Exposure Surface Analysis for Assessing Relative

Visual Vulnerability and Aesthetics


Joseph K. Berry and Martin, Chris

Geography Department, University of Denver

2050 East Iliff Avenue, Denver, CO 80208-0183



Viewshed analysis has been used for years in assessing what can and can’t be seen from a location. Early computer techniques solved complicated geometric equations to determine line-of-sight connectivity from a single viewpoint. More recently, grid-based procedures use surface analysis to determine visual exposure from sets of points, lines or polygon features in a fraction of the time. The information generated identifies a visual exposure density surface of relative connectivity from viewer locations to all other locations in entire project area. The procedure considers viewer characteristics, distance, exposure angle, terrain, and intervening conditions. Sensitivity analysis enables resource managers to address “what if” scenarios and determine visual impact alternatives. This paper discusses the important concepts, considerations and practical issues surrounding visual exposure assessment and its integration into decision-making. A visual vulnerability project for the Black Canyon of the Gunnison / Curecanti National Recreation Area (Colorado) serves to illustrate the procedures within a practical context. 




Visual considerations are an increasingly important component in effective natural resources planning and management. While visual analysis capabilities have been part of the GIS toolbox since the 1970s, they only have been sporadically used in operational settings. In part, the lack of use can be attributed to the unavailability of a comprehensive elevation data set and the inaccessibility of required computer power.


With the advent of the National Elevation Dataset (NED) and ever increasing computer power, times have changed. Whereas manual mapping procedures are too tedious and inaccurate for widespread use, computer-assisted visual analysis is simply a click away. The lack of a familiar manual procedure as a conceptual reference, however, has clouded its application. Many resource managers are uncomfortable employing a “black box” button without at least a conceptual understanding of how it works and verification of its accuracy. Once this initial hesitancy is overcome, visual analysis likely will become as commonplace in forest management as the Biltmore Stick was in the 1930s or the pickup truck in the 1950s.




The algorithm for determining visual connectivity uses a series of expanding rings to determine relative elevation differences from the viewer position to all other map locations. Elevation differences that are less than those in previous rings are not seen.


Figure 1. Calculating a viewshed.


The top portion of figure 1 illustrates the procedure. The ratio of the elevation difference (rise indicated as striped boxes) to the distance away (run indicated as the dotted line) is used to determine visual connectivity. Whenever the ratio exceeds the previous ratio, the location is marked as seen (red); when it fails it is marked as not seen (grey).  


To conceptualize the procedure, imagine a searchlight illuminating portions of a landscape. As the searchlight revolves about a viewer location the lit areas identify visually connected locations. Shadowed areas identify locations that cannot be seen from the viewer (nor can they see the viewer). The result is a viewshed map as shown draped over the elevation surface in figure 1. Additional considerations, such as tree canopy, viewer height and view angle/distance, provide a more complete rendering of visual connectivity.


Figure 2. Calculating a visual exposure map.


The top portion of figure 2 shows the viewsheds from three different viewer locations. Each map identifies the locations within the project area that are visually connected to the specified viewer location. Note that there appears to be considerable overlap among the “seen” (red) areas on the three maps. Also note that most of the right side of the project area isn’t seen from any of the locations (grey).


A visual exposure map is generated by noting the number of times each location is seen from a set of viewer locations. In figure 2 this process is illustrated by adding the three separate viewshed maps together. The resulting visual exposure map in the bottom of the figure contains four values—0= not seen (grey), 1= one time seen (red), 2= two times seen (green) and 3= three times seen (blue)—forming a relative exposure scale.


Figure 3. Calculating a visual vulnerability map.


The top portion of figure 3 shows the result considering the entire road network as a set of viewer locations. In addition, the different road types are weighted by the number of cars per hour. In this instance the total “number of cars” replaces the “number of times seen” for each location in the project area.


The effect is that extra importance is given to road types having more cars yielding a weighted visual exposure map.  The relative scale extends from 0 (not seen; grey) to 1 (one car-location visually connected; dark green) through 12,614 (lots and lots of visual exposure to cars; dark red). In turn, this map was reclassified to identify areas with high visual exposure—greater than 5,500 car-locations (yellow through red)—for a map of visual vulnerability.


A visual vulnerability map can be useful in planning and decision-making. To a resource planner it identifies areas that certain development alternative could be a big “eyesore.” To a backcountry developer it identifies areas whose views are dominated by roads and likely a poor choice for “serenity acres.” Before visual analysis procedures were developed, visceral visions of visual connectivity were conjured-up with knitted-brows focused on topographic maps tacked to a wall. Visual vulnerability analysis makes this information readily available in an objective and detailed format. 




The previous section described procedures for characterizing visual vulnerability. The approach identified “sensitive viewer locations” then calculated the relative visual exposure to the feature for all other locations in a project area. In a sense, a feature such as a highway is treated as an elongated eyeball similar to a fly’s compound eye composed of a series of small lenses—each grid cell being analogous to a single lens.


Figure 4. Visual connectivity to a map feature (Profile Rock) identifies the number of times each location sees the extended feature.   


In figure 4, Profile Rock is composed of 120 grid cells (think lenses) positioned in the center of the project area. In determining visual exposure to Profile Rock, the computer calculates straight line connectivity from one of its cells to all other locations based on its relative position on the elevation surface. Depending on the unique configuration of the terrain some areas are marked as seen and others are not.


The process is repeated for all of the cells defining Profile Rock and a running count of the “number of times seen” is kept for each map location. The top right inset displays the resulting visual exposure from not seen (VE= 0; grey) to the entire feature being visible (VE= 120; red). As you might suspect, a large amount of the opposing hillside has a great view of Profile Rock. The southeast plateau, on the other hand, doesn’t even know it exists.


The lower portion of figure 4 extends on the concept of visual exposure by introducing distance. It is common sense that something near you (foreground) has more visual impact than something way off in the distance (background). A proximity map from the viewer feature is generated and distance zones can be intersected with the visual exposure map (lower-right inset in figure 4). The small map on the extreme right shows visual exposure for just distance Zones 1 and 2 (600m reach). The accompanying table summarizes the average visual exposure to Profile Rock within each distance zone—much higher for Zones 1 and 2 (34.3 and 41.3) than the more distant zones (27.3 or less).


The ability to establish weighted visual exposure for features is critical in deriving an aesthetic map.  In this application various features are scaled in terms of their relative beauty— 0 to 10 for increasing pretty places and 0 to -10 for increasing ugly places.


For example, Profile Rock represents a most strikingly beautiful natural scene and therefore is assigned a “10.”  However, Eyesore Mine wash is one of the ugliest places to behold so it is assigned a “-9.”  In calculating weighted visual exposure, the aesthetic value at a viewer location is added to each location within its viewshed.  The result is high positive values for locations that are connected to a lot of very pretty places; high negative values for connections to a lot very ugly places; zero, or neutral, if not connected to any pretty or ugly places or they cancel out.


Figure 5. An aesthetic map determines the relative attractiveness of views from a location by considering the weighted visual exposure to pretty and ugly places.


The top portion of figure 5 shows the aesthetics ratings for several “pretty and ugly” features in the area as 2D plots then as draped on the 3D terrain surface. The 2D plots in the bottom portion of the figure identify the total weighted visual exposure for connections from each map location to pretty and ugly places.


The overall aesthetic map draped on the 3D terrain surface in the center is analogous to calculating net profit—revenue (think Pretty) minus expenses (think Ugly). It reports the net aesthetics of any location in the project area by simply adding the Pretty_wExposure and Ugly_wExposure maps. Areas with negative values (reds) have more ugly things within their view than pretty ones and are likely poor places for a scenic trail. Positive locations (greens), on the other hand, are areas where most folks would prefer to hike.


Figure 6. Weighted visual exposure map for an ongoing visual assessment in a national recreation area.


Real-world applications of visual assessment are taking hold.  For example, a senior honors thesis project is underway at the University of Denver by a student intern with the National Park Service. The project will develop visual vulnerability maps from the reservoir in the center of the park and a major highway running through the park (right side of the figure 6). In addition, aesthetic maps will be generated based on visual exposure to pretty and ugly places in the park.


The project uses the Park Service’s ArcGIS database to manage map layers. Viewer locations, such as US Highway 50, are converted from vector to raster format matching the digital elevation data (30 meter cell size). In this example, the highway and elevation grid layers are passed to the MapCalc software system for calculating visual exposure (right side of figure 6). Exposure values for this subset of the park ranges from “Not Seen” (light grey) to “Very Exposed” to the highway (dark red). The 3D plot in the lower portion shows the visual exposure map draped on the terrain surface and was generated using the Surfer software.




Visual quality is an important consideration in natural resources planning and management. Understanding the visual impact of a proposed activity, such as a clear-cut, on the landscape is critical. Simply overlaying the proposed area on a visual vulnerability map aids in a quick assessment of the proposal’s overall impact. Subsequent project-level analyses identify specific impacted areas. In a similar manner, a visual aesthetic map can help identify areas with good views that could be incorporated into siting recreational activities, such as a hiking trail. The interplay of visual vulnerability and aesthetics has been too complex for traditional management practices. However, data availability, powerful computers and a new paradigm of spatial processing have converged to make visual analysis an important new tool in a resource manager’s toolbox.


Author’s Notes:


1.      This paper is based on two “Beyond Mapping” columns appearing in GeoWorld, April and March, 2003. Additional information on visual analysis is provided online at www.innovativegis.com/basis/, select the “Map Analysis” online book, “Topic 15, Deriving and Using Visual Exposure Maps.”

2.      Figures in this paper used MapCalcTM software by Red Hen Systems, Inc. An educational CD with online text, exercises and databases for “hands-on” experience in these and other grid analysis procedures is available for US$21.95 plus shipping and handling from www.redhhensystems.com.

3.      Chris Martin’s senior honors project on visual assessment for the Black Canyon of the Gunnison / Curecanti National Recreation Area (Colorado) will be completed in time for a June graduation, 2003 and posted under “Column Supplements” at www.innovativegis.com/basis/. 




Joseph K. Berry is a leading consultant and educator in the application of Geographic Information Systems (GIS) technology and the principal of Berry & Associates // Spatial Information Systems (BASIS), consultants and software developers in GIS. He is the author of the “Beyond Mapping” column for GEOWorld magazine since 1990. He has written over two hundred papers on the theory and application of map analysis and is the author of the popular books Beyond Mapping and Spatial Reasoning. Dr. Berry is a Special Faculty member at Colorado State University and the W. M. Keck Scholar in Geosciences at the University of Denver. He holds a B.S. degree in forestry, a M.S. in business management and a Ph.D. emphasizing remote sensing and land use planning. Website: www.innovativegis.com; Email: jberry@innovativegis.com.


Chris Martin is a senior in geography at the University of Denver and student intern with the National Park Service. Email chmartin@du.edu.