Single ping uncertainty

Atle et al: We have some very nice ADP wave data from the "surfzone" of a coral reef for which I would like to know the expected single ping velocity uncertainty - the reason is that I see a linear relation between the pressure std and the velocity std as one would expect for linear waves, but there is an offset. My bet is that this offset (ca. 8 cm/s) is part turbulence and part noise.  I am asking for single ping uncertainty because it looks like in wave mode the velocities are single pings. In our case we had a 2MHz system with a 0.75 bin for wave measurements. Is there a general formula for this uncertainty given frequency and bin size? Looking at the Brumley et al 1991 IEEE paper (on broadbands) it would appear so, but I thought I would check with you.

Stephen

Dear Stephen

We have had three great Nortek seminars in rapid succession but I am afraid they came at awful price when it comes to our ability to follow up on forum questions and emails that have come in.  My apologies.

In wave mode, we average for as many pings as there is room for within the averaging interval.  For a 2 MHz unit, the number of pings (along each beam) is 7 for 2Hz wave data and 15 for 1Hz data.  This gives an uncertainty (std. dev. of noise variance) of about 2.5 cm/s and 7.5 m/s in the vertical/horizontal for 2 Hz wave data and 1.8 cm/s and 5.3 cm/s for 1 Hz data. If you have collected wave data at 2 Hz, this should mean that the numbers are quite close to your estimate of 8 cm/s.

We have not published any noise numbers for the Aquadopp profilers but we use the Cramér–Rao lower bound during development to make sure that  we are not way off.  The numbers that are presented in the right-hand column of the deployment software are semi-empirical, in the sense that it is checked against actual data but scaled with the transmit pulse, receive window (bin size) and number of pings.

Best regards, Atle Lohrmann

Thanks