The Petaga Point Site - 21ML11
A Geophysical Investigation in Support of the Kathio Archaeology Day Public Excavation Program
A geophysical investigation was conducted at the site of the Kathio Archaeology Day Public Research Program (KADPRP). The investigation consisted of ground penetrating radar (GPR), magnetic field gradient, and electrical resistance surveys over a 35 m x 30 m area that encompassed the 2006-2010 KADPRP excavation units. The 5 square meters of KADPRP excavations have unearthed remnants of an undocumented excavation block associated with the University of Minnesota's 1960's era research at Petaga Point, as well as intact portions of an archaeological horizon that appears to represent a burned house feature.
The U of M's previous excavations within this portion of the Petaga Point site occurred during the 1966 and 1967 field seasons. Although the complete results of these excavations have not been published, a plan map depicting the outlines of a burned house (and it's surrounding excavation block) was published, as was a brief description of the excavation results, which included a reference to additional undocumented test excavations in the vicinity of the documented house (emphasis added below). Additional and more detailed project background may be found here and here.
An excerpt from "The Northern Margin of the Prairie Peninsula" by Elden Johnson, 1971
It is the house type associated with two Kathio Focus sites at the recently excavated Mille Lacs sites which may show the closest parallel with the early Missouri tradition sites. Bleed excavated a corner of one such house at Petaga Point on the last day of his 1966 field season (Bleed 1969:10-11) and the excavation was extended the following field season by the author. The house floor as it was exposed is shown in the diagrammatic sketch (Fig. 3 [reproduced below]). The house is rectangular in plan, measures approximately 6 x 11 meters with an additional 2 meter entrance passage, and has a floor depressed approximately 50 cm below the original ground surface. Post molds are peripheral along the four walls and the main fire pit is offset and located toward the entrance. No interior post molds indicating a roof support pattern were found, and the house is oriented on a northwest-southeast axis. The structure had been burned and masses of charcoal and partially burned wood lay over the floor. Eastern triangular points and Kathio sherds lay under the charred material on the original floor surface at several places within the structure. Test excavations at adjacent locations on the site marked by a distinctive color change in the grass cover located two other structures, one of which had also been burned. The remaining undisturbed grass cover on the site indicates the presence of at least 5 structures, clustered together but in no formal spatial arrangement. There is no indication of any defensive fortification at the site. House forms similar to those at Petaga Point have been excavated at the Cooper site where they occur stratigraphically underlying a later protohistoric Midewakanton Dakota occupation level.
The objective of the geophysical investigation was to place the KADPRP excavation in context with both the buried archaeological feature(s) and the documented and undocumented 1966-1967 excavation blocks. Geophysical fieldwork was conducted during over the periods 9-10 August and , 2011.
The results of our investigation are presented in the form of three figures. Figure 1 is an interactive graphic summarizing the geophysical survey results and presenting some initial interpretations of these data. Figure 2 uses a modified display range and palette to highlight high-amplitude patterning in the vicinity of the KADPRP excavation units. Figure 3 provides the estimated locations of documented and undocumented 1966-1967 University of Minnesota excavation blocks.
All project data and imagery are available for download at the bottom of the page.
Figure1: Interactive Data Analysis and Initial Interpretations
The following interactive graphic presents the results of the geophysical investigation along with some initial interpretations of the data. Click the button labeled autoplay to view the GPR data sequence.
a current Flash Player is required for viewing this interactive graphic
Figure 2: High-amplitude patterning near KADPRP excavation units
The following four images emphasize high-amplitude signal in the vicinity of the KADPRP excavation units. Click on the thumbnails to view at a larger scale. The display range and palette is presented below the images. The display range is expressed in terms of standard deviations (σ) from the mean value of each data set.
15-25 cm below surface
20-30 cm below surface
25-35 cm below surface
30-40 cm below surface
(mouse over to view excavation plan map)
Figure 3: Estimated locations of 1966-1967 excavation blocks
The estimated locations of the documented and undocumented 1966-1967 excavations at Petaga point are provided in this interactive graphic. The documented plan map was scaled, rotated to magnetic north (note: magnetic north in 1966 was about 6 degrees east of magnetic north at present), and it's location was inferred from the magnetic data. The location and geometry of the undocumented plan map was inferred from a combination of the magnetic and GPR data.
Discussion and Recommendations
The GPR functions by sending high frequency electromagnetic waves into the ground from a transmitter antenna. Some of these waves are reflected back to the surface as they encounter abrupt vertical changes in the dielectric permittivity or electrical conductivity of the matrix through which they are traveling, and are detected by a receiver antenna. Diffuse vertical changes in these properties do not produce significant reflections.
In the intact portions of the KADPRP excavations an archaeological horizon was encountered. The most significant stata within this cultural horizon was a distinct burned layer, which consists of the charred remains of a protohistoric house feature. The stata overlying the burned layer is described as a "displaced soil that was placed back over the house (perhaps to create an earthlodge)". The boundary between the burn layer and overlying strata is described as "sharp".
Based on our understanding of the source of GPR reflections and the observed archaeological evidence, it seems reasonable to conclude that the high amplitude GPR reflections visible in depth slices from between approximately 10 cm and 35 cm below surface represent the burned layer. The relative amplitude of these reflections is likely proportional to the thickness and density of the burned layer, as well as the "sharpness" of the boundary between the burned layer and overlying strata. In areas where bioturbation has mixed the burned layer and overlying strata (in other words, has reduced the sharpness of the boundary) we can expect a reduced signal response.
The limits of both the documented and undocumented U of M excavation blocks appear to be largely invisible to the GPR, which suggests that there is little to no dielectric contrast between the excavation block edges and the back-filled soils.
The U of M excavation blocks do appear to be partially visible in the magnetic field gradient survey results, with the interior of the blocks appearing as a faint magnetic low (negative data appear as lighter shades in these plots). Negative signal may be induced by a lack of magnetically enhanced A-horizon soils within the blocks, while positive signal (dark areas of plot) occur at the edges of the blocks and over baulks within the blocks.
The magnetic data are dominated by the presence of relatively intense bipolar remanent magnetic fields - often referred to as magnetic dipoles. Unfortunately these dipoles tended to mask or obscure the more subtle induced magnetic signal response associated with the 1960's excavation blocks and in-situ archaeology. Some of these magnetic dipoles are undoubtedly associated with the presence of igneous and metamorphic cobbles in the glacial till, while other fields are likely caused by modern metal debris associated with archaeological investigations (pinflags, gutter spikes, nails), or iron trash and debris associated with the historic farmstead. Another possible source of magnetic dipoles are iron artifacts associated with the protohistoric occupation of the site.
The intensity of magnetic dipoles decreases with greater depth below surface, while the width of anomalies tends to increase with greater depths below surface. With this in mind it is appropriate to interpret small, “bright” magnetic dipoles as originating close to the ground surface and weaker more diffuse anomalies as originating deeper in the soil profile. A final source of very intense signal within the magnetic data is lightning induced remanent magnetization (LIRM). An explanation of LIRM anomalies may be found here and here. The LIRM signal is also a form of signal clutter that may have obscured archaeological signal of interest.
The estimated location of the documented 1966-1967 U of M excavation block was arrived at through examination of the geophysical data. Although evidence of excavation block edges in the data is rather sparse, the GPR data do appear to provide circumstantial corroborating evidence in the form of a long linear anomaly that leads out of the extended entryway of the house to the north-northwest. This linear anomaly may represent soil that was compacted by foot traffic leading to and from the entryway. We can say with certainty that compaction does cause a detectable GPR signal response in these soils due to the distinct "two-track" linear anomaly created by maintenance vehicle traffic over the site.
The location of the undocumented 1967 excavation block was estimated after a careful examination of both the magnetic and GPR data. The eastern portion of this block appears as a rectilinear magnetic anomaly, and although the block itself is not visible in the GPR data, it's presence can be surmised by the lack of a strong GPR signal response from the burned layer in previously excavated areas.
It should be emphasized that the initial interpretations offered above are preliminary. Competent interpretation must be a cooperative process involving both the archaeological geophysicist and the archaeologists who are most familiar with the specific cultural context of the site. These initial interpretations should be reviewed by archaeologists familiar with Petaga Point. Comparison of survey results with the range of expected feature types and intra-site patterning may result in different or elaborated interpretation.
Testing (or ground truthing), if it may be performed, will greatly inform interpretation of these data, and is strongly recommended. Verification (or refutation) of preliminary interpretations and insights into feature composition and geology can allow us to revise or elaborate our interpretations, and to do so with greater confidence.
Raw Geophysical Data and Imagery Download
Click here to download geophysical data and images. Please note that the data have associated explanatory metadata files. Standalone versions of the interactive graphics may be downloaded here. The interactive graphics may be opened on any computer that has a web browser and FLASH player installed.