By Lena Weber, Ph.D., GIS Applications Manager, C.T. Male Associates

Catskill Mountains from the home of artist Frederick Church (from the collection of Olana State Historic Site, New York State Office of Parks, Recreation and Historic Preservation).

Olana State Historic Site, the 250-acre National Historic Landmark near Hudson, New York, is in itself a work of art. This was the home that Frederick Church (1826�1900), among the foremost artists in the Hudson River School of painting, created from farmland he purchased in 1860. While landscapes from Olana were often the subject of his paintings, his main Olana artwork consisted of the views themselves, carefully constructed to present nature as artistic arrangement.

The Olana Partnership (TOP), the nonprofit support arm of the Olana State Historic Site, decided in 2004 to conduct a GIS-based viewshed analysis from multiple viewpoints within the site as proposed development of surrounding properties threatened to impair the historic vistas. Several studies had already been done: in 1986, 19 nearby properties were evaluated to establish viewshed protection strategies; in 1996, a study assessed potential visibility of new structures to the south and southeast of Olana.

In 2004, TOP wanted a complete 360-degree viewshed analysis from multiple viewpoints on the site to assess viewshed impacts of any proposed development within a four-mile radius.

Using ArcReader and the viewshed model results, an example parcel (outlined in yellow) is visible from viewpoints 3 and 4, even with the reconstructed historic landscape case.

TOP called the GIS Group of C.T. Male Associates, P.C., in Latham, New York, to conduct the viewshed analysis because of the company's viewshed analysis experience using ArcGIS with the Spatial Analyst extension. Also, in addition to hard-copy maps, C.T. Male proposed delivering an ArcReader project to allow TOP staff digital access to the study results. The analysis itself required multiple viewshed calculations, using topographic information plus historic landscape data. As with any viewshed model, field testing was required.

In approaching the project, the first step was to collect and collate relevant GIS data:

• United States Geological Survey 7.5-minute 10-by-10-meter pixel resolution digital elevation models (DEMs)
• Historic landscape documentation
• Selected viewpoints
• Balloon sites for model calibration

The study area was a four-mile radius centered on the Olana house, consistent with the 1996 study. Additional GIS data included roads, town boundaries, tax parcels, and aerial photography to show the visible/nonvisible viewshed in relation to other geographic features.

Model Building

Scanned map showing historic landscape reconstruction and the six viewpoints. The green markings refer to reinterpretations of where to draw the tree lines for the viewshed.

Viewshed analysis was performed using ArcGIS (ArcView) with the ArcGIS Spatial Analyst extension. DEMs were converted to grid format as the primary calculation layer, for both viewsheds and the three-dimensional "hillshade" effect. On the viewshed maps, the hillshade layer shows the relationship between terrain and visibility. The model was calibrated using estimated viewpoint locations, and preliminary viewshed calculations were mapped for fieldwork.

Balloons were flown to a height of 100 feet on a day with good visibility and minimal foliage. TOP staff suggested general balloon locations, and C.T. Male's landscape architect selected specific sites in the field based on accessibility and visibility as predicted by preliminary viewshed analysis. From each Olana viewpoint, staff attempted to see the balloons with and without field glasses. Balloon and viewpoint locations were GPS surveyed.

Field visibility was then compared with GIS calculations. Inconsistencies could be explained by vegetation; in several cases, viewpoints were repositioned to accommodate landscapes. GIS calculations in ArcView produced viewshed grids from each viewpoint, using two-meter observer height and the topographic grid.

Viewshed Calculations

Overlooking the Catskills from the Olana house site (Photo by Christa Hay, C.T. Male).

When producing viewshed calculations, many factors need to be considered. For instance, winds bend trees, seasons change vegetation volumes, structures are built or demolished, trees grow or are destroyed by lightning and logging, and flash floods erode slopes. Nevertheless, GIS analysis using easily available data is the most objective and robust method of viewshed calculation. Obstructions close to the observer have more impact. Obstructions at a distance must be higher than the views that are behind them. Therefore, the viewshed calculation using only topographic features, while the most conservative, is also the most robust. The following are some limitations to this type of viewshed calculation:

• Mathematically, model trees act as a constant-height impervious wall and may cause viewshed calculations to understate visibility.
• Some atmospheric or man-made landscape obstructions are not considered in the calculations, such as leaf-on/leaf-off seasons, deciduous/coniferous tree classification, crown height, bushiness, lateral spread of individual trees, clouds, and fog.
• Observers are assumed to have perfect vision and their position remains fixed.
• For this study, six viewpoints were selected. Other off-site areas may be visible from additional places on the Olana site.

Project Results

A set of maps titled Olana Topography (no trees) Viewshed was produced, showing the color-coded viewshed grid and the various values of the visible landscape when only the topography of the earth's surface acted to impede visibility. To include historic landscape information, treed areas, shown on a document entitled Olana Historic Reconstruction Plan 1996 by landscape architect Robert Toole, were digitized and given a height value of 18 meters. This was added to the DEM as input for historic viewshed calculations. These two sets of maps were delivered along with a report containing sufficient technical detail so that the project can be reproduced.

Olana viewshed from viewpoint 3 (Olana House) using only topographic features to calculate visibility. Pink colors indicate areas visible from the Olana House viewpoint.

ArcReader was installed at TOP's Olana office, along with an ArcReader application that contains both sets of viewshed grids, labeled roads, tax parcels, base DEM data, and 2004 aerial mosaics for the study area. This mapping application, with read-only access to the data, has much GIS functionality and no significant learning curve. All of these data layers overlay each other and can be turned on/off. The staff can zoom to any specific tax parcel�located by tax parcel ID number or street address—and view its potential visibility from each viewpoint under each viewshed scenario.

TOP staff could then make an initial determination on the viewshed impact of any specific construction within the four-mile radius area. Proposed developments will fall into one of four categories:

• Well within the visible portion of the historic landscape viewshed from one or more viewpoints
• Well within the visible portion of the topographic viewshed from one or more viewpoints
• Not visible from any viewpoint
• Near the viewshed boundary

In the first three cases, technical complexity has been removed from viewshed impact decisions. Only the last case requires more research.

Conclusion

The visibility of the entire study area can now be determined scale-independently, without incurring more consulting costs. As residential and commercial development pressures continue to escalate, the need to balance historic preservation with economic growth places greater and greater pressure on TOP. This cost-effective methodology enables TOP to recognize and preserve Church's historic panoramas; for developers, instead of costly site-specific research, viewshed impacts are assessed in advance and alternative "smart-growth" solutions get a head start.

More Information

For more information, contact Kimberly J. Lamay, director of Administration and Public Affairs, the Olana Partnership (e-mail: kimberlylamay.top@olana.org); Lena Weber, GIS manager, C.T. Male Associates (e-mail: l.weber@ctmale.com); or Christa Hay, GIS specialist, C.T. Male Associates (e-mail: c.hay@ctmale.com). Graphics, analysis, and ideas were contributed by Christa Hay and Pat Collins, C.T. Male Associates, P.C.; and Kimberly Lamay, the Olana Partnership.