Hi, I have looked at the 165E data from the narrowband 150kHz adcps. Orienting the two deployments wa3 and wa4 with respect to one another is a challenging problem. I have created some comparison plots and discussions as listed below. prof_wa4_dec.gif: There is considerable temporal variability in the target strength so I would not expect the difference between consecutive 4 day mean profiles of target strength to be linear or constant. In figure prof_wa4_dec.gif I show profiles of the last four days of deployment wa3 and the first four days of deployment wa4, as you have done. In the second panel, I show profiles from wa4 of the 4 day mean from 10 December 2002 to 14 December 2002 and the 4 day mean from 14 December 2002 to 18 December 2002. Even for the same deployment, the difference in 4 day consecutive profiles is not constant. I can tell you almost for certain that there is not a depth mismatch on the order of 10-15m. When I map the data to depths, I compare the location of the surface determined from the maximum target strength with the depth of the transducer determined by two pressure sensors. The maximum difference I have had is 6.5m in the mean and I go back and adjust the mapping of the data to eliminate this difference. From all my intercomparisons, I expect my depths to be mapped to an accuracy of better than 5m. I do occasionally put in the wrong number when I run my programs though, so it is good to check. prof_tar_4pyr.gif: Looking at the previous plot, it did appear that the 50m of depth closest to the transducer appeared to have unreasonably low target strengths compared to the shallower values. To look at target strength variability and at values near the transducer, I made prof_tar_4pyr.gif. I divided wa3 into 4 time periods and computed a mean target strength profile for each time period (solid lines). I did the same for wa4 (dashed lines). As I look at this plot, I see that there is variability within a deployment, that the wa4 profiles are perhaps 6-8 dB too high compared with wa3, and that the 50m closest to the transducer have target strengths that are too low. It appears to me that the equations do not appropriately convert echo amplitude to target strength within about 50m of the transducer head. Target strength is a challenging measurement. As I stated before distributing these data, our adcps were not calibrated to allow computation of absolute target strength from the echo amplitude, nor do I adjust for the temperature at the transducer. You have asked if there is any way to reference the deployments to each other. My first reaction is no, however I will share some of my observations with you that may allow you to implement a referencing and perhaps understand the expected magnitude of the referencing errors. To begin with, the adcp measures the magnitude of the return signal for each of the 4 transducers. The measurements are time gated to obtain 44 measurements (bins) at increasingly shallower depths from the upward looking transducer. The echo amplitude measurement is collected at the end of the listening period for the velocity measurement, so I adjust the depth of the echo amplitude measurement versus the depth of the velocity measurement accordingly. As stated in my tech memo, I then use a historical sound velocity profile for each site and map the 44 bins to actual depths. As stated above, I also use the reflection of the surface to determine the depth of the transducer, and compare the result with the pressure gages attached to the adcp. This allows me to verify the accuracy of my depth mapping. Since I want to resolve the surface, I actually record data for time intervals that would represent distances that would be 50 to 100m above the surface. If the measurements were far enough "above" the surface, the amplitude signal would flatten out (become nearly constant) and is representative of the back ground noise level of the instrument. prof_amp_tar_wa5.gif: In prof_amp_tar_wa5.gif, the left set of profiles are the mean amplitudes for each of the four beams for the WA5 deployment. Note that as the profile approaches "-115m" depth (above the surface), that the amplitude signal becomes nearly vertical, which indicates an approximate background noise level for each of the four beams. When I developed my processing software, I had noted that the target strengths of the four beams would align better if I removed this approximate background noise level for each beam. The target strengths to the right on this plot have NOT had the back ground noise level removed from the echo amplitudes before conversion. For most of my earlier data I removed the background noise level for the target strength computation. However, the practice is not great. Often I would have deployments where the transducer was deployed so deep that I only got about 40 to 50m above the surface, and did not have a good representation of the background noise for each beam. Even for this deployment, where the amplitude is nearly constant from about -80 to -115, the method of forcing the noise levels to align (ie the vertical amplitude line from -80m to -115m), would not exactly align the amplitude in the water column. However it does improve the alignment. An initial difference range of 6dB between the target strength profiles in the water column would be reduced to about 2dB if the amplitude profiles were all referenced to the approximate background noise level. I will demonstrate this in plots that I discuss later. For this plot, I recomputed the target strength for the wa5 deployment such that I did NOT remove the background noise and also mapped the data "above" the surface as shown on the right side of this plot. The apparent increase in target strength from -40m to -120m is not real, but rather the result of taking a constant echo amplitude value and adjusting it for supposed sound absorption and beam spreading. prof_amp_tar_wa4.gif: Similar mean deployment amplitudes and target strengths (without removing background noise-no referencing) for WA4. prof_amp_tar_wa3.gif: Similar mean deployment amplitudes and target strengths (without removing background noise-no referencing) for WA3. prof_amp_tar_all.gif Similar mean deployment amplitudes and target strengths (without removing background noise-no referencing) for WA3,WA4 and WA5. In the earlier data that I sent you, I had attempted to reference the amplitudes to the approximate background noise level. However, since the reference value approximated from the portion of the profile above the surface was not very precise, nor work all that well, I was not very particular about what I used for referencing. For the explanations below, I have recomputed the target strength using the mean echo amplitude for each beam for each deployment at 80m above the surface as the reference instrument noise. Since my noise estimate (referencing echo amplitude) can actually be greater than the measured echo amplitude, I reduced the referencing echo amplitude by 15 dB for all profiles. This increases the target strength at all depths by 15dB, but is a constant offset. In prof_ref80m_wa3.gif, prof_ref80m_wa4.gif and prof_ref80m_wa5.gif I have plotted the new mean deployment profiles of referenced target strength for each beam. One would expect the 4 beams to measure the same target strength from the same instrument. Note that the spread in subsurface target strength values for the three deployments still have a spread of 1 to 2 dB, but are greatly improved over the unreferenced target strengths shown in earlier plots. In prof_ref80m_all.gif I have overplotted the referenced mean target strength for each beam for the three deployments on the same plot. At about 100m depth, the spread of target strengths is about 6-7 dB, versus the unreferenced profiles with a spread of target strength values at 100m of 11-12 dB. Further it is obvious from prof_ref80m_wa4.gif that beam 2 in deployment wa4 is too large by 1.5-2 dB and if this could be corrected, the spread of referenced target strengths for all three deployments would be reduced. Another note: above the surface the target strengths for deployment wa4 are smaller than for wa3 and wa5. This is because although the echo amplitude signal above the surface has been set equal for all deployments and beams, that the depth of the transducer for wa4 is shallower than the other two deployments. Therefore the adjustment for beam spreading and absorbtion, which increases the measurement with distance from the transducer when converting amplitude to target strength, is less for the shallower deployment. Soooo, I am not an electrical engineer and I do not know the best way to determine the background noise level from the data. Most studies using target strengths from adcps have had the instruments calibrated for target strength measurements prior to deployment. I did look at the consistancy of each beam noise level at 80m above the surface. In wa5_sclr.gif beams 1,2 and 3 seem faily consistant with respect to one another, but beam 4 seems to drift about 1 dB from the beginning to the end of the deployment. In wa4_sclr.gif the beam measurements are fairly consistent with a bit more fluctuations in beam 4. In wa3_sclr.gif beams 1,2,and 3 seem consistant with respect to one another, but beam 4 seems to have problems. Therefore, when I computed a referencing noise level for beam 4, I used values from 20 June to 15 August only. Thus, there appear to be factors that increase the noise level measured above the surface that do not correlate with increase in subsurface amplitude. I have not yet looked at the magnitude of the difference in temperature at the transducer heads nor what effect adjusting for transducer temperature would have on the magnitude of the target strength. I will try to do a simple estimate of the magnitude of the change and let you know. The effect would be constant with depth, so would only be a factor for trying to reference the deployments to one another. Note that most of my deployments have the transducer deeper than these three deployments, and the information above the surface that I have illustrated in these plots would not be available. Well I hope this helps. Let me know how you want to proceed. Regards, Tricia