Remote Sensing of Great Kiva Depressions at Cedar Mesa, Utah Report on Hampton Funded Research R.G. Matson UBC William D. Lipe Washington State University Periodic gatherings and recognized places of veneration help integrate residentially dispersed populations, such as those occupying Cedar Mesa. They contribute to the construction of a cultural landscape reflecting shared traditions. In the US Southwest, Great Kivas are perhaps the largest such integrative structures. In May, 2012, we used auger testing and geophysical prospecting to investigate four possible Great Kivas. Previously three possible Great Kivas were mapped and surface collected. As part of this 2012 project a fourth was also mapped and collected. In addition we mapped a defensive/plaza site and a large shrine site. The research was primarily supported by a grant from the Hampton Fund at U.B.C. with secondary support from the National Trust for Historic Preservation (USA). Sampling surveys in the 1970s (Fig. 1) had identified numerous small dispersed habitation and limited activity sites dating to AD 1060-1270 (late Pueblo II [P II] and Pueblo III [P III]) in central Cedar Mesa (Lipe and Matson, 1971, 2007; Lipe et al. 2012; Matson and Lipe 1975, 1978; Matson et al. 1988, 1990). Community centre sites (large sites with public architecture) were not found in that survey of some 800 km2 and approximately 500 sites. Shrines were first recognized by archaeologists in the northern Rio Grande area in New Mexico and not generally recognized in regions north of the San Juan River in Colorado and Utah (which includes the Cedar Mesa area) until the 1990s. In years subsequent to our 1970s survey, both community centers and places of veneration (shrines) were identified by others on Cedar Mesa (Hurst and Till 2009; Severance 1999, 2008). Two such possible “Chaco Outliers” community centres were the subject of our SSHRC funded investigations in 2009 (Matson and Lipe 2011; Lipe et al. 2012). These two, the Et Al and Owen sites, were mapped and collected then. Also the possible isolated Great Kiva HST site, (Fig. 2) located during our pre-field reconnaissance in 2008, was mapped. The HST site also had a sample of surface ceramic sherds analyzed via a “pick-up and put-down” non-collection procedure (Lipe et al. 2012). As part of our 2008 reconnaissance we re-visited a 1970s collected site, B12-1(42Sa4146) and recognized that what had heretofore been an anomalous feature was clearly a shrine which was intervisible with the Et Al site. Our May 2012 fieldwork focused on some of these sites. In addition to the Late Pueblo II (AD 1060- 1140) HST site (Fig. 2) we recognized that there was another site (not part of the 1970s investigations, but known to us then) in the Todie drainage (Fig. 3) that was similar to HST. It also showed a large depression without much evidence of surface architecture (although more than at the HST site), indicating that it was also a possible isolated Great Kiva. HST also has an associated prehistoric road trace, a feature not present at the Todie site. The Todie site was also mapped and collected in May 2012. Great Kivas? The main effort of the 2012 investigations was to get information on the nature of the four large depressions (at the Et Al, Owen, HST and Todie sites). We used two broad approaches, the use of 7cm augers to produce “drill logs” and remote sensing. Preliminary results of the remote sensing component, including a Hampton Report, Dec. 24, 2013 –1 combination of electromagnetic (EM) conductivity and ground penetrating radar (GPR) have been reported separately (Cross 2012). In addition to physical investigations at Cedar Mesa sites, results of previous GPR reconnaisssance of a comparable feature at a nearby Comb Wash site (Osburn 2005) were reproduced for calibration. As it stands, we are continuing to assess the correlation between geophysical and augering results with an aim to more fully understand related archaeological implications. Findings presented here are based largely on the interpretation of auger tests. We begin with our investigation of the two possible isolated Great Kivas, the HST and Todie sites. The ceramics analyzed from the 2009 work showed that the occupation of the HST site was in the expected “Chacoan times” (AD 1060-1140) when both Chaco outliers and Great Kivas are found over much of the northern Southwest, but the ceramics were of the “Kayenta” tradition which has never before been associated with Great Kivas (Matson and Lipe 2011). Thus it would be surprising if this was the location of a Great Kiva. Although the Todie site was not collected or analyzed in the 1970s we did visit it a number of times then and later (Matson in his investigations in 1991 and 1992, for example [Matson 1994]) and saw it as a “good” P II “Prudden Unit”. A Prudden Unit site is a group of surface rooms (circa 4-12) associated with a kiva (not a “Great Kiva”), and usually interpreted as an extended family residence. In later Pueblo III sites, these surface rooms are often masonry, but on Cedar Mesa in P II times these surface rooms are almost always “jacal” or wattle and daub, with perhaps sandstone slabs along the ground at the bottom of the jacal walls. Thus the presence (or absence) of the surface rooms is not always easy to determine by surface 1015 8.5 9.5 1010 9 10 9.5 rubble R1 1000 Datum 1 A Road 425a 28202 8 2 5 10 10 4 9.6 3 10 7 EM1 990 9 rubble rubble 6 R3 R2 TR1 TR2 Artifact DL2 Scatter 4 m radius 980 Grid North True 10.5 North DL1 2 m radius 10 0510 m 90 100 110 120 130 H.S.T. Site Site Datum Artifact Scatter 425a 28201 Arbitrary 10 m height Auger Hole map by J. Morin, R. Dickie, TR GPR Transect J. Clark, N. Fast, S. Matson EM Electrical Resistivity Transect Maximum extent of kiva Minimum extent of kiva Figure 2. HST site contour map. Hampton Report, Dec. 24, 2013 –3 540 96.8 97.2 97.4 97 97.8 97.6 Todie Flat Site 99 Feature 3 98 42Sa30148 530 97.6 J.Clark, N.Fast, S.Matson 99.2 99.6 98.2 99.4 Datum, at arbitrary 97.8 98.4 100meters elevation 98 98.8 Feature 6 Auger Hole 98.2 99.6 520 DL Dog Leash 98.4 GPR1 GPR2 98.6 Feature 9 Wall Feature 4 98.8 99 Depression Feature 1 6 4 510 3 99.2 Midden 8 1 99.8 99.4 Rubble 99.6 5 2 DL5 7 99.8 100 Feature 10 Maximum extent of 99.8 Feature 2 DL3 kiva, 7.0 m diameter 500 Feature 7 Datum A DL2 Minimum extent of 100.4 DL6 97.4 kiva, 4.5 m diameter Datum B 100.2 DL4 Main contour at DL1 DL7 97.6 1 meter interval 10 Feature 5 490 Feature 8 Intermediate contours 9 at 0.2 meter interval 97.8 99.8 True 100 480 98 North 99.6 99.6 99 98.8 Feature 11 99.4 98.4 0 5 10 meters 99.2 meters 98.2 470 450meters 460 470 480 490 500 510 520 Figure 3. Contour map of the Todie Kiva site inspection. Although the ceramics collected from the Todie site in 2012 are undergoing analysis now, preliminary inspection shows they include many of the expected Late P II (circa 1100) “Mesa Verde” tradition sherds (Blinman 2009) rather than the Kayenta ceramics found on the HST site. We had reason to doubt that the depressions on these two sites were Great Kivas even though the surface depressions appeared to be of the proper “Great Kiva” size (Fig. 2 & 3). Radar velocity varies with soil composition, texture and moisture content. Note: Approximate depth scale based on Indicated depth scale is provided for approximate reference only. estimated radar velocity v = 0.15 m/ns. Fundamental reference is two-way transit time in nanoseconds (ns). Figure 4. Ground Penetrating Radar results at the Todie Kiva site. Hampton Report, Dec. 24, 2013 –4 The Todie site provided some of the best correlations between auger and remote sensing. As Fig. 3 shows the augering demonstrated the presence of a kiva with a minimum diameter size of 4.5m and a maximum of 7m, a large kiva, but clearly not a Great Kiva. The GPR of transect 1 is in close agreement with the auger results, but the agreement is not as good in transect 2 (Fig. 4). This kiva had a floor (with abundant burned roof fall immediately above it) approximately 170cm below the current surface, or 190cm below the surface if the current depression is subtracted. Also found during the mapping of the Todie site is the very small Prudden unit noted as Features 7 & 8 in Fig. 3. The small depression immediately to the south of the small row of masonry rooms (Feature 7) was tested by auger and found to be a kiva probably only about 2m in diameter with a floor 85-90cm below the surface. The HST site also produced good results with the augering, but neither of the remote sensing techniques produced results that were clearly correlated. It can only be said that the EM-38 “Apparent Conductivity” technique was not in disagreement with the augering (Fig. 2, 5). The augering shows a minimum kiva diameter of 4.5m and a maximum of 7m. The floor of the kiva was between 165 and 185cm below the current surface and 195-225cm if the current depression is subtracted.