Dealing with Rebels

The GPS units in differential mode (DGPS), which are to be used in a Hydrographic Survey are usually proved to the clients prior to the commencement of survey. Even though there are  several proven methods to perform this verification, the most common and simple method is to log the position of a known station using the equipment to be used, for a certain period of time and average the fixes to compare it with the established station co-ordinates.  With negligible mean difference and standard deviation of the observed and Known coordinates, we prove that the DGPS Unit has the prescribed accuracy level.

 But anyone who has worked with GPS receivers in the field knows that they are prone to occasional position jumps due to various reasons, even if the user has obtained commercial codes for precise positioning. A positional jump (Also referred as Bad coordinates) means the GPS receiver shows a different geographical position rather than its own position due to an algorithm error, and by naming it as a “Jump” it is evident that on the subsequent solution, it reverts back to its original Position. Often these odd position jumps are far apart, even to be on the opposite side of globe. If these occasional jumps are not removed from the observed data set to prove the positional integrity of an equipment against a known coordinate, it is going to affect the average positional value, sometimes in a more profound way than you can ever imagine.

In a world, that believe in statistics, where we are in a constant pressure to remain inside the threshold limits, we can’t afford to have a high standard deviation value in our reports. Certain GPS receivers are provided with filters, that allows one to omit values if subsequent observation differs in certain value parameters, like you can define that any position shift of more than 5 meters in two subsequent fixes should be discarded. But almost all the calibration procedures insist in removing all the filters from GPS equipment prior to calibration.

So the only solution to this problem is to remove these positional jumps from the observed data by filtering it manually before considering it in the report.

For this purpose, in SURge, we have a data cleaning utility to single out and clear these positional jumps

That’s why we show the following warning before opening the data cleaning utility…

In SURge, we understand the value of our profession and never allows any kind data manipulation as almost all the reports are directly generated from the raw observation files, which gives the user no room to manipulate and helps to attain the confidence of the client.