The recent correspondence from William Trayfors was superb ("Lessons to be learned from BVI groundings," Issue No. 82). I have obtained the relevant charts with the intention of retelling the story to my piloting classes. Although Mr. Trayfors makes several important points, I would suggest that there are a few corrections and additions to be made.
1. The corrections required to convert between chart locations and WGS-84 positions from GPS should have been 0.24 minutes of latitude and 0.15 minutes of longitude, not 0.24 and 0.15 nautical miles. There is no difference for the latitude, but the distinction is very important for longitude. One minute of longitude is one mile at the equator, but becomes shorter as latitude increases. In this location the difference is about 5%.
2. I have used GPS with charts requiring this kind of conversion, and I think it is a treacherous business. The simplicity of the math belies the confusion created. In the case described, converting from WGS-84 to chart positions requires the addition of a factor to latitude and the subtraction of another factor from the longitude. Converting from chart to WGS-84 positions reverses the processes: subtraction for latitude and addition for longitude. It sounds simple, but even when I was calmly planning the next day’s sail, repeated checks could not dispel my uncertainty about having done it all correctly. Performing the conversions while in motion near some hazard was worse. Thankfully, I can only imagine trying to do them during the excitement of a grounding. If you doubt that errors would creep in, I think the article proves my point. While correctly stating that the longitude factor is to be subtracted to get a chart position, the factor was obviously added to get the "corrected" chart position that is given.
3. With those difficulties in mind, it would clearly be preferable for the GPS unit to display positions that could be plotted directly on the chart in use. In referring to WGS-84, the article failed to make it clear that GPS units allow the user to select a chart datum. The unit will then output positions that match with charts using the specified datum. Horizontal datum is not clearly stated on all charts, so it is not always that easy; however, users should be aware of this capability and make use of it when appropriate. Most of the GPS users with whom I have talked do not even know their own units have such a function, and I think this story missed an excellent opportunity to tell them.
4. I believe that cautioning people not to rely too heavily upon these electronic tools often falls on deaf ears. People will continue to use their black boxes unless they really understand the problems with them. Katama’s grounding is a good example to use since it appears to have resulted not just from reliance on GPS, but on a lack of recognition of the chart datum as an issue and the use of a GPS that was set to a datum different from the chart. It could be argued that the grounding would not have occurred at all if chart datum had been better understood or if the GPS had been properly set to the chart datum. I will make that argument in my classes in the hope that it will prompt the students to learn more about both their GPS and their charts.
Phillip Jones is a physician, a U.S. Power Squadron member, and an amateur radio operator living in New Orleans.
Bill Trayfors responds:
Phillip Jones makes some excellent points.
1. He is, of course, correct in noting that 0.15 minutes of longitude does not exactly equal 0.15 nautical miles except at the equator. The difference at 18° north (Virgin Islands) is very small (5%) and I chose to ignore it while making the main point. At higher latitudes the difference can be very significant.
2. I agree completely with his notes about the uncertainties of applying corrections to charted positions in order to match GPS output. One is never quite certain that the corrections have been computed and applied correctly.
3. Yes, many GPS units allow the user to choose among various chart datum "standards." I have tried using this capability several times on my two GPS units and concluded thatfor me at leastthis is not a good solution. First, the user must determine which datum was used in constructing the chart. Second, considerable manipulation of the GPS units is necessary to find and choose the right datum (thus leaving the GPS in a "nonstandard" condition, i.e., not set to WGS-84). Having done this, the user is never quite certain that GPS positions will actually match charted positions. In my trials from known nautical positions, I found considerable error even after choosing a chart datum matching that of my chart, leaving me wondering whether the chart or the GPS datum or the GPS system was responsible for the lingering error. Bottom line for me: I’d prefer to use a known standard (GPS WGS-84 datum) and assume a larger error in charted positions.
4. I agree that many mariners will likely continue to use their black boxes without taking account of datum errors. But we must try to get the hazards of this practice more widely known and accepted. To this end, I’m collaborating with members of the scientific community on an article which focuses specifically on chart datum differences and GPS errors. The challenge is to take some pretty esoteric stuff and translate it for the general boating community.
I would like to reiterate a point made in my article: I have no direct knowledge that Katama’s grounding was due to GPS-datum errors. The circumstances and available figures certainly point in that direction, but other causes are possible. Whatever the actual cause, the case of Katama serves conveniently to draw the GPS-datum error lesson.
Military aircraft pilots are taught that "the environment is there to kill you; don’t give up anything to it," meaning one should use all available data and techniques to stay alive. The nautical equivalent might be, "GPS is a wonderful tool. Use it, carefully, but don’t give up your common sense and, especially, your other navigational tools and skills."