|LiDAR Analyses of Prehistoric Earthworks in Ross County, Ohio|
|Written by William F. Romain and Jarrod Burks|
|Monday, 03 March 2008|
In earlier articles (Romain and Burks 2008a, 2008b), we demonstrated the use of LiDAR technology as applied to the Great Hopewell Road (the Van Voorhis section, 33Li401) and the Newark Earthworks. In this article we head south, down the Scioto River Valley, and show the results of LiDAR imaging for select sites in Ross County, Ohio, a region noted for its many prehistoric earthworks (see endnote 1).
LiDAR is an acronym for Light Detection and Ranging. The concept is similar to radar except that, in the case of LiDAR, instead of radio waves, laser light is aimed at a target and used to determine the distance to that target. In the case of aerial LiDAR, this distance information is combined with GPS (Global Positioning Satellite) data. The result is a precise set of three-dimensional coordinates for any given point on the ground. During a fly-over of a selected area, thousands of laser pulses are bounced off the ground, and the buildings and trees on it. The resulting LiDAR data, with some filtering and processing, can then be used to create very accurate surface maps. Because of the data density and scale of the surveys, LiDAR-generated surface maps offer the possibility of showing patterns on the ground that might not otherwise be visible to the naked eye. As we demonstrate in this article, the implications of this technology for prehistoric earthwork identification and interpretation are enormous.
Hopewell Mound Group
Appropriately, we begin our tour of Ross County earthworks with Hopewell Mound Group. The embankments, ditches, and many mounds of this site cover well over 100 acres along the north bank of the North Fork of Paint Creek, about 15 miles upstream from the Scioto River. Figure 1 shows a map of the Hopewell site as published by Squier and Davis (1848:Pl. X), those intrepid 19th century gentleman-archaeologists who, in the span of just a few years, surveyed and gathered maps others had made of 80-plus Ohio earthworks.
A LiDAR image of Hopewell Mound Group is provided in Figure 2. In this image, and others in this report, vertical heights are exaggerated to better show relevant features (see endnote 1). The surface model in Figure 2 was generated from more than two million LiDAR points. In this figure, higher elevations are shown in red, with lower elevations decreasing through yellow, green, and blue, respectively. The perimeter walls of the main structure are readily apparent, especially the north wall which has not been damaged as much as others are, by plowing.
In Figure 3, morphological details become even more apparent as we zoom-in and change the lighting and perspective. For reference, the red arrow in Figure 1 shows the angle of view provided by Figure 3. As Figure 3 shows, the north and south walls of the Square enclosure are visible in topographic data, though their appearance is somewhat muddled by a variety of more recent linear features (fence rows and/or plow marks) in the same area. This is important because the Square is impossible to see on the ground with the naked eye and has been difficult to detect in geophysical surveys. Also evident are several mounds located within the main earthwork enclosure. The large mound near the top-right of the image in Figure 3 is Mound 25, famous for having been the largest known Hopewell mound (about 500 ft long by 180 ft wide) before it was completely excavated in the 1920s by the Ohio Historical Society (Shetrone 1926).
We indicate (with blue arrows) in Figure 1, that a north-south profile through the earthwork is included in the Squier and Davis map. Figure 4 shows the same profile, generated in this case from the LiDAR data. The outside ditch and low wall that define the north perimeter of the main enclosure are evident in the LiDAR profile. So too, is the mound shown in the Squier and Davis profile—Mound 23, which is one of the few at this site that has not been completely excavated. From the profile data we find that the bluff under the north edge of the main enclosure is about 50 feet higher than the terrace level that the rest of the earthwork is situated upon.
Conspicuously absent from the LiDAR images are the interior enclosures, including the large circle west of Mound 23, the enclosure surrounding Mound 25, and a new, small circle found in 2001 to the northeast of Mound 25 (see Pederson, Burks, and Dancey 2001).
Many other features at Hopewell Mound Group could be commented upon. As this report is intended only as a brief introduction to the use of LiDAR, however, we will move on to the next site.
The Baum earthwork is located in the Paint Creek Valley, on the south side of Paint Creek, just below Spruce Hill and about a mile and a half from the town of Bourneville. Figure 5 shows Squier and Davis’s (1848:Pl. XXI, no.1) map of the enclosure. Figure 6 is a 1964 aerial photo of the site—one of the better aerial views. Looking at the aerial photo alone, one would be hard-pressed to find the geometric enclosure without additional references. A LiDAR image of the same area (Figure 7), however, clearly reveals significant portions of the earthwork.
Of interest is the fact that we can use the LiDAR data to determine the size of the earthwork. Measurement of the Baum Square using the LiDAR data (Figure 8) shows it to be about 1,056 feet in width along the western wall, measuring from the crests of the perpendicular walls (see endnote 2).
In Figure 9, a close-up view of the Baum Square with greatly exaggerated height shows very faint traces of the four gateway mounds that appear on the Squier and Davis map.
The Seip enclosure, another of the tripartite enclosures of Ross County, is located roughly four miles upstream from Baum. As with most other large geometric earthworks, very little of the original earthwork can be recognized by the naked eye at ground level due to the effects of agricultural plowing and erosion. Figure 10 shows the earthwork as represented by Squier and Davis (1848:Pl. XXI, no. 2). Figure 11 is a LiDAR image of the same area. As can be seen, several parts of the original earthwork are visible in the LiDAR image, including sections of the Large Circle, Small Circle, and northeast corner of the Square. The large mound in red is the now-reconstructed Seip-Pricer Mound, while the lower mound to the northeast is the Seip Conjoined mound, both of which have been excavated by the Ohio Historical Society.
Notably visible in the LiDAR image are several old stream channels that appear to have cut into the original boundaries of the earthwork. These channels are shown by the deeper shades of blue. Clearly, flooding is a problem in this area! Whether or not the earthwork is still intact to some degree among these stream channels at the southern edge of the enclosure has yet to be demonstrated.
Mound City is located on the west bank of the Scioto River, three miles north of Chillicothe. The interior mounds and perimeter walls of the site have been restored to the heights measured by Squier and Davis in the 1840s. The configuration of the mounds and the wall are based on excavations at each of the mounds and in the walls in 1920 and from the early 1960s through about 1980. Detailed analyses presented elsewhere also reveal that restoration of the perimeter wall is accurate in terms of its configuration and orientation (Romain 2000:125-129, 2005:3A-8).
Figure 12 shows a recent aerial view of the site. Figure 13 shows the same perspective using LiDAR imaging. The apparent lumpiness in the walls in this image may be caused by interference from vegetation. The linear features crosscutting the enclosure, especially visible in the lower right corner of the enclosure, are old roads that were part of Camp Sherman—a World War I training camp that used to cover a large area on the west bank of the Scioto River, including all of Mound City, from 1916 through about 1920. Camp Sherman is responsible for flattening a large number of earthworks and mounds north of Chillicothe. Paradoxically, it probably also is the reason why Hopewell Culture National Historical Park exists today. This park began in 1923, after Camp Sherman, when the Federal Government still owned the land containing the remains of Mound City. Much of the rest of the federally-owned land in the area was given to the state or returned to private ownership.
LiDAR provides us with the ability to assess earthworks relative to the surrounding landscape. In the case of Mound City, one of the observations we note is that the Scioto River is very close to eroding-away the northeast corner of the earthwork. As shown by Figure 14, a distance of only 50 feet separates the earthwork from the edge of the river-eroded terrace that it is situated on.
One can also see the ‘lost’ borrow pit, (or perhaps it is a deep earth or water feature) that appears on the Squier and Davis map at the northeast corner (upper right) of the enclosure. While all of the other so-called borrow pits have been reconstructed, this one escaped notice until the late 1990s. Excavations in this pit in 1998 by National Park Service archaeologists identified an intact layer of Hopewell debris buried in the pit. Should a strong flood erode the nearby bank, this feature too would be in jeopardy. Erosion by rivers and streams is a real threat to Ohio earthworks, and it has claimed portions of many.
The Hopeton earthwork sprawls across a high terrace on the east side of the Scioto River, northeast from Mound City. Squier and Davis's (1848:Pl. XVII) somewhat stylized map of Hopeton appears in Figure 15. Figure 16 shows a LiDAR image of the earthwork and the surrounding landforms. Several things are worth noting about this image. First we see that the earthwork is situated on the highest terrace in this bend in the river; the next level up is the top of the bluff, outside the floodplain. Also visible in this image is a large gravel-mining operation and its impact on the landscape, including large, deep pits (water adversely impacts LiDAR surveys, which is why the surface of the water in the quarry pit looks unusual), and, most distressingly, an artificially-made hill located just to the northwest of the earthwork. Unfortunately, this eyesore blocks any view of the alignment of this earthwork to the summer solstice sunset through its diagonal axis (Romain 2005:3A-9--3A-10).
One of the things the LiDAR image shows is that the large Hopeton circle is not a geometrically perfect circle. This attribute of the Circle has been noted by others who have surveyed the site (e.g., Thomas 1894:474). In this case, the LiDAR data allow us to measure the Circle directly. As Figure 17 shows, from these data, it appears that the long diameter of the Circle as measured roughly east to west, from crest to crest, is about 1,054 feet in length. The east side (right side in Figure 16) of the large Circle runs up onto the slope along the base of the bluff. Perhaps the elongation of the east side of the circle was necessary to maintain this seemingly important diameter of 1,054 feet?
In connection with the so-called Hopeton Square, we note that the east side of the square enclosure is not a contiguous, straight-line embankment. Indeed, this observation is corroborated by other sources to include the ground survey map made by Thomas (1894:Pl. XXXV), recent magnetic surveys by the National Park Service, and most aerial photos of the site. Figure 18 shows a selection of these representations.
Intriguing too is that, in the 1938 aerial photo (Figure 18a), as well as in the Squier and Davis map (Figure 15), two circular features are visible along the eastern side of the Square. Both have been detected in the National Park Service’s geophysical surveys. Also noteworthy is that, in the 1938 aerial view, the parallel walls shown by Squier and Davis are visible. To date, these walls have eluded geophysical detection, except perhaps for a small segment revealed by a limited conductivity survey conducted in 2001 by Berle Clay in the area of the small circle that is attached to the north edge of the walls. None of these smaller features, the parallel walls or the circles east of the square, however, are visible in the LiDAR data. Two hundred years of plowing and other land modifications have had a profound impact on these earthworks.
High Bank Works
The High Bank earthwork is located at the confluence of the Scioto River and Paint Creek, roughly four miles south of Chillicothe. Figure 19 shows Squier and Davis’s (1848:Pl. XVI) map of the earthwork. The primary components of the site include a large circle connected to a large octagon. There are numerous secondary enclosures, some of which were apparently linked together by low, irregularly shaped walls southwest of the octagon. While the large circle and octagon are still visible today, the low connector embankments to the southwest have not been seen since Squier and Davis’s time.
Figures 20 and 21 show LiDAR images of the earthwork. Worth noting with reference to Figure 20 is that, several of the depressions shown on the Squier and Davis map (labeled “dug holes”) are visible in the LiDAR image. From the LiDAR data, and as shown by Figure 21, we are able to ascertain that the Large Circle is approximately 1,054 feet in diameter as measured along the longitudinal axis of the circle-octagon pair.
Figure 22 shows an interesting finding. In this case, the LiDAR data reveal that the area where the longitudinal axis of the earthwork intersects the Large Circle is elevated, which may be the result of years of embankment erosion, or it could be evidence of a conscious choice in construction technique. As the profile in Figure 22 shows, the embankment wall at point A is higher than the walls at B and C. The difference in elevation is about four feet. A similar situation is found at the Newark Octagon and Observatory Circle. In the Newark case, the Observatory Mound is incorporated into the Observatory Circle in such a way that its summit provides a view down the longitudinal axis of the Octagon-Circle complex.
Certainly this correspondence might be coincidental. However, in view of other shape and size similarities between the Newark and High Bank earthworks (see Squier and Davis 1848:71; Hively and Horn 1984:S92; Romain 2000:55-56) the possibility that this attribute was recognized by the Hopewell and perhaps intentionally selected-for in the layout of the High Bank earthwork is worth noting.
In the Squier and Davis map of High Bank (Figure 20), we note that a set of embankments lead from the earthwork toward a secondary group of enclosures along the bluff, overlooking the Scioto River floodplain. What may be traces of this feature are found in the Figure 20 LiDAR image. The location of the feature differs somewhat from the location represented by Squier and Davis. However, such discrepancies are not uncommon when comparing Squier and Davis’s maps to ground-truthed features at other sites. In any event, Figure 23 shows a LiDAR profile through the area of interest. In this figure, the blue dashed lines indicate the trajectory and crests of the proposed walls. The red dashed line is the profile across the feature. According to the profile data, the walls are about one-foot in height, while the area between the walls extends down to a lower elevation than would normally be expressed in the slope between the outside flanking areas. In cross-section, this configuration closely resembles the profiles of the Great Hopewell Road and the Marietta Sacra Via (see Romain and Burks 2008a).
The Junction Group is an unusual cluster of small enclosures and a few mounds near the southwest edge of Chillicothe, at the confluence (or junction) of Paint Creek and the North Fork of Paint Creek. As the crow flies, this is about five miles northwest of High Bank. Figure 24 shows Squier and Davis’s (1848:Pl. XXII) map of the site. Each of the enclosures consists of an embankment surrounding a ditch. Today this site is located in an agricultural field, and it is nearly impossible to see the earthworks from the ground. From the air (Figure 25), with the right kind of crop cover, moisture conditions, and light angle, some of the earthworks are still visible as crop marks. In this case, in early June of 2007, the young corn plants are growing differently on the earthworks.
Figure 26 shows a LiDAR image of the same area. Agricultural plowing has seriously degraded the earthwork complex. As a result, only a few of the original enclosures can be seen in the LiDAR image. Only three earthwork features are definitely visible: the two larger circles and the large rounded square.
Figure 27 shows the results of cross-section profiles across these three earthworks. Of interest is that all three profiles show the same general configuration: an outside wall with an inside ditch. Furthermore, the center of each of these three earthworks is sloped, as if a low mound once existed within each. In the case of one of the circles, Squier and Davis do in fact show a central mound (Figure 24).
Many more Ross County earthworks could have been included in this report. Our objective at this point, however, is simply to introduce the subject of LiDAR as it can be applied to the large prehistoric earthwork complexes of Ross County. In fact, LiDAR analyses of archaeological features have only just begun in Ohio.
In the 19th century, when many of these sites were first mapped, thousands of years of erosion had already obscured the earthwork outlines. Thus, many of these earthwork sites have hidden discoveries waiting to be revealed by modern investigators and LiDAR imagery represents one avenue for re-examining these monumental sites. Most importantly, because LiDAR assessments of large areas can be made quickly and with great accuracy, we now have a means to search for the remains of all of Ohio’s hundreds of earthwork sites and assess their conditions—a project long overdue.
The authors wish to acknowledge the assistance of The Ohio State University Newark Earthworks Center (NEC). The NEC is our partner in this continuing research project. In particular, we wish to thank Dr. Richard Shiels and Maarti Chatsmith for their continuing sponsorship and encouragement.
Special thanks are extended to Greg Rouse for providing access to the Ross County LiDAR LAS files and to Brian Redmond for his editorial comments.
Hively, Ray, and Robert Horn
1984 Hopewellian Geometry and Astronomy at High Bank. Archaeoastronomy Supplement to Vol. 15, Journal for the History of Astronomy 7:S85-S100.
Perderson, Jennifer, Jarrod Burks, and William S. Dancey
2001 Hopewell Mound Group: Data Collection in 2001. Current Research in Ohio Archaeology 2001, http://www.ohioarchaeology.org/joomla/index.php?option=com_content&task=view&id=46&Itemid=32 , accessed February 29, 2008.
Romain, William F.
2000 Mysteries of the Hopewell: Astronomers, Geometers, and Magicians of the Eastern Woodlands. Akron, Ohio: University of Akron Press.
2005 Appendix 3.1: Summary Report on the Orientations and Alignments of the Ohio Hopewell Geometric Enclosures. In Gathering Hopewell: Society, Ritual, and Ritual Interaction, edited by C. Carr and T. Case. New York: Kluwer Academic Publishers.
Romain, William F., and Jarrod Burks
2008a LiDAR Imaging of the Great Hopewell Road. Current Research in Ohio Archaeology 2008, http://www.ohioarchaeology.org/joomla/index.php?option=com_content&task=view&id=231&Itemid=32 , accessed February 4, 2008.
2008b LiDAR Assessment of the Newark Earthworks. Current Research in Ohio Archaeology 2008, http://www.ohioarchaeology.org/joomla/index.php?option=com_content&task=view&id=232&Itemid=32 , accessed February 11, 2006.
Shetrone, Henry C.
1926 Exploration of the Hopewell Group of Prehistoric Earthworks. Ohio Archaeological and Historical Quarterly 35:1-227.
Squier, Ephraim G., and Edwin H. Davis
1848 Ancient Monuments of the Mississippi Valley; Comprising the Results of Extensive Original Surveys and Explorations. Smithsonian Contributions to Knowledge Vol. 1. Washington, D.C.: Smithsonian Institution.
1894 Report on the Mound Explorations of the Bureau of Ethnology for the Years 1890-1891. In Twelfth Annual Report of the Bureau of American Ethnology. Washington, D.C.: Smithsonian Institution.
1 It is important to recognize that the process of creating a LiDAR image is part science and part art. Similar to a map, or drawing, what the viewer is presented with reflects what the image-maker wishes to show and is based on a series of subjective decisions. By changing perspective, lighting intensity and angle, height exaggeration, and color, very different results are achieved. Certain features can be exaggerated, others minimized. Indeed, features such as earthwork walls can be made very visible, just barely visible, or not visible at all.
On a more technical note, caution needs to be taken when using LiDAR images to assess potential astronomical alignments. LiDAR data typically utilize state plane coordinate systems. Grid north as indicated in such coordinate systems is not necessarily equivalent to astronomic north (i.e., true north). Sometimes the difference between grid north and true north can be one degree, or more – depending on the geographic area. For an observer on the ground, a one-degree difference is the equivalent of two sun diameters. Measured distances are not subject to the same provisos – but see endnote 2.
2 As with any measurable data, there is a range of error associated with earthwork measurements. Using software program features, we can zoom-in on the embankment walls. And using specific points identified in a point cloud rather than surface model image, we can determine the exact distance between points. Still, there is an element of uncertainty in determining exactly what point defines the center, or crest, of an embankment wall. Earthwork walls and corners are not perfectly symmetrical. Thus in the case of a LiDAR image, what one observer judges to be the center may differ from someone else’s opinion. In the case of the Baum and other LiDAR images discussed in this report, we estimate this uncertainty is about plus or minus four feet.
Only registered users can write comments.
Powered by AkoComment 2.0!
|Last Updated ( Monday, 03 March 2008 )|
|< Prev||Next >|