Well, at last we have rolled out the first corrections to all of the data logged for TeamSurv. Part of the reason that it has taken so long is because we have also been implementing a framework that makes it easy for us to roll out new or improved algorithms in the future.

As well as the corrections (of which more below), we have also made enhancements to the validation of uploaded data, so much of the rogue data is either corrected automatically or filtered out. Also, now that we have about 2.5 million points logged, we have implemented zoom tables, so as you zoom out it doesn't draw all of the data points any more.

We have rolled out corrections for the orientation of the vessel, and also for tide heights.

Vessel Orientation

We offset the GPS position to that of the depth transducer, using your calibration data and either the compass heading or the GPS Course over Ground. Ideally we use the compass heading, as this allows for tidal streams and drift or leeway, but when this is not available from your instruments we use the GPS COG instead.

Vertically, we use the depth of your transducer from the calibration, and the depth below transducer output from your depth sounder.

Tide Height Corrections

Predicted tide heights are generated from TideWizard, which gives harmonic tide predctions for al primary and secondary tidal stations in and around each of the trials areas (note that there are no tides in the Baltic). This is much more accurate than using the the time and height differences for secondary ports, as given in the tide tables.

We then assign an area of influence to each tidal station. The computer initially generates a first pass of this, and then it is reviewed and modified as necessary manually.

Now, for each point in the track, the tide height prediction for the appropriate tidal station is worked out, and applied to the logged depth to correct it to sea level.

Forthcoming Corrections

We will shortly roll out sea level corrections, which adjusting for actual, rather than predicted tide levels. These corrections use tide gauge networks and apply the difference between predicted and measured tides at those locations. This correction is seperate from tide level corrections as the tide gauge networks are much sparser than the tide height predictions.

We are also implementing corrections for the speed of sound. As a depth sounder is an accoustic device, changes in the speed of sound affect the recorded depths. These can be quite significant, e.g. 6% of the depth in the cold, fresh water of the Curonian Lagoon in Lithuania. Again, we use networks of data monitoring stations to give us the temperature and salinity data, upon which the speed of sound depends, and then calculate the corrections.