Remote Sensing Survey Preliminary Report Dillard Archaeological Site, Crow Canyon, CO

Figure 1 Dillard archaeological site and geophysical survey interpretations, June 2012.

Submitted by: Meg Watters, PhD Co-PI, Remote Sensing & Visualization Coordinator Oregon Public Broadcasting Table of Contents Overview ...... 5 Introduction ...... 7 Site Control Survey...... 7 Geophysical Methods, Principles, and Equipment ...... 9 Magentometry ...... 9 Conductivity / Magnetic Susceptibility ...... 11 Resistance ...... 13 Geophysical Data Interpretations...... 16 Magentometry ...... 18 Resistance ...... 28 Conductivity / Magnetic Susceptibility ...... 33 Conductivity ...... 33 Magnetic Susceptibility...... 38 Discussion of geophysical survey results ...... 43 Airborne LiDAR Principles and Results...... 51 Conclusions and Recommendations...... 52 Acknowledgements and Credits ...... 53 Time Team America ...... 53 References ...... 54 Table of Figures

Figure 1 Dillard archaeological site and geophysical survey interpretations, June 2012...... 1 Figure 2 New site features identified through geophysical surveys...... 6 Figure 3 Geophysical survey areas covered during Time Team America project...... 8 Figure 4 The magnetic anomaly produced by a kiln is aligned to the dip and direction of the Earth’s magnetic field (From Clark 1996)...... 9 Figure 5 Duncan McKinnon with the Bartington 601 dual array fluxgate gradiometer...... 11 Figure 6 Electromagnetic induction diagram...... 11 Figure 7 Bryan Haley with the EM38 conductivity meter...... 13 Figure 8 The flow of current from a single current source and resulting potential distribution...... 14 Figure 9 A general four electrode array...... 14 Figure 10 The Twin-electrode array commonly used in archaeology...... 15 Figure 11 Duncan McKinnon with the RM 15 resistivity meter and electrode array...... 16 Figure 12 Geophysical survey areas at Dillard site...... 17 Figure 13 Magnetic gradient survey results...... 19 Figure 14 Magnetic anomalies caused by iron stakes and/or nails are seen as mono-poles (A) or di-poles (B) with an orientation to magnetic north...... 20 Figure 15 Magnetic survey results with site surface features...... 22 Figure 16 Interpreted magnetic survey results (the red arc identifies part of the Great Kiva berm)...... 23 Figure 17 Interpreted magnetic survey results with site surface features...... 24 Figure 18 Auger positions for ground-truthing geophysical survey anomalies...... 26 Figure 19 New site features mapped through geophysical surveys and ground truthed. Pit house and pit house like features are purple polygons and blue points identify pits associated with a potential fence.27 Figure 20 Resistance survey, Dillard...... 29 Figure 21 Interpreted resistance survey...... 30 Figure 22 Interpreted resistance survey with overlain site features...... 31 Figure 23 Interpreted resistance survey with overlain ground-truthed archaeological features identified during geophysical surveys...... 32 Figure 24 Conductivity survey results...... 34 Figure 25 Interpreted conductivity survey results...... 35 Figure 26 Interpreted conductivity survey results with site features...... 36 Figure 27 Interpreted conductivity survey with ground-truthed archaeological features...... 37 Figure 28 Magnetic susceptibility survey results...... 39 Figure 29 Interpreted magnetic susceptibility survey results...... 40 Figure 30 Interpreted magnetic susceptibility survey results with site features...... 41 Figure 31 Interpreted magnetic susceptibility survey results with overlain ground-truthed archaeological features...... 42 Figure 32 Comparison of magnetic gradient (A & C) and resistance (B &D) big pit house (red) and possible ritual pit house (yellow) anomalies...... 43 Figure 33 Contrasting soils (red and dark brown) can be seen that define the edge of the big pit house feature. (Image courtesy of Crow Canyon Archaeological Center.) ...... 44 Figure 34 The floor of a section of the possible ritual pit house feature with a sipapu, ritual fire circle, and ash pit. (Image courtesy of Crow Canyon Archaeological Center.)...... 45 Figure 35 Interpretations of all the geophysical survey methods at Dillard...... 46 Figure 36 Core samples over geophysical anomalies help identify cultural (red points) and non-cultural (yellow points) site features...... 47 Figure 37 Ground-truthed archaeological features (purple are pit houses or pit house like structures and navy blue are individual pits) overlain on site interpretations and surface features...... 48 Figure 38 Mapped site features overlain on interpretations of geophysical surveys...... 49 Figure 39 Pit house anomaly comparison. The pit house features as ground-truthed are positioned in the center of each sample above...... 51 Figure 40 LiDAR DTM and the broader landscape with BM III site distribution in reference to the Dillard site (red)...... 52 Overview Time Team America joined the Crow Canyon Archaeological Center (CCAC) in June 2013 to investigate the Dillard site, a community center during the Basketmaker III (BM III) period, A.D. 500-725. The Time Team America challenge at the Dillard site was to (1) try to determine the site population, (2) to better understand why a Great Kiva was built here, (3) to gain insight into the organization of the site, (5) its context to the broader landscape, and (6) what this meant for the development of community.

BM III settlements cannot be identified or analyzed from the ground surface, they simply are not visible. Over the past three years CCAC investigations have identified eight pit structures to the south of the great kiva through systematic auguring, excavation, and a small amount of resistivity. In the three days of geophysical surveys conducted by Time Team America that included magnetic gradient, conductivity, and resistivity surveys (Charles 2012), an additional eight to nine pit structures and a variety of other pit-like anomalies were identified in the area to the north of the Great Kiva (Figure 2). Ground-truthing (excavation and auguring) geophysical anomalies helped provide material that was able to provide dates the site, from A. D. 610 to 670. One pit structure identified through geophysical surveys was partially investigated and revealed a suite of ritual features, often found in BM III structures, in this instance the orientation, formality of construction and closing the of the features suggests a relationship to the great kiva. Auguring of selected single pit anomalies suggests a fence line feature that runs between the great kiva and cluster of pit structures to the north.

Ground penetrating radar was tested on different parts of the site and ruled out as an effective technique due to the soil properties and failure of GPR to record any useful information. (Part of this testing was conducted over the balk of the excavation trench at the Great Kiva where the stone foundation of the kiva should easily have been recorded if ground conditions were amenable.) On May 9, 2012 an airborne LiDAR survey was conducted by Paul Kinder and Adam Riley from the Natural Resource Analysis Center, West Virginia University. As part of the flight, Meg Watters (Time Team America, Oregon Public Broadcasting, OPB) and Shanna Diedrerichs (site director, CCAC) were filmed in the air discussing the Dillard site, LiDAR technology, and what might be captured in the LiDAR data. The goal of the airborne LiDAR survey was to attempt to identify possible site features and to contribute to a broader landscape perspective for interpreting the Dillard site. Figure 2 New site features identified through geophysical surveys. Introduction Geophysical survey methods of the sub-surface and 3D laser scanning of existing environments provide cost-effective means for capturing archaeological information for site recording, investigation, and management. Using non-invasive sub-surface and surface mapping methods can document the basic structure and layout of site. In instances where historic properties are active sites with maintenance and potential development impact demands, these methods can guide placement of expensive excavations and contribute to site impact strategies when dealing with upgrade of site infrastructure (such as utilities, and landscape management); thus providing large cost savings while reducing destructive impact upon important archaeological remains.

Geophysical survey methods can provide primary information on site settlement patterns. The continued application and development of broad area coverage for archaeological assessment has begun to introduce an alternative perspective into regional, or landscape archaeology (David and Payne 1997; Kvamme 2003). Because geophysical surveys are able to cover large areas in comparison to the limited extent of archaeological excavations, the information they provide introduces a new component to the concept of the archaeological landscape. Broad area geophysical surveys provide information on the structure and organization of a site enabling the study of spatial patterns and relationships relevant to research questions. In addition to the large-scale perspective of the site, geophysical survey results also provide a high-resolution focus on individual site features.

Geophysical surveys measure different subsurface properties at regular intervals across broad areas. Contrasting properties in a relatively homogeneous soil can identify buried objects or features such as foundations, compacted earthen surfaces, pits, stone walls, middens, hearths and any number of archaeological features. The different physical properties of the features, measured either in contrast to their surrounding matrix, or as recorded at the surface are referred to as ‘anomalies’ until they are able to be ground-truthed through excavation or other methods such as soil coring.

Different geophysical methods are sensitive to specific properties, such as magnetic fields, or the flow of an electrical current in the earth. Employing a combination of methods over a survey area can help provide information as to the nature, or material, of an anomaly, thus providing insight for site interpretation. Mapping the distribution of anomalies over a large area can help in the recognition of anomalies generated through cultural activities revealing the spatial distribution and association with site features (Kvamme 2003).

Geophysical surveys can provide important information for help in site planning and preservation. These non-invasive methods can help establish priorities and identify areas for further invasive investigations, or for preservation and management. They are a fast and cost-effective method for gaining insight to what is buried beneath the ground. Geophysical survey results can be spatially integrated with other data relevant to archaeological investigations to provide a comprehensive record of the site environment, both below, and above ground.

Site Control Survey The Time Team America geophysical survey grids were established by CCAC surveyors and tied into their survey control. Due to the ground cover and inherent data collection rate for different geophysical survey methods, coverage was different for each method. Magnetic gradient survey covered the entire research area targeted (north of the Great Kiva to site fence boundary), while conductivity / magnetic susceptibility and resistance surveys covered smaller areas (due to method and rate of data collection). The resistance survey was conducted by a group from the Center of Southwest Studies at Fort Lewis College, directed by Mona Charles. The survey results from their work during the 3 day period of the Time Team America project are presented here. Additional site coverage of this and other related sites can be read in Charles, M.C., 2012. Electrical Resistance Survey of Three Sites in the Indian Camp Ranch, Montezuma County, Colorado; Report prepared for Crow Canyon Archaeological Center Cortez, Colorado.

Figure 3 Geophysical survey areas covered during Time Team America project. Geophysical Methods, Principles, and Equipment

Magentometry Magnetometers are passive instruments that measure the magnetic field strength a specific location on the surface of the Earth. The Earth’s magnetic field varies depending on location relative to the earth’s equator and can be visualized as a large bar magnet that is tilted 11 degrees from the axis of rotation (Heimmer and Devore 1995). Over a small area and in homogeneous soils, the magnetic field is expected to be uniform (Weymouth 1986). A subsurface target can be detected with magnetic survey as a deviation from this background field reading. The resultant anomaly often has a dipolar form aligned with the dip and direction of the Earth’s field (Figure 4). The most common unit of measure is the nanoTesla (nT).

The magnetic signal of a target is composed of two parameters: induced and remnant magnetism (Reynolds 1997). A magnetometer measures the remnant magnetism of a target, which is permanent and may be caused by the presence of highly magnetic rock compounds or thermal alterations to soils which have high iron content (Heimmer and Devore 1995). Magnetization caused by thermal alteration is called thermoremanence and it occurs at maximum expression at temperatures above about 600 degrees Celsius, but there is some effect at any elevated temperature (Aitken 1964).

Figure 4 The magnetic anomaly produced by a kiln is aligned to the dip and direction of the Earth’s magnetic field (From Clark 1996).

Induced magnetism is only visible in the presence of magnetizing field. However, the Earth serves as a constant magnetizing agent and, therefore, it can be sensed by a magnetometer. The induced magnetism is generally referred to as magnetic susceptibility. Magnetic susceptibility is greater in the topsoil and soils that are organically rich, but often produces relatively subtle anomalies (Clark 1996). Therefore, excavations that rearrange the topsoil are sometimes evident in magnetic surveys, but these are rather weak in strength. The Geonics EM38B conductivity meter can better measure the induced magnetism of the ground.

Magnetic anomalies produced by archaeological targets are often much weaker than signals produced by other sources, usually between 1 nT and 100 nT (Aitken 1961). However, anomalies produced by historic period targets are usually much greater than this range. Archaeological objects that may produce magnetic anomalies include fireplaces, furnaces, burnt clay floors, hearths, kilns, daub, bricks, and walls composed of magnetically anomalous rocks such as basalt (Aitken 1964; Hasek 1999).

Another type of target visible magnetically is ferrous, or iron containing materials (Aitken 1964). Archaeological targets such as historic nails can sometimes be mapped using magnetometers. However, more recent ferrous objects, such as power lines, cars, buried pipes, and surface trash, can easily obscure archaeological targets (Heimmer and De Vore 1995). Some advantages to the use of fluxgate instruments are their relative insensitivity to steep magnetic gradients and their speed of acquisition is better (Reynolds 1997). Fluxgate instruments have become the workhorse for archaeological geophysical survey in Britain and the United States (Clark 1996).

The magnetic gradiometer was developed in the 1990s and uses two sensor heads. The primary advantage of a gradiometer system is that no correction for diurnal drift is necessary (Reynolds 1997, Bevan 1998). In addition, they are much less affected by nearby objects with steep magnetic gradients, such as large masses iron (Bevan 1998). Also, gradiometers tend to emphasize shallow anomalies, a benefit for archaeological survey. One disadvantage is that the accuracy is dependent on a consistent orientation of the sensors (Bevan 1998, Hasek 1999).

Interpretation of magnetic imagery begins by identifying anomalies, which may have strong high and low amplitude values (Bevan 1998). Next, metal objects can be identified from the shape and amplitude. Anomalies with strong, narrowly spaced dipoles or strong monopoles are usually produced by ferrous metal objects. If targets are relatively large and the amplitude is not extreme, the shape may be approximated in the magnetic imagery (Bevan 1998).

Little information about the depth of a target is obtained with magnetic survey. In some cases, the half- width rule can be used to estimate target depth. The half- width rule depends on the amplitude drop off for readings over a target and assumes a simple and regular target shape (Bevan 1998). However, except for buried iron targets, this technique is often not useful for archaeological targets.

The Bartington 601 fluxgate gradiometer was used for the magnetic survey (Figure 5). Magnetometry survey parameters were: 0.125 m sample rate 0.5 meter transect spacing Zig-zag data collection method (survey grid SW corner to grid NE corner)

The magnetic survey data were processed using Geoplot 3.0. Processing techniques included de-spiking, grid/transect mean zeroing, 3 x 3 low pass and 10 x 10 high pass filtering. Once processed, data were interpolated along the x axis. Figure 5 Duncan McKinnon with the Bartington 601 dual array fluxgate gradiometer. Conductivity / Magnetic Susceptibility Electromagnetic (EM) induction instrumentation uses a near surface transmitter coil to emit radio frequency electromagnetic waves into the subsurface. Objects in the subsurface respond by generating eddy currents, producing a secondary electromagnetic field (Figure 6). This secondary electromagnetic field is proportional to conductivity and detected by a receiver coil on the instrument and recorded by an attached data-logger (Bevan 1983; Clay 2006).

Figure 6 Electromagnetic induction diagram. The Geonics Limited EM38B was used in the survey (Figure 7) and allows for simultaneous collection of both quadrature-phase (electromagnetic conductivity) and in-phase (magnetic susceptibility) components. Electromagnetic conductivity measures the “ability of the soil to conduct an electric current” (Clay 2006) and is recorded in siemens (mS/m). Theoretically, electromagnetic conductivity is the inverse of resistivity although methods for recording each are completely different (voltage, sample spacing, soil, volume, sensitivity to metals) and results may not match entirely. The transmission of the quadrature-phase component of the induced electromagnetic field signal is related to the mineral and chemical composition of the soil. Soils high in clay and/or saline composition will produce higher conductivity measurements; whereas soils composed of sand and/or silt will produce a lower conductivity measurement. Levels of soil moisture also have a dramatic impact on conductivity measurements where increased moisture will cause higher conductivity readings (Clay 2006).

Magnetic susceptibility measures “a material’s ability to be magnetized” (Dalan 2006). It is different from magnetic gradiometry in that susceptibility is an active measurement recorded in the presence of an induced magnetic field. The transmission of the in-phase component of the induced electromagnetic field is based on the presence of a magnetic topsoil matrix being greater in magnetism than proximate soil matrix or materials. The increase in magnetism in topsoil is the result of pedogenesis enhancement from hematite, magnetite and maghematite minerals. Additionally, changes to the magnetic composition of the soil can be caused by human activity, such as fire or the movement of magnetically rich topsoil (Dalan 2006).

Both quadrature phase and in phase readings were simultaneously collected for each station, relating to conductivity and magnetic susceptibility properties respectively. This specification results in a maximum depth sensitivity of about 1 m for the conductivity. For the magnetic susceptibility, the penetration is significantly shallower.

Conductivity survey data sampling: 2 samples per meter 0.5 meter transect spacing Parallel data collection method (all transects travel grid south to north)

The EM data were processed using Geoplot 3.0. Null values were added in a text editor so that grid lengths and widths were in multiples of 10 meters and these were used to create a single composite data set.

Data processing methods include a despike operation and a 3X3 low pass, as well as the addition of a 10 X 10 high pass filter to a second version. The magnetic susceptibility data was processed in a similar fashion, without the creation of the second high pass filtered version. Figure 7 Bryan Haley with the EM38 conductivity meter.

Resistance Resistance survey was conducted on site by a group from Fort Lewis College led by Mona Charles. Results of the survey conducted during the Time Team America filming are integrated into this report. After filming, additional areas of the site were surveyed with resistivity, results are presented in Charles, M. 2012. Electrical Resistance Survey of Three Sites in the Indian Camp Ranch, Montezuma County, Colorado. Report prepared for: Crow Canyon Archaeological Center, Cortez, CO.

Resistance survey is designed to measure the electrical resistance of the earth in order to provide information on the subsurface structure. The electrical properties of the earth are recorded as a function of depth and / or horizontal distance. An electrical current is introduced into the earth through electrodes and the resulting potential distribution is sampled at the ground surface. The measured apparent resistivity provides information on the magnitude and distribution of the electrical resistivities in the volume of the sampled subsurface (Griffiths and King 1981).

An electric current is caused by the flow of charged particles and is measured in amperes (amps). Amperage expresses the amount of charge that passes any point in a circuit in one second. A measurement of the ground resistivity is made by passing an electrical current into the ground through an electrode acting as the current source (Figure 8). A second electrode, or current sink, enables the electrical current to exit from the ground completing the circuit. The current flows into the earth in all directions from the source electrode. Figure 8 The flow of current from a single current source and resulting potential distribution.

The most common electrode configurations are linear arrays that contain two current electrodes (A and B) that are the current source and sink of equal strength, and two potential electrodes (M and N) that measure the difference in potential between two points (Figure 9).

Figure 9 A general four electrode array.

If the ground is inhomogeneous and a fixed electrode array is moved or the electrode spacing is varied during survey, the calculated resistivity will vary for each measurement. The resistivity of the earth can vary greatly depending on the composition and structure of the material and ground water saturation. Not only does resistivity vary with rock formations, it also varies from deposit to deposit and on a macro scale within individual deposits depending upon their structure. Resistivity values can vary greatly due to the unconsolidated nature of near-surface materials. The principles provided for basic rock formations can be followed when considering the structure of the near surface and resistivity mapping for archaeological applications (Griffiths and King 1981).

The nature of the archaeological features, the mineral content and compaction of soils in which they are buried, and the saturation levels of the subsurface all affect earth resistivity. The saturation of the subsurface is dependent on rainfall, soil composition and compaction and subsequent percolation rates, evaporation rates, and water take-up through the roots of vegetation. Weather and geological conditions impact on the effectiveness of resistance surveys in archaeological applications and dictate careful consideration of resulting data (Clark 1996). A number of electrode arrays are used in resistance surveys. The array, or configuration, refers to the arrangement of electrodes. Linear arrays, which are used more commonly, consist of two current electrodes (A and B) and two potential electrodes (M and N). The twin-electrode array is the most popular for archaeological surveys. Due to the relative speed of data collection, the benefits of the resulting survey include a high lateral resolution and depth of investigation relative to the spacing of the mobile electrodes (Apparao et al. 1969; Apparao and Roy 1971). The basic twin electrode array used in archaeological applications can be seen in Figure 8 where single current (A) and potential (M) electrodes are set with a fixed distance (a) with the second pair of electrodes (B and N) are placed at a distance 30 times the spacing (a) of the primary electrodes (A and M) and fixed separation distance (a) the same as the mobile probe spacing (Figure 10).

Figure 10 The Twin-electrode array commonly used in archaeology.

The depth of investigation can be defined as the depth at which a thin horizontal layer makes the maximum contribution to the total measured signal at the surface (Barker 1989; Evjen 1938; Roy and Apparao 1971; Roy 1972). The separation distance and positions of the current and potential electrodes fundamentally contribute to calculating the most accurate depth estimation. The depth of investigation of electrode arrays should be the depth with which a measurement of apparent resistivity is best associated. Although there is no single depth of investigation, a single value is more useful to have as a reference. The most practically useful value is the median depth (Edwards 1966; Barker 1989). The median depth is defined as the depth from below which and from above which 50% of the signal originates.

The RM15 resistivity meter was used for survey at the Dillard site (Figure 11). Figure 11 Duncan McKinnon with the RM 15 resistivity meter and electrode array.

Geophysical Data Interpretations The geophysical survey area was focused to cover the Dillard site from the Great Kiva to the northern perimeter of the area bounded by a fence. The fence was removed to enable effective survey with instruments sensitive to iron. Figure 12 shows the area surveyed by Time Team America. Each data image in this report has a key that defines map layers. EM survey area refers to Electromagnetic Induction survey, or conductivity and magnetic susceptibility. Other abbreviations in the key will include: mag – magnetic gradient survey; res – resistance survey; cond – conductivity survey; and mag sus – magnetic susceptibility survey.

Site features included in some of the data maps were provided by Crow Canyon Archaeological Center. Figure 12 Geophysical survey areas at Dillard site. All of the geophysical survey results were imported to a project GIS using ArcMap 10.1. Data are rectified into the GIS project and polygon files are created to identify and map interpreted anomalies. Data results are presented below with and without interpretations. This is done so that the client may look at the data and consider what they may see based upon their viewpoint and expertise. Magentometry The magnetic gradient survey mapped 8 to 9 new pit house like structures, a number of individual pits and numerous anomalies that may represent additional buried archaeological features. Figure 13 shows the results of the magnetic gradient survey, magnetic gradient values range from – 3.56 (white) to 3.39 (black) nT. Figure 13 Magnetic gradient survey results. When interpreting data it is vital to consider surface features and the signatures that they might introduce to the final data maps. A number of things can contribute to the data maps including surface brush or trees, these may knock sensors out of alignment and introduce data spikes. While attempt to filter out data spikes, this site had significant obstacles to survey around and through, some area surface features having more impact on the final data than others. Piles of surface rock, midden, or backfill can also appear in the final data plans depending on their geophysical values. Having the surface feature information for the Dillard site complements the data interpretation process but all interpretations must be assessed visually in the field to see if there may be additional surface features that could be the cause of potential anomalies.

Figure 15 (and subsequent surface feature overlays on other survey methods) shows the site features overlain on the geophysical map of the magnetic gradient survey. The windrows stand out clearly in the magnetic data as strong, linear magnetic anomalies. Iron stakes or nails also stand out clearly as very highly contrasting black and white anomalies, either as monopoles (Figure 14, A), or as dipoles oriented to magnetic north (Figure 14, B).

A B

Figure 14 Magnetic anomalies caused by iron stakes and/or nails are seen as mono-poles (A) or di-poles (B) with an orientation to magnetic north.

Figure 16 shows the final interpretation for the magnetic gradient survey (Figure 17 shows interpretations with overlain surface features). Interpretations were divided into three categories: individual anomalies, mag_AOI (general areas of interest), and pit house – AOI (specific areas surrounding the pit houses). The individual anomalies are interpreted over many positive (black) and negative (white) anomalies, as well as areas of varying magnetic field strength. Because we ground- truthed a sample of individual magnetic anomalies that turned out to be single pits, interpretations in this report include many, if not most, of the point anomalies in the survey data. I felt that I could not skip over any, as pit fill may vary, thus the magnetic field strength may vary from pit to pit. There may be a few instances where some changes in the data are not annotated. The line had to be drawn somewhere with the interpretations, thus what is included in these maps are to the best of my ability taking all information into consideration (surface features, ground-truthed features, possible rodent burrows, etc…).

Areas of interest are identified through what appears as a ‘clustering’ of anomalies, these areas should be closely considered on the ground surface and may represent activity areas. Areas of interest associated with the pit houses suggest possible activity that may be related to the pit house and activities that may be associated with them. There are many nuances within this data in particular and an attempt to identify different categories for further consideration was made. Continued ground- truthing will greatly assist in a more intuitive interpretation of these data and hopefully, this can be a case study for investigation of similar sites in the region.

(Though not interpreted in the GIS, the Great Kiva berm is clearly visible in the magnetic gradient data, Figure 16, red line.) Figure 15 Magnetic survey results with site surface features. Figure 16 Interpreted magnetic survey results (the red arc identifies part of the Great Kiva berm). Figure 17 Interpreted magnetic survey results with site surface features. As part of Time Team America, geophysical data are used in part to select areas to excavate in the 3 day format. Because of the excellent results of the magnetic survey, a series of excavation units, coring transects, and individual core locations were identified to help characterize what the magnetic survey (and additional surveys as their results were investigated) data was mapping. Figure 18 shows most of the cores / augur locations that were conducted. Yellow points identify non-cultural samples and red points identify cultural samples that were retrieved through coring. Figure 19 shows the final interpretation of the archaeological features mapped through the geophysical surveys and ground- truthed through coring and excavation.

The identification of pit house (and pit house like) structures through the geophysical surveys (Figure 19, dark purple) begins to reveal the distribution of structures and helps estimate the population of the site. As part of the ground-truthing, anomalies that appeared as a double ‘ring’ of magnetic points encircling several pit structures to the north of the Great Kiva were sampled; preliminary results identify four of these point anomalies (navy blue) as pits and thus, are interpreted by site archaeologists as an alignment of postholes that would have been associated with a fence. This reveals not only information on the organization of space but also begins to provide insight to social and community organization. Following on the research by Kvamme at Huff Village (2003), further investigation of individual magnetic anomalies may provide insight to much more information related to site population, the season that it was abandoned and more. Figure 18 Auger positions for ground-truthing geophysical survey anomalies. Figure 19 New site features mapped through geophysical surveys and ground truthed. Pit house and pit house like features are purple polygons and blue points identify pits associated with a potential fence. Resistance The resistance survey covered (Figure 20) a limited part of the entire survey area but clearly mapped (Figure 21) pit house and pit house like structures, as well as provides a number of anomalies for further investigation. Figure 22 presents the surface features overlain on resistance data and interpretations. Resistance data values range from 8.42 (white) - 29.63 ohms (black).

Resistance survey mapped 4 of the 6 pit house / pit house like features (in the survey area it covered) identified through magnetic survey and coring (Figure 23, purple polygons). Resistance survey did not identify any of the single pit anomalies that were mapped through magnetic gradient survey, ground- truthed and interpreted as pit features representing a fence (Figure 23, navy blue points). Figure 20 Resistance survey, Dillard. Figure 21 Interpreted resistance survey. Figure 22 Interpreted resistance survey with overlain site features. Figure 23 Interpreted resistance survey with overlain ground-truthed archaeological features identified during geophysical surveys. Conductivity / Magnetic Susceptibility Conductivity (Figure 25) and magnetic susceptibility (Figure 28) surveys map a number of anomalies that may be related to archaeological features. The nature of these surveys may reveal more subtle features that relate to areas associated with archaeological structures (as revealed through resistivity and magnetic surveys). It is important to remember the issues of ground surface obstacles as discussed earlier in the report, the EM38B conductivity meter is sensitive to instrument orientation and height above the ground surface. If knocked off of axis or lifted to avoid surface obstacles, artifacts may be introduced into the final data, visual inspection of conductivity and magnetic susceptibility plan maps and the survey surface are recommended when reviewing data.

Conductivity Conductivity survey maps 4 of the 7 pit house (pit house like) features as ground-truthed in the field with Time Team America. It is interesting to note that the conductivity survey (individual anomalies and Area of Interest) are located adjacent to the pit house (pit house like) features. These areas of contrasting conductivity values may represent deposition and / or compaction as a result of activities that might (because of vicinity) be related to the pit house (pit house like) structures. Conductivity values range from 10.2 (white) to 12.33 (black) mS/m, a very fine threshold.

Figure 26 shows conductivity data interpretations. Note the purple polygons that are labeled cond_BPH_InteriorFeature. In this instance, I was interpreting the conductivity data in isolation, attempting not to be influenced by the anomalies interpreted (and ground-truthed) in the other geophysical survey methods. However, knowing that magnetic gradient and resistance surveys both mapped a ‘big’ pit house (BPH) feature, I wanted to see how this appeared in the conductivity data. I saw a higher conductivity anomaly in the location of the big pit house and within this anomaly, smaller individual areas of higher contrasting conductivity values. The thought was to try to interpret features within the actual pit house (as Kvamme did at Whistling Elk, interpreting the central fire hearth and other pit house features in the magnetic gradient data, 2006). But, when compared to the magnetic and resistance survey data, it appears that the conductivity polygon that ‘defined’ the area of the big pit house structure overlapped with the footprint, but also continued to the northwest of the pit house. It is in this area adjacent to the pit house that the individual high-conductivity point anomalies are identified (cond_BPH_InteriorFeature). It will be interesting to investigate this area more to see what, if any, features might be discovered and how they might relate to the big pit house.

Figure 27 shows that conductivity data did not map the individual pits that were mapped through magnetic gradient survey and are interpreted to indicate a fence line. Figure 24 Conductivity survey results. Figure 25 Interpreted conductivity survey results. Figure 26 Interpreted conductivity survey results with site features. Figure 27 Interpreted conductivity survey with ground-truthed archaeological features. Magnetic Susceptibility Magnetic susceptibility survey identified anomalies at 5 of the 7 pit house (pit house like) features in its survey area and a number of additional areas of interest for further investigation. Some surface features such as the windrow and foot path are clearly visible in this data. The pits identified through the magnetic gradient survey that define a possible fence line are not mapped through magnetic susceptibility survey. The data values for the magnetic susceptibility survey range from 0.13 (white) to 0.50 SI (black).

Figure 28 shows the results of the magnetic susceptibility survey; interpretations are shown in Figure 29. Figure 30 shows the site surface features overlain on the interpreted magnetic susceptibility results and Figure 31 overlays the ground-truthed archaeological features identified through ground truthing of the geophysical anomalies. Figure 28 Magnetic susceptibility survey results. Figure 29 Interpreted magnetic susceptibility survey results. Figure 30 Interpreted magnetic susceptibility survey results with site features. Figure 31 Interpreted magnetic susceptibility survey results with overlain ground-truthed archaeological features. Discussion of geophysical survey results Two features that appeared as strong anomalies in the resistance and magnetic gradient data were selected for further investigation on the first day of Time Team America’s project; the big pit house and a second pit house or pit house like structure. Figure 32 shows the survey results (A magnetometry, B resistance) with the outlined pit house anomaly (big pit house in red, possible ritual pit house in yellow) in each method (C magnetometry, D resistance.)

Figure 32 Comparison of magnetic gradient (A & C) and resistance (B &D) big pit house (red) and possible ritual pit house (yellow) anomalies. Ground-truthing of both of these pit house anomalies feeds back to the interpretation of the geophysical data. The site surface area of the big pit house was cleared down to the top of the pit house feature, defining the edges of the pit house structure (Figure 33). A sample trench was excavated through the second pit house feature to the feature floor; initial interpretations suggest this is a typical pit house that has an alignment of features that include a ritual sipapu, fire hearth, and ash pit, (Figure 34). It is interesting to note the complementary feature anomalies of the magnetic gradient and resistance surveys. Magnetic gradient appears to have relatively the same strength over the central part of both pit house features; resistance survey appears to give a more accurate outline of the extent of the big pit house feature.

Figure 33 Contrasting soils (red and dark brown) can be seen that define the edge of the big pit house feature. (Image courtesy of Crow Canyon Archaeological Center.) Figure 34 The floor of a section of the second pit house feature with a ritual sipapu, fire hearth, and ash pit. (Image courtesy of Crow Canyon Archaeological Center.)

Viewing all of the geophysical survey interpretations together presents a colorful picture (Figure 35). Obviously, there is a lot going on in this area. To help better understand the geophysical surveys, and thus more intuitively interpret the data and archaeological nature of the site, initial ground-truthing through coring (Figure 36) helped identify areas that contained cultural materials, enabling interpretation of possibly 9 pit house or pit house like features and a series of aligned pits that are interpreted as a fence (Figure 37). Combining the geophysical interpretations with the ground-truthing and overlaying the site surface features (Figure 38) enables a new approach to considering the buried archaeological resource at the Dillard site. Figure 35 Interpretations of all the geophysical survey methods at Dillard. Figure 36 Core samples over geophysical anomalies help identify cultural (red points) and non-cultural (yellow points) site features. Figure 37 Ground-truthed archaeological features (purple are pit houses or pit house like structures and navy blue are individual pits) overlain on site interpretations and surface features. Figure 38 Mapped site features overlain on interpretations of geophysical surveys. One of the primary pieces of information we are gaining from these surveys and ground-truthing investigations is an understanding that pit house features, though not visible on the ground surface, area visible through geophysical surveys. However, it is important to note that no two pit house features appear the same in any of the geophysical survey methods. To best comprehend this statement, Figure 39 presents each of the pit houses as mapped, or not, through geophysical surveys. Figure 39 Pit house anomaly comparison. The pit house features as ground-truthed are positioned in the center of each sample above.

The point of comparing the geophysical signatures over each of the pit houses shows that thus far, there is no specific ‘guideline’ to data interpretation nor a specific ‘signature’ in any of the geophysical survey methods for pit house features. The general conclusions from this comparison of geophysical survey methods are:

1. Magnetic gradient was both the fastest and most effective tool for mapping buried archaeological features at the Dillard site. 2. Resistance was a much slower method, but effective in mapping some of the pit house features. 3. Conductivity and magnetic susceptibility were slightly faster survey methods than the resistance, but were not consistent at discerning the actual location of pit house features with a high level of confidence.

It is interesting to note, is that these two latter methods may provide valuable information to activities related to the pit house (pit house like) features.

Airborne LiDAR Principles and Results Airborne LiDAR, or light detection and ranging, measures the height of the ground surface and any features (i.e. trees, buildings) that may be on it and provides high definition and accurate models of the landscape to a resolution of 1 m to 0.5 m in archaeological applications. LiDAR uses a pulsed laser beam that scans from side to side as a plane flies at a low altitude over the survey area. 20,000 to 100,000 points per second build the ground model. In post-processing the first returns can be removed from the data providing a ‘bare earth’ model (or Digital Terrain Model, DTM) that accurately represents the ground surface.

The airborne LiDAR data were acquired by the NRAC, West Virginia University. NRAC operates an OPTECH ALTM-3100C airborne laser (small-footprint) mapping system. The system integrates a laser altimeter, a high-end Applanix Pos/AV Inertial Measurement Unit (IMU), also called an Inertial Navigation System (INS), and a dual frequency NovAtel GPS receiver. This integrated system is capable of 100 kHz operation at an operating height of 1,100 meters (3,609 feet). LiDAR technology offers fast, real-time collection of three-dimensional points that are employed in the creation of Digital Elevation Models (DEMs), Digital Terrain Models (DTM), landscape feature extraction, forest stand structure analysis, as well as many other research applications.

Data were collected in multiple, low altitude acquisition passes over the core area of the Dillard site (Figure 40) to yield ground LiDAR point densities of 15-20 per square meter (vertical accuracy of 15 cm or better). Integrated data have a vertical error of 15cm or less at the 95% confidence level for areas of open terrain and moderate slopes of 10 degrees or less (based off manufacturer’s specifications). Data are recorded in the applicable Universal Transverse Mercator (UTM) zone, NAD83 datum (CORS96) while heights are orthometric, referenced to the North American Vertical Datum of 1988 (NAVD88) using GEOID09.

The resulting ‘bare earth’ model from the LiDAR data provides an excellent model of the landscape, Viewing the site and interpreted features draped on a LiDAR digital terrain model (DTM) shows its location within the broader natural and cultural landscape. The site’s orientation to local landmarks such as the San Juan Mountains to the east, the Mesa Verde questa to the south, and lone Ute Mountain to the west confirm the site’s expansive view shed of the prehistoric Mesa Verde Region. Despite this emphasis on view shed, the LiDAR DEM demonstrates that the site sits on one of many low lying ridges, making it easily accessible to the 107 known BMIII habitation sites in the surrounding settlement providing insight on its role in the larger community.

Figure 40 LiDAR DTM and the broader landscape with BM III site distribution in reference to the Dillard site (red).

Conclusions and Recommendations

The geophysical methods used for the Dillard site survey were very effective at mapping archaeological features as well as providing potential contextual information related to individual pit house features. We are able to make a preliminary statement about the distribution of structures and an estimate of the population of the site. The ground-truthed features and remaining geophysical interpretations provide important information to consider not only the use and organization of space at the Dillard site but also provide insight to social and community organization. This work proves that magnetic gradient survey is the most effective tool for mapping Basket Maker III sites in the Crow Canyon region (i.e. similar archaeological features, ground matrix, and environment). It is recommended that for future site exploration magnetic gradient would be the most effective tool. Depending on further site investigations at the Dillard site, conductivity / magnetic susceptibility (electromagnetic induction with the EM36B) survey may provide additional information related to the space around pit house or pit house like features. While a bit on the slow side for data collection, resistivity is an excellent survey method if conducted during a season when the earth is not entirely dry, as it depends on some ground moisture for successful survey.

Additional regional landscape investigation through traditional GIS and LiDAR 3D model analyses may provide interesting insight to the location of the Great Kiva at the Dillard site.

Acknowledgements and Credits Sincere thanks are given to Shanna Diederichs for her continued involvement with the geophysical data and site interpretation long after the Time Team America crew departed the field. Thanks also go to Shirley Powell, Steve Copeland, Scott Ortman, Grant Coffey, Caitlin Sommer, Mark Varien, Dylan Schwindt and the rest of the Crow Canyon Archaeology Center archaeologists and staff. Our work could not have been completed without the support and assistance from Jane Dillard, for this we are sincerely thankful.

Time Team America This work was undertaken as part of the filming of Season 2 of the PBS prime-time program Time Team America. The program is co-produced by Oregon Public Broadcasting and Videotext LLC and funded entirely by a National Science Foundation Informal Science Education grant. Meg Watters is a co-PI1 on the grant and the Remote Sensing and Visualization Coordinator for the television program. Members of the Time Team America geophysical survey team include Bryan Haley, Tulane University, and Duncan McKinnon, University of Arkansas2. Adam Riley and Paul Kinder, West Virginia University, Natural Resource Analysis Center, performed the LiDAR survey and delivered processed data to Watters. The material in this report is based upon the work supported by the National Science Foundation under Grant number 1114113. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.

1 Noel Broadbent, Smithsonian Institute is the other co-PI and Dave Davis from OPB is the project PI. 2 Thanks to Duncan and Bryan for all of their hard work in and out of the field with data collection, processing, interpretation and contributing to this report. References Aitken, M. J. 1961. Physics and Archaeology. Interscience Publishers Inc., New York.

Apparao, A., A. Roy & K. Mallik. 1969. Resistivity model experiments. Geoexploration 7:45-54.

Apparao, A. & A. ROY. 1971. Resistivity model experiments. Geoexploration 9:195-205.

Barker, R. 1989. Depth of investigation of a generalized collinear 4-electrode array. Geophysics 54:1031-1037.

Bevan, B. 1983. Electromagnetics for Mapping Buried Earth Features. Journal of Field Archaeology 10:47-54.

Bevan, B. 1998. Geophysical Exploration for Archaeology: An Introduction to Geophysical Exploration. Midwest Archaeological Center, Special Report No. 1. National Park Service, Lincoln, Nebraska.

Charles, M. 2012. Electrical Resistance Survey of Three Sites in the Indian Camp Ranch, Montezuma County, Colorado. Report prepared for: Crow Canyon Archaeological Center, Cortez, CO.

Clark, A. 1996. Seeing Beneath the Soil, Revised Edition. London: B. T. Batsford Ltd.

Clay, B. 2006. Conductivity (EM) Survey: A Survival Manual. Cultural Resource Analysts, Inc.

Crutchley, S. and Crow, P. 2009. The Light Fantastic: Using airborne LiDAR in archaeological survey, Swindon: English Heritage.

Dalan, R. 2006. A geophysical approach to buried site detection using down-hole susceptibitliy and soil magnetic techniques, Archaeological Prospection 13:182-206.

David, A. and Payne, A. 1997. Geophysical Surveys within the Stonehenge Landscape: A Review of Past Endeavour and Future Potential. Proceedings of the British Academy 92:73-111.

Evjen, H. 1938. Depth factor and resolving power of electrical measurements. Geophysics 3:78-95 . Griffiths, D. & R. King. 1981. Applied Geophysics for Geologists & Engineers. London: Pergamon Press.

Hasek, V. 1999. Methodology of Geophysical Research in Archaeology. BAR International Series 769. ArchaeoPress, Oxford, England.

Heimmer, D. and S. DeVore 1995. Near-Surface, High Resolution Geophysical Methods for Cultural Resource Management and Archaeological Investigations. U.S. Department of the Interior, National Park Service.

Kvamme, K. 2003. Geophysical Surveys as Landscape Archaeology. American Antiquity 68 (3):435-457. Kvamme, K. 2006. Magnetometry: Nature’s Gift to Archaeology, (ed.) J. Johnson, An Explicitly North American Perspective, University of Alabama Press.

Roy, A. 1972. Depth of investigation in Wenner, three-electrode and dipole-dipole DC resistivity methods. Geophysical Prospection 20:329-340.

Roy, A. & A. Apparo. 1971. Depth of investigation in direct current methods. Geophysics 36:943-959.

Weymouth, J. 1986. Magnetic Survey of Archaeological Sites in the Knife River Indian Village National Historic Site: Small and Non-Village Sites. Manuscript, Midwest Archaeological Center, Lincoln.

OK, looks like: Cond: 10.2 - 12.33 mS/m MagSus: .13 - .50 SI Mag: -3.56 - 3.39 nT Res: 8.42 - 29.63 ohms