AWAC tilt limits

Hi again!

I have a question about correction of wave data when the AWAC is tilted between 5 and 10 degrees: is there a way to correct the wave data? I can retrieve the WAP data and of course I can see -99999 default values for some ensembles and some variables, but is there a way to correct the remaining values?

When pitch and roll are more than 10degrees, can you please confirm that the data is completely lost and there is nothing to do on it?

Thanks a lot for the information!

Camille.

Hello Torstein

We've been measuring waves on brasilian shores for more than 2 years now, with more than 60 deployments so far, always on water depths between 10 and 20 m. We had excessive tilt problems sometimes, and I agree that 10 degrees is more or less the limit for obtaining good AST data. 

This week I got something new: when opening my latest deployment data on STORM, I noted that we had tilt well over this limit, up to 25 degrees of combined pitch&roll. Nonetheless, the AST data seems all right, with no data been erased from the time series, since there was no burst with more than 10 % bad detects. When I processed the data using only AST, the measured values seem very coherent with the environment where the AWAC was deployed. I've also compared the results with the results from the pressure data and it also seems reasonable, with the mean period from the pressure data a bit higher than the AST (what is very logical from our experience). I've also tested the pitch&roll sensors after the deployment and they are just fine.

As I was using the 10 degree limit to exclude data, now I'm on a bit of a dilema: I have to prove to my client that the AST data is all right even with the high tilt. The big difference on this deployment from the other we are used to do is the water detph: on this last one it was 33 m, well above the depth we are used to work with. Do you think the bigger depth might be the reason why AST worked even with a higher tilt?

Any ideas are welcomed...

Thanks,

Pedro

Previously Torstein Pedersen wrote:

The following question was recently presented on the forum in another discussion thread.  I have moved it here since it is important to understand for system performance and deserves its own attention.

 

Hello mister Atle

I'm write from Chile, and i need answer the following questions to AWAC 600 kHz

1. Confirm that the Pitch an Roll readings must be between +-5° for clasificate a data, how, good data

2. What is the Heading range? Is the same that Pitch and Roll; +-5°

3. Is posible that Nortek, check the attitude parameters?

thanks, best regards,

Juan Pablo

(The file size is 5.022 KB) impossible attach.

Tilt

Juan, for the wave measurements the tilt becomes important for the performance of the AST.  The velocity measurements and processing are not as sensitive to tilt. 

The acoustic energy from the AST ping reflecting from the surface becomes more more reduced as an AWAC becomes more and more tilted.  The magic number for permissible tilt is 10 degrees; above this one will see that finding the ocean surface becomes challenging when attemping wave measurements.  What often happens is that the there are too many samples which we refer to as Bad Detects  and a wave burst can not be properly processed.  It is for this very reason that we do not bother to use the three other beams for AST measurements; these beams are angled at 25 degrees from vertical and would give poor results.

The resulting tilt from the combined pitch and roll seems to really deteriorate above 10 degrees, however we at Nortek encourage our users to try their best to have the AWAC deployed within 5 degrees of vertical.  This may seem challenging for some types of deployments, so you may want to consider either deploying with a diver who can manually position the frame or use a gimbal which ensures vertical orientation.  

As a side note, an AWAC deployed on a subsurface buoy has demonstrated to maintain vertical orientation within our specified tilt limits and AST performance is typically very similar to a bottom mounted AWAC.

 

Heading

The heading should really be stable for a bottom mounted if it is moving then you will likely see a small degree of error in the directional wave estimates (as well as current measurements).   I suspect that if this happens it is during large wave events and/or shallow water deployments where bottom currents from the waves are strong.  If you are seeing this throughout the deployment then you may want to reconsider how the AWAC is fixed to the bottom (anchored or weighted).  

But back to Juan's original question, if the heading changes during a wave burst, then this will be reflected in the directional estimate.  How much it changes,  how it changes (back and forth or one big change), and when the heading changed in the burst it changes will all effect the directional estimate.  I suspect that oscillating heading will simply smear the directional estimates but the mean may still be "reasonable".  Large jumps and when they occur in the burst are more difficult to say how much they effect the resulting wave estimates.

We have no firm recogmendation for heading limits, but I like your suggestion of keeping it under 5 degrees

One solution to the problem of heading changes is to configure the AWAC wave measurements to collect data in SUV mode.  This will account for changes in the heading for each wave sample

 

 

 Juan, I would be happy to have a look at your data.  You may post it on our ftp site:

ftp://nortek-as.com/pub/incoming

After you have posted the file, please send an email to inquiry@nortek.no to let us know it is there and tell us what you would like like us to have a look at.

 

kind regards,

Torstein

 

Hi!

This is a question about the processing methodology to export the wap file, and error codes.

I have a .wpr file corresponding to a deployment of a 600kHz AWAC. This instrument has been deployed at a location which is approximately 40m deep.

The pitch and roll values are between 1 and 7 degrees in absolute values.

The .wpr has been processed using the PDWaveBase library, with a nProcMethod=MLMST 4 , and nSpectrum=2 (back up spectra for AST, base for other)  (inputs of the PDSetProcPar function).

According to the PDwaveBase documentation, "The nSpectrum parameter specifies which estimate will be used to estimate the non-directional spectrum and associated parameters. For the AWAC with AST it is encouraged to use the “auto” setting of 2. This will always use the AST and default to the pressure or velocity based on the measured depth.”

When launching the PDwaveBase to get the wap file, I find that in the colum corresponding to the Hm0 and Tp there are -99999 values for all the bursts, and for some other parameters, this value appear from time to time.

The error code (last column of the wap file) shows mainly these errors:

1027 = 100 0000 0011 : High AST Data Loss + no pressure + low pressure
1043 = 100 0001 0011 : High AST Data Loss + no pressure + low pressure+ Unreasonable Estimate

High AST Data Loss = This indicates that the too many points were lost during the data clean-up or despiking steps.  The level at which the data loss in the AST time series is considered excessive is 10% of the data.

What can be the cause of this?

40meters deployment is in the range of AWAC measurements, how is it possible to have so much AST data loss and no pressure record??

Shall we process the .wpr using a SUV method? What are the ranges of depth best suitable for each kind of processing methodology?

Find attached to this message a folder which contains the .whd and the .wap file.

Thanks in advance for your help

Camille.