the intensive code minimization. Probably the biggest impact causing minimization of code is to use relative instead of absolute coordinates, reusing of geometry when applying two or more different styles to one object, using symbols together with centrally defined styles (symbology).
The symbology is an important part of topographic maps. Symbols for 1:24,000-scale quadrangle maps are defined in two technical documents, Publication Symbols Part 5 and 6 6 (Supplementary Symbol Specification) that are related to Primary Series Quadrangle Standards [USGS, 2003c] These define symbol names, technical parameters like number, size specification, color and remarks on how symbols should be used correctly. Figure 7 presents an example of a symbol for a primary highway.
Based upon these standards, it is possible to create the same symbology for map objects that are found in the paper version of topographic maps. SVG offers many advantages for the creation of symbology for 1:24,000-scale topographic maps. Symbols can be applied to map objects two different ways. According to a level of symbol complexity, symbology or style can be applied to a map object either directly by styles or using predefined symbols. The application of the SVG “style” to map objects as SVG elements is not a complicated process, and there are a variety of ways to handle such definitions. SVG offers four methods of applying style to elements:
Direct application of style (using the first two methods listed above) allows for the quickest method to test the results of style application, whereas style sheets (internal or external) allow for cleaner, more organized documents. Since this effort is directed towards smaller file sizes for the SVG document, CSS styles provide a better solution. They make the document clearer and reduce the amount of code.
One of the most useful features in SVG is the ability to establish an object (or group of objects) as a symbolic group (‘symbol’ tag). By doing so, this “symbol” can be referenced multiple times throughout the document by ‘use’ tag. This is useful especially for point features. For example, a point feature like a church or spot height is defined as a symbol and then it can be later referenced as many times as needed.
For complex map features, such as multi-lines which consist of several different lines with different styles, a method suggested by [Neumann, A. and Isakowski, I.] ) leads to a very effective solution. To create a multi-line, they propose to re-use the geometry of a single line. This is useful for instance in case of roads (class1 and class2) where lines consist of two or three styles. Figure 8 displays the code and corresponding graphic example:
Figure 8: Multi-line composed of three styles as proposed by [Neumann, A. and Isakowski, I.]
Symbology for USGS large scale topographic maps can be defined universally as a symbol library and used for any map of this type. This can be done by an external CSS file. Names of symbol classes defined in CSS should follow National Mapping Program Technical Instructions – Part6, Publication Symbols [USGS, 2003c] " and the particular number under which specific symbol can be found. Names should also include their geometric (like area, line, point or text), and thematic affiliation (hypsography, hydrography, and so on), based upon which the rendering order in the SVG document would be consider. This could be an example of a symbol style:
hp300line - where hp refers to the thematic affiliation (Hypsography), 300 refers to its number in the “National Mapping Program Technical Instructions – Part6, Publication Symbols”, and line refers to its geometric property.
Interactivity represents an important method that allows users to better extract topographic information from a web map. The number of such interactive features that can be implemented is basically unlimited and can be developed to serve specific applications. Zooming, turning layers on and off, providing coordinate readouts, and displaying attributes through mouseover functions represent basic functions that would allow interactive manipulation with a topographic map.
The architecture of the web page with the embedded SVG map, project title, text, and navigational bar is based on XHTML. The navigation bar consists of checkboxes that represent all thematic vector layers and raster shaded relief. The layers can be separately turned on/off according to user needs. UTM coordinates display plane X and Y coordinates in the UTM projection. The Z-coordinate depicts the actual elevation in meters based on an underlying DEM as it was proposed in [Neumann, A. and Isakowski, I.] study. Zooming, panning, and resetting to the original view are controlled through the Adobe SVG Viewer functions. Small icons which are enlarged on mouse-over provide the instructions on how to use zoom in, zoom out, pan, and returning to the original view. Under Additional Info, a user can find the link to the SVG legend. In this legend, all symbols that are used in the SVG map are explained and classified into the thematic layers.
Figure 10: USGS topographic map 1:24,000 web page: http://ptolemy.unomaha.edu/~pavlicko/
File sizes for whole topo-sheets in SVG (all thematic layers) would vary, depending on the information content of the map (number of contours, etc.). In order to investigate SVG file sizes, research was conducting on several 1:24,000-scale topographic maps from Wyoming, Montana, and Nebraska. These SVG files were all between 3 to 4 MB in size. Such large files are not appropriate for the Web at this time. We propose that, at a maximum, approximately only one quarter of a topo-sheet is appropriate for presentation through the Web. To distribute topographic maps for a larger area of a one topo-sheet or encompassing many map sheets using SVG, it would be necessary to create a database that could distribute small parts of an individual topographic map on demand. Methods for delivering larger geographic datasets for large scale topographic maps are proposed by [Neumann, A.]
Figure 12: The area of the SVG map project (grey color) on the 7.5-minute quadrangle map (North Omaha, NE). The area represents about 1/4 of the map.
This study has demonstrated that Scalable Vector Graphics (SVG) is a promising vector technology for delivering high quality interactive topographic maps via the Internet. Vector formats, in comparison to their raster counterparts, provide many advantages such as the quality of the graphical representation, maintenance and actualization, interactivity, and extensibility through other web programming languages. The corresponding implementations for the distribution such maps in raster form leave much to be desired in the graphic quality as well as interactivity.
There are few studies dealing with distribution of large scale topographic maps over the Internet in vector form, using either open-standard or proprietary technology. In this aspect, this work can be an inspiration for the future research. It has been shown that USGS 7.5-minute quadrangle maps can be successfully implemented and disseminated with SVG. SVG proves that it might be a perspective format in the future for the distribution of this type of map. The graphic quality and manipulation possibilities with such SVG topographic maps is far superior to the raster implementations.
However, there are still some limitations to SVG that have to be taken into the consideration. The size of SVG files for the distribution of large scale topographic maps is a certain limitation. According to the results, only portions (subsets) of 1:24,000-scale topographic map sheet can be distributed, with some interactive elements. To distribute topographic maps for a larger area or encompassing many map sheets using SVG, it would be necessary to create a database that could distribute small parts of an individual topographic map on demand.
Link to the project site: http://ptolemy.unomaha.edu/~pavlicko/
XHTML rendition created by gcapaper Web Publisher v2.0, © 2001-3 Schema Software Inc.