INTRODUCTION:

Salvors, Inc. has used the previous two seasons (1995 & 1996) as test seasons for the use of Differential Global Positioning Systems (DGPS) as a method of position location for salvage activities. The data presented in this report is from the 1996 season. The M/V Endeavor, captained by John Brandon, was equipped by Salvors, Inc. to receive DGPS locations. Additionally, other subcontractors were given the opportunity to purchase DGPS equipment and use it for testing in exchange for a discount on their contracts with Salvors, Inc. Craig McKinnon, captain of the M/V Gold Duster, took advantage of this offer. The readings contained in this report are from these two vessels, the M/V Endeavor and the M/V Gold Duster. The readings they took are from three sites: the Cabin Wreck, Riomar, and the Cannon Wreck. Their readings indicate that the DGPS method of positioning does offer a reasonable degree of accuracy that can be obtained at an affordable price.

EQUIPMENT:

Initially, Salvors, Inc. purchased the Differential Corrections, Inc., (DCI) premium one-meter accuracy service at the rate of $600 per year for the testing purposes. The M/V Endeavor was equipped with the mounted Garmin GPS receiver and a DCI RDS 3000 Differential Signal Receiver. Problems with this unit were encountered, however, as the Endeavor was unable to receive DCI's signal while working the southernmost contract areas, specifically, the Douglas Beach Site. Even with an antenna booster, the Endeavor was unable to receive the signal. In 1996, the Coast Guard officially opened the Cape Canaveral broadcast station, the signal from which has a broadcast range of 200 nautical miles. At this time, the Endeavor was equipped with a Garmin GBR 21 Differential Signal Receiver and they were able to receive the Coast Guard's signal on the Douglas Beach site. Craig McKinnon also purchased Garmin equipment for the M/V Gold Duster, specifically the Garmin 75 GPS receiver and the Garmin GBR 21 Differential Signal Receiver. The Garmin equipment advertises an accuracy of 5-10 meters (16.4 - 32.8 feet) with differential corrections but has the capacity to read to the three meter level of accuracy (9.8 feet), that is, the GPS receiver reads to one decimal place in the seconds column of degrees, minutes, and seconds readout. The signal broadcast from the Cape Canaveral Station is operating at an accuracy of 2.5 meters (8.2 feet) 95% of the time within 10 nautical miles from the station and 3.5 meters (11.5 feet) 95% of the time for the next 100 nautical miles. Cape Canaveral plans to update its technology in the future to a level of 1 meter (3.3 feet) 95% of the time within the first 100 nautical miles and 2 meters (6 feet) 95% of the time for the next 100 nautical miles. This technology is already in use at the Egmont Key station (Schenk, personal communication, 1997).

METHODOLOGY:

The Receivers for both vessels were mounted at the helm, but the antennas were placed in different positions. The antenna on the M/V Endeavor (a forty-foot vessel) was mounted between the prop-wash deflectors and readings were taken directly over the excavation unit. The antenna on the M/V Gold Duster (a 26 foot vessel) was mounted just forward of the helm and readings were taken at the boat's position. One reading was taken for each vessel at each location. The dates from which readings were selected for plotting were chosen at random from the sites and dates available with two sets of data (both sextant and DGPS readings). Adjacent readings from the available data were selected in order to see if the DGPS readings demonstrated the same patterning as the sextant readings. A total of 13 readings were selected for use in this report. More readings were plotted to ensure consistent accuracy; however, they were eliminated in this report in order to eliminate cluttered plots (Figure 1 demonstrates how close together the readings plot). These readings come from both vessels and three different sites.

The sextant readings were plotted using a standard three arm protractor and the standard site basemaps provided by the State of Florida Division of Historical Resources. DGPS readings, which were originally set to receive in NAD 83, were converted to UTM readings, set to NAD 27, using the Corpscon program produced by the U.S. Army Corps of Engineers. This was done in order to plot these readings on the standard basemaps, provided by the State, which contain UTM lines that were plotted utilizing 1927 datums (NAD 27). It should be noted that there is a small degree of error present in the conversion process (Challstrom, personal communication 1996). It is the recommendation of Salvors, Inc., that this error factor be eliminated in the future by eliminating the basemaps and plotting the locations in a programs that has maps based on 1983 datums (NAD 83). Those are the datums that the differential corrections from Cape Canaveral are based on, and therefore, NAD 83 is the datum that needs to be used by the GPS receiver in order to avoid compounding the error factor. If the State does not wish to eliminate the basemaps, then updating the maps to reflect UTM lines based on NAD 83 datums can eliminate their error factor. The difference between the 1927 datums and the more accurate 1983 datums is approximately 200 meters for the Northing reading and 20 meters for the Easting reading (Please see Table 1 on next page for a comparison of the UTM reading for these plotted areas in NAD 27 and the same UTM readings in NAD 83).

RESULTS:

The results of the plotted readings can be seen in Figures 1-7. All of the DGPS readings plotted between three and seven meters (9.8 and 22.9 feet) from the sextant readings. Figure 1 is the plots of the M/V Endeavor's readings from the Cabin Wreck, and they demonstrate the close proximity of the DGPS readings to the sextant plots. Figure 4 demonstrates the same for the Endeavor's readings on the Riomar Site, and Figure 7 demonstrates the proximity of the plots for the M/V Gold Duster's readings on the Cannon Wreck. The DGPS plots demonstrated the same patterning as the sextant reading plots. This is illustrated in Figures 2,3,5, and 6. Figure 2 demonstrates plotted readings for the first three excavation units on Cabin Wreck, and Figure 3 demonstrates plotted readings for the next three excavation units on Cabin Wreck. Figure 5 demonstrates plotted readings for the first three readings on Riomar, and Figure 6 demonstrates plotted readings for the last two plotted readings on Riomar. The plots for the Gold Duster do not show as consistent a pattern as those of the Endeavor's; however, this may be due to a difference in the way the captains of the two vessels work, and the fact that the antenna on the Gold Duster was mounted forward of the helm. Because of this the readings may be reflecting the position of the front of the boat rather that the actual position of the excavation units.

TABLE 1: NAD 83 AND NAD 27 DATA COMPARISONS

Figure 1: Cabin Wreck Plotted Reading
Figure 2: First Three Plots on Cabin Wreck
Figure 3: Last Three Readings on Cabin Wreck
Figure 4: Riomar Plotted Readings
Figure 5: First Three Readings on Riomar Site
Figure 6: Last Two Readings on Riomar Site
Figure 7: Cannon Wreck Plotted Readings

CONCLUSION:
    There are numerous factors that need to be taken into consideration when assessing the benefits and drawbacks of changing from a sextant and beach marker system of position location to Differential Global Positioning Systems (DGPS).  The accuracy of the equipment itself is one consideration, and the data in this report demonstrates that the systems tested in 1996, that is the Garmin GPS receivers and the Garmin GBR 21 Differential Signal Receivers, have an accuracy of between three and seven meters.  Eliminating the errors that result from the need to convert these readings into NAD 27 readings, and position fix averaging may further increase the accuracy.  Additionally, requiring a standard positioning of the antenna on the back of the boat will increase the accuracy of the actual position of the excavation units.  This may not be possible on some vessels, however, and it may be necessary to have data recorders record the vessel’s heading in order to re-establish its exact positioning when the reading was taken.  It should be taken into consideration, however, that there are other factors affecting the accuracy of the readings.

    Equally as important as the accuracy of the equipment is the accuracy of the equipment operator and the person who plots the readings.  Changing to DGPS as a system of position location will reduce the number of opportunities for operator errors.  The sextant and beach marker system, while theoretically extremely accurate, provides numerous opportunities for errors.  The first, of course, is the taking of the reading itself.  Failure to consistently take readings from the same position on the vessel, particularly a larger vessel, rough sea conditions, inexperience on the part of the operator in addition to the fact that this system relies on physical variations such as eyesight, and general failure to consistently take readings in a meticulous manner are factors that could affect the accuracy of the readings that result.  Furthermore, there is another major area in which errors could occur, and that is in the plotting of the readings.  The basemaps are marked with triangular symbols with a point in the middle to represent the beach markers.  Upon consulting with several subcontractors, it is evident that there are variations in the way they plot their locations.  Some will align the three-arm protractor with the point in the middle of the triangles, as was done for this report, while others will align it with the apex of the triangle.  Still others may simply make sure that the middle line of the arms of the protractor falls somewhere on each of the triangles they are using.  This point was illustrated when, during the course of preparing this report, two different people plotting the sextant points on the maps came up with variations in where each had plotted specific points.  It was therefore decided that one person should plot all of the data in an effort to remain consistent.  While the errors that result from these areas where variations can occur might not be huge, they may certainly exceed three to seven meters.  DGPS eliminates these opportunities for human error.  The most an operator will have to do once the equipment is set up and operational is to write down several readings and average them.  The equipment does the rest.  Errors within the system will still result in more consistent readings that those of the sextant and beach marker system, once the myriad of variables is considered.

    One final benefit of changing to DGPS is the fact that the data will be far easier to plot and access in the future.  The annual relocation of beach markers, as well as the need to re-position them during the season due to storms and vandalism, has resulted in a number of different positions throughout the years.  As the State is no doubt aware, computerizing the data for easier and more efficient access will most likely require a new program written specifically for this task.  The program will have to encompass converting the sextant angles into, ultimately, UTM readings, choosing the year and thereby the positions for the beach markers that year, choosing the specific beach markers, which may not necessarily be adjacent to one another, and finally, plotting the locations.  The plotting and access of future DGPS readings will undoubtedly be far easier.  The data will simply have to be downloaded into an existing plotting program and plotted.

    In summation, it appears that the benefits of changing to DGPS far outweigh the drawbacks.  While in theory the sextant and beach marker system may be somewhat more accurate that affordable DGPS systems, in reality, with all of the variables that must be considered, DGPS will ultimately provide more accurate data.  Additionally, as the technology is updated, the accuracy will increase and the cost of more accurate systems will become more affordable.  It is the feeling of Salvors, Inc., that accurate DFPS readings can be obtained at a reasonable cost (Please see the equipment recommendations that follow this report).  It is, therefore, the recommendation of Salvors, Inc., that the Archaeological Guidelines be changed to incorporate DGPS as the primary method .

Back to Commercial Salvage & Archaeology: Communications & Progress - 1999