Memory and Power Use Question

Dear Forum,

I'd appreciate your comments on the following deployment profile.

We plan to deploy an Aquadopp Profiler in an estuary channel, with ssc as high as 2500-3000 mg/l during catchment flood events. The tide range is up to 3m and water depth 0.4m at low tide. We have an instrumented mast in the channel with 3 OBS (sample at 2Hz, 600 samples every 10 minutes). Because of external power and comms we have no major sampling limitations.


The propsed set-up for the Profiler is:

>Power Level: high
>Interval between bursts: 10 minutes
>burst length: 5 or 7.5 minutes
>blanking distance: 0.2m
>cell size 0.1m x 30 cells

Can someone confirm max sampling rate for profiler is 1 Hz ?

>(2) wave bursts
>
>Because we can, we thought about doing the following also:
>
>cell size 0.1m
># of samples 2048
>sample rate: 2 Hz
>Interval 1800s

The software calculates that for above setup 2.3Mb of data
collected per day and 10% of 50 Wh. I note that if I reduce # of 0.1m cells from 30 to 1, the memory does not change and power drops from 10% to 1%.  Why would that be ?

Your comments appreciated

best regards

Andrew Swales

Dear Andrew,

The Aquadopp Profiler can either measure profiles or it can take a burst of single cell measurements with a fixed blanking distance of 0.20 m. The single cell burst measurement was developed for PUV wave measurements, hence the term Wave bursts selection in the Advanced setup dialog.

The timing for the profile/wave measurements is given by:
Profile interval - The time between consecutive profiles.
Average interval - The time period the Aquadopp Profiler will average internally before the output of one profile.
Wave burst interval - The time between each set of wave burst measurements is done.

Example:
The Aquadopp Profiler is started at 1:00:00 p.m. with profile interval of 600 seconds, average interval of 50 seconds, wave burst interval of 1800 seconds doing 512 burst measurements at 2 Hz. The data will then be output in the following manner:
1:00:50 pm Profile #1 output
1:10:50 pm Profile #2
1:20:50 pm Profile #3
1:30:50 pm Profile #4
1:30:51 pm Wave burst header #1 output
1:30:52 pm Start of single cell output at 2 Hz
|
1:35:08 pm End of single cell output at 2 Hz
1:40:50 pm Profile #5
1:50:50 pm Profile #6
2:00:50 pm Profile #7
2:00:51 pm Wave burst header #2 output
2:00:52 pm Start of single cell output at 2 Hz
|
2:05:08 pm End of single cell output at 2 Hz
2:10:50 pm Profile #8
etc.

Now for the answer to your questions.

1) The maximum sampling rate for profiles is 1 Hz. To achieve this sampling rate you will need to measure profiles continuously. If you enable wave burst measurement, the minimum profile interval will be 60 seconds.

2) The reason why the memory does not change is simply a rounding effect. With the setup you specified the wave data will take up nearly all of the 2.3MB of data. To get around the rounding effect you can choose 10 or even 100 days, you will then see that the required memory does decrease when you reduce the number of cells.

As for the actual deployment you say that you have external power and comms. I assume you can then handle large sets of data as well, so I would suggest the following setup.

Profile interval: 1 second
Average interval: 1 second
Power level: High
Blanking distance: 0.05 m
Cell size: 0.10 m
Number of cells: 30
Wave bursts: Off

The data can then be stored in the PC. If the deployment duration is so long that it will fill up the recorder in the Aquadopp Profiler it is important to use the Start Data Collection selection, if not the instrument will stop taking data when the recorder is full. To ease the data processing the Aquadopp Profiler and the OBS sensors should be started simultaneously. It is probably also a good idea to use the Automatic Filename feature in the AquaPro software. You can choose how many minutes of data you store in each file. For example one file every hour.

If you want 2 Hz sampling rate there are a couple of things to be aware of. You must then sample in wave mode, so you will only get one cell for each measurement. In addition, that cell will have a fixed blanking of 0.20 m. It is possible to configure the instrument to do a wave burst measurement every 10 minutes, for example 1024 samples at 2 Hz. This would match the sampling of the OBS sensors.

- Sven

Dear Sven,

thanks for the reply to my questions. This has generated more ! Can you please tell me:

- for wave bursts, which cell is measured (I assume it is the first cell beyond blanking distance) or can it be selected ?

- for any given cell size, what is the sampling volume ?

- stratification. Given that I have conductivity and temp data at 3 elevations, what are the effects of vertical density variations, and how best to correct for them.

- the Aquadopp has a 5 Mb memory. We will obviously need to download (and clear?) the memory regularly. You comment on the need to use "Start Data Collection" otherwise the Aquadopp will stop data collection when the memory (recorder?) is full. Does your approach negate the need to clear the memory manually each time ?

thanks in advance

Andrew Swales