Ground Penetrating Radar (GPR) data were collected on site SCrI-333, a pre-historic Chumash village, near Fraser Point on Santa Cruz Island, California. A 60 x 60 meter area within the site was surveyed in July 2002 by members of the GDEP program. The site exhibits about 38 circular-to-oval depressions that correspond to pit houses (Wilcoxon 1985). Around the rims of the pit houses there are extensive accumulations of shell middens. The latter are layers of deposited shell. Radiocarbon dating (Wilcoxon 1985) indicates that site SCrI-333 was occupied from 6,000 to 1,800 YBP (years before present), making this early site highly significant in terms of its extent and complexity. The GPR method measures electrical properties in the earth. In the SCrI-333 setting, GPR should be able to show differences between sterile soil and shell rich middens. Over 190 GPR data files were carefully examined for the high reflection areas that might represent pit house floors and shell deposits. Two distinct reflection zone areas were visible in the data. The deeper zone appeared between 1-2 meter in depth and it may correspond to whole red abalone shell concentrations. The upper half correlates more to surficial pit house patterns.

Santa Cruz is the largest island off the coast of California (Fig. 1). Located between Anacapa and Santa Rosa islands, it lies from 19 to 25 miles off the adjacent mainland coast between Ventura and Santa Barbara. The earliest evidence of human habitation on the island could be as old as 10,000 years before present, left by the hunter-gatherer people who preceded the more settled Chumash. For at least the last 6,000 years, Santa Cruz and most of the other Channel Islands were occupied by the Chumash. At the time of first European contact, Santa Cruz supported 11 villages and about 1,100 people (http://www.tnccalifornia.org/).

Site SCrI-333 (formally known as SCrl-3) is located near Fraser Point on the southwestern portion of the Island. The site is characterized by deep layers of deposited shell and extensive pit houses (Fig. 2). Many archæological surveys and excavations have been conducted on this site. The most recent were by Mr. Larry Wilcoxon in the 1970's and early 1980's and the 2002 GDEP Program with Dr. Michael Glassow of UCSB.

The objectives for this research were:

• Continue geophysical work on site
• Process computer data from 2002 GPR survey
• Assess potential for identifying house floors, red abalone shell layers, and other features of interest
• Provide framework for augering programs to identify anomalies
• Look at correlation between surface features and geophysical anomalies

Geophysical methods are very important in archæology. Doing geophysical surveys will help archæologists understand and interpret archæological sites in a more scientific way. Ground-Penetrating Radar (GPR) and other geophysical methods will not only preserve non-renewable sites, but can reconstruct a site, find burials, uncover deeply buried sites, etc. Tens of thousands of archeological projects are conducted each year in the U.S.: Only 1% have geophysical surveys conducted prior to excavation.

GPR uses high frequency pulsed electromagnetic waves (generally 10 MHz to 1,000 MHz) to acquire subsurface information. Energy is propagated downward into the ground and is reflected back to the surface from boundaries at which there are electrical property contrasts (http://fate.clu-in.org/) (Conyers and Goodman 1997).

The pithouses were shaped like half an orange, about 3-5 meters in diameter. In site SCrI-333 there are about 38 circular-to-oval depressions, some representing pit houses. The depressions form curvilinear patterns. The Chumash Indians used to throw their garbage to the side of the houses. Over time the garbage piled up and compressed. The GPR expectations are to identify house floors and dense concentrations of red abalone shell. As red abalone is a very sensitive climate change indicator, dense concentrations of this material at certain depths show the existence of different climate conditions at the time of deposit (Glassow 1993).

In theory, as the houses were destroyed, the materials that were used to construct the houses were collapsed onto the floors (Fig. 3). Over time the materials decomposed and they became new layers in the pit, new potential GPR reflectors. Wind would blow sand, shell, grass, and other materials into the pit. Sand and silt infill will likely appear somewhat transparent and non-reflective to the GPR. Layer thicknesess would be roughly equivalent to the length of occupation, with youngest materials associated with the top layers, and older materials at the base. The GPR expectations for the natural processes are that the hard pack floors should show as prominent reflectors, pit house fill should be fairly non-reflective, and there should be reflections associated with dense abalone shell concentrations, particularly for whole shell lenses.

The GPR data were obtained at site SCrI-333 on Santa Cruz Island in July 2002 (Fig. 4). At site SCrI-333, nine 20 x 20 meter grids were set up, for a total area of 60 x 60 meters. The GPR antenna was moved south to north for each profile starting at the southwest corner of each grid. The line spacing was at one meter intervals. The GPR was set for two-way travel time of 60 nanoseconds (ns), which translates to approximately 3 meters (there is some vertical exaggeration) (Dietz 2002).

Our objective was to locate the pit house floors and other anomalies with the GPR data collected. There were 191 data files to view. Radview software (http://www.geophysical.com/) was used to open and view the GPR data. Based on the orientation and color of the constituent soils and deposits, we (Cruz and Cortez) were able to pick out and map the reflection amplitudes in site SCrI-333 (Fig. 5). The higher the amplitude, the more visible the reflection, marked by yellow and white. Medium amplitude colors are blue and purple. Lower amplitudes, blues and blacks, indicate small differences in the electrical properties between layers.

We were able to approximate the locations of subterranean shell deposits. We found that the GPR anomalies were at various depths within the 60 X 10 -9 seconds (60 ns) record. So data points were split into two groups: upper half and lower half. The upper half anomalies clustered in the upper 20-30 ns range, approximately 0-1 meters. The lower half anomalies clustered in the 30 to 50 ns range, approximately 1 to 2.5 meters.

When we finished picking the reflection areas, we created a database in MS-Excel ™ in which we input 100 for high reflection areas, 50 for medium reflection areas and 20 for low reflection areas. The Excel spreadsheet generated was moved into Transform ™ software (1992-1996 Fortner Research LLC version 3.3) to create an image or map. The new image was then moved to Adobe Photoshop ™, in which was cropped and layered the maps to create a composite. The layers are as follows: the new GPR data interpretation, Glassow's topographic map, and Wilcoxson's archæological drawing.

As in any research work, there are error factors that we must acknowledge. Occasionally error is embedded in mapping: For example, in my research there was one extra data line collected in all nine grids. Wilcoxson's map may be misaligned by about 1 to 2 meters from the geophysical grid. Dr. Glassow's topographic map has a box-like figure in the middle of the map, representing a measurement error.

In the summer of 2003, an augering project was conducted to verify the target anomalies on site SCrI-333. The project was led by Dr. Glassow. The results were consistent with the predictions. Within 1-2 meters there was evidence of dense concentration of abalone shells. Within the pit house depressions, the subsurface deposits were also as predicted: wind blown sand and fragments of shell.

There is a great need for geophysical methods in archæology. Geophysical instruments bring valuable insights into cultural and natural processes of archæological sites formation. In addition, there is minimal disturbance to important cultures properties and the results can be used to guide archæological exploration and excavation to maximize data collection.

Ultimately, this kind of research can contribute to a better understanding of prehistory and cultural change. Our studies' results support the idea that future archæological work would benefit greatly from a well planned geophysical program on the Channel Islands.

I would like to thank Dr. Elizabeth Ambos and Dr. Daniel Larson for their support and encouragement throughout these days and to Michael Cruz for his knowledge and help. Thanks to Mr. Keith Miller from Lakewood High and Sachiko Sakai from UCSB, Dr. Mike Glassow from UCSB, and Terry and Matt for their support. Thanks also to Tom Tran for his help in the computer lab and to Dr. Laura Breece for getting me into archæology. Ann, Jenn, friends and family deserve recognition for their support, as do Crisanne Hazen and all the GDEP students and staff.

I would also like to thank the NSF for their grant #GEO-1009891.

• Conyers, L. B., Goodman, D. 1997. Ground Penetrating Radar; An Introduction for Archæologists. AltaMira Press, Walnut Creek, California.
• Dietz, A. 2002. Detailed Stratigraphic Analysis of Archæological Sites Compared With Geophysical Surveys: A Case Study from Chumash Sites on Santa Cruz Island. GDEP 2002.
• Glassow, M.A. 1993. Changes in Subsistence on Marine Resources through 7,000 years of Prehistory on Santa Cruz Island, California pp.75-94
• Salazar, J. 2002. Aerial Photography Case Studies: Santa Cruz Island Site SCrI-333 and Navan Fort.
• Sturtevant, W. C. (ed.). 1978. Handbook of American Indians, 8th Volume Smithsonian Institution, Washington, pp. 505-537.
• Wilcoxon, L. 1985. Subsistence and Site Structure: An Approach for Deriving Cultural Information From Coastal Shell Middens.
• http://www.tnccalifornia.org/our_proj-santa_Cruz_island/closeup.asp
• http://fate.clu_in.org/gpr.asp?techtypeid=41
• http://www.sbnature.org/research/anthr/chumash/daily.htm