Page 151 - Green - Maritime Archaeology: A Technical Handbook. 2nd ed

130 Maritime Archaeology: A Technical Handbook, Second Edition
only show green measurements. Do not delete any measurements at this stage.
If after rejecting one measurement the situation does not improve, look for any other obviously wrong measurements and reject them one by one, readjusting in between. Do not reject so many measurements that less than four are available on each point. If enough measurements are rejected then almost any network can be made to fit.
For baseline measurements the blunders tend to be longer than the “true” values, because snagging or bowing always makes these measure- ments longer. If a short blunder occurs because of a misread tape, then the value tends to be considerably shorter and is obviously wrong. Check dive logs to see if any measurements are likely to be blunders and check if they are candidates for rejection. If they are suspect but you still want to use them, then increase the estimated error for the measurement and the adjustment will rely on them less.
Start with a small group of points and measurements and get them right. Ignore any other measurements until later. When the small set is correct start adding more points and their measurements. Once a small set is assumed to be right you can temporarily fix their positions be- fore including more points and measurements. If some of the included measurements are wrong, then they will be easier to identify. It is essential that the positions of these points are freed once the whole network is to be adjusted.
Where multiple measurements have been made be wary of rejecting all of the measurements except one just to make the network fit. As multiple measurements have been made the most likely value should be the average of all of the measurements.
15. Insufficient Measurements
When rejecting measurements you must ensure that too many measure- ments are not rejected. To position any point we need three measurements; additional measurements are used as checks to ensure that the position is correct. As a minimum there should be three baseline measurements plus one depth measurement for each point.
If it becomes necessary to reject more measurements than this, repeat measurements should be made. To allow readmission of rejected measure- ments you should not delete measurements but ignore them. Once a network has been positioned successfully then some measurements could be deleted, however; it is better to leave them in the project in case the survey is to be reprocessed again later.