Processing of sontek deployment pm438 at 0, 140w On the unix machine I have defined: setenv bcprog ~OCRD/plimpton/ndbc/prog setenv bcs ~OCRD/plimpton/ndbc/sontek setenv bcsraw ~OCRD/plimpton/ndbc/sontek/raw setenv bcsplots ~OCRD/plimpton/ndbc/sontek/plots setenv bcsproc ~OCRD/plimpton/ndbc/sontek/proc Also for adcp: setenv bcadcp ~OCRD/plimpton/ndbc/adcp setenv bcraw ~OCRD/plimpton/ndbc/adcp/raw setenv bcplots ~OCRD/plimpton/ndbc/adcp/plots setenv bcproc ~OCRD/plimpton/ndbc/adcp/proc setenv bcsen ~OCRD/plimpton/ndbc/adcp/sensors from deployment log: anchor drop at 3 26 23 5 04 start data at least 2 hours after drop start at 5 26 23 5 04 from recovery log: release 16 53 12 9 04 stop data at 16 50 12 9 04 first convert *.dat files to ascii files add 90 sec to data since data is recorded for 3 minutes centered on the even 10minute time intervals. Rotate 9 degrees for 140W. Correspondence of depth with sontek serial number from deployment and recovery logs: $bcprog/con_sontek5 < $bcsproc/con_sontek.in note the start and end time of files: $bcsraw/pm438egr.005 1710 0 15 5 2004 1940 0 13 9 2004 $bcsraw/pm438egr.010 1550 0 14 5 2004 1950 0 13 9 2004 $bcsraw/pm438egr.025 1610 0 14 5 2004 2020 0 13 9 2004 $bcsraw/pm438egr.045 21 0 0 14 5 2004 2030 0 13 9 2004 $bcsraw/pm438egr.080 16 0 0 14 5 2004 20 0 0 13 9 2004 $bcsraw/pm438egr.120 0 0 0 28 4 2004 2010 0 13 9 2004 Next plot the 10 minute data - ! I will provide plot software if you get DISPLAA $bcsplots/plsontek_5m.indat pm438_005m_1.gif, pm438_005m_2.gif, pm438_005m_3.gif, pm438_005m_4.gif $bcsplots/plsontek_10m.indat pm438_010m_1.gif, pm438_010m_2.gif, pm438_010m_3.gif, pm438_010m_4.gif $bcsplots/plsontek_25m.indat pm438_025m_1.gif, pm438_025m_2.gif, pm438_025m_3.gif, pm438_025m_4.gif $bcsplots/plsontek_45m.indat pm438_045m_1.gif, pm438_045m_2.gif, pm438_045m_3.gif, pm438_045m_4.gif $bcsplots/plsontek_80m.indat pm438_080m_1.gif, pm438_080m_2.gif, pm438_080m_3.gif, pm438_080m_4.gif $bcsplots/plsontek_120m.indat pm438_120m_1.gif, pm438_120m_2.gif, pm438_120m_3.gif, pm438_120m_4.gif Rotate vertical velocity into the speed and recompute u and v from new speed and old direction. New file keeps old vertical velocity. $bcprog/rot_vert_to_speed < $bcsproc/rot_vert_to_speed.in makes in $bcsraw pm438egr_pr.005, pm438egr_pr.010, pm438egr_pr.025, pm438egr_pr.045, pm438egr_pr.080, pm438egr_pr.120 Convert the data to ram format so rest of processing per module proceedures: $bcprog/convert_sontek_to_ram_2 < $bcsproc/convert_sontek_to_ram.in Make flag data with no edits for comparing with adcp data: /usr/local/nxram/bin/flag_data $bcsraw/curr438a.ram $bcsraw/flag438a_noedits.txt makes curr438a.flg copy curr438a.flg to curr438a_noedits.flg ! so there is a copy for you to see /usr/local/nxram/bin/davg $bcsraw/curr438a.flg makes curr438a.davg copy curr438a.davg to curr438a_noedits.davg ! so there is a copy for you to see Make plots of adcp/sontek comparisons. ! I will provide plot software if you get DISPLAA $bcsplots/cmp_sontek_adcp.indat $bcsplots/reg_sontek_adcp.indat makes: $bcsplots: cmp_005m.gif cmp_025m.gif cmp_080m.gif cmp_010m.gif cmp_045m.gif cmp_120m.gif reg_005m_spdir.gif reg_025m_spdir.gif reg_080m_spdir.gif reg_010m_spdir.gif reg_045m_spdir.gif reg_120m_spdir.gif reg_005m_uv.gif reg_025m_uv.gif reg_080m_uv.gif reg_010m_uv.gif reg_045m_uv.gif reg_120m_uv.gif Note that the instruments do not measure direction well at low speeds. For these regressions exclude directions where the speed for either instrument was less than 5 cm/s. This sometimes results in n: for direction less than n: for u, v, and speed. Next begin to edit sontek data: at 80m there are two intensity drops in beam 3. The second intensity drop coincides with an increase in the ADCP - Sontek velocity difference. Set sontek data bad beginning 2004236200000 There are spikes in the 120m data from 2004246220000 to 2004246222000 for beam 1 intensity, u deviation, u, sp, vert vel, and compass direction. (pm438_120m_1.gif, pm438_120m_2.gif, pm438_120m_3.gif, pm438_120m_4.gif). Set the data bad for this time period. Note: On 245 2004 (1Sep04), the 120m D204 sontek appeared to slip on the wire and rotate ~100 degrees. (pm438_120m_3.gif) This did not appear to cause error in the data. Likewise on 200 2004 (18Jul04), the 80m D129 sontek appeared to slip ~60 degrees on the wire with no error in data. (pm438_080m_3.gif) No apparent fishing pulls from pressure data or Sonja's editting of temperature. The Q4 values set for the temperature at 500m does not apply to shallower sonteks. /usr/local/nxram/bin/flag_data $bcsraw/curr438a.ram $bcsraw/flag438a_testedits.txt makes curr438a.flg copy curr438a.flg to curr438a_testedits.flg ! so there is a copy for you to see /usr/local/nxram/bin/davg $bcsraw/curr438a.flg makes curr438a.davg copy curr438a.davg to curr438a_testedits.davg ! so there is a copy for you to see remake adcp/sontek comparison plots for 80m and 120m ! I will provide plot software if you get DISPLAA $bcsplots/cmp_sontek_adcp.indat $bcsplots/reg_sontek_adcp.indat makes: $bcsplots: cmp_080m_testedits.gif cmp_120m_testedits.gif reg_080m_spdir_testedits.gif reg_120m_spdir_testedits.gif reg_080m_uv_testedits.gif reg_120m_uv_testedits.gif Note that the apparent lack of agreement in the September 2004 meridional velocity at 120m may well be due to ocean variability and there is no reason to set the sontek data bad. Note also that the sontek tends to underestimate the speed at 120m where scatterers are less abundant. This is typical of the 3mHz sonteks. Sonteks when compared with adcp data (25m, 45m, 80m (editted), 120m (editted)) all have mean differences less than 5cm/s and 5degrees and rms of differences less than 10 cm/s and 10 degrees, so set the quality of the data at these depths to Q1. (I am still determining the exact cutoffs for quality 1 data, but this is a good guess. Leave the 10m and 25m data that were compared with extrapolated adcp data as Q2 since differences probably due to adcp extrapolation rather than problems with sonteks. The short distance needed to extrapolate for the 25m extrapolated adcp data allowed me to be able to assign the 25m sontek data as Q1. At this point I do not generally adjust the sonteks according to the sontek compass calibration. However, I keep track of differences with adcp and sontek calibrations to verify quality of future use of sontek compass calibration, and may reprocess these data after I have confidence in the sontek compass calibrations. direction comparison: CA09 and PM438 1st column is the mean direction difference between the adcp with compass calibration applied and the original sontek direction (see $bcplots/pltreg_all_pm438a.gif and pltreg_all_pm438b.gif), then the pre and post calibrations for the sontek compass are listed, then the difference between the adcp and the sontek with the mean pre and post sontek calibration removed. Don't consider the 5m and 10m data as adcp is extrapolated. adcp-sontek sontek calib adcp-calibrated sontek pre post 5m D138 -3.92 0.2 0.0 extrap data 10m D69 -1.14 -0.3 -1.0 extrap data 25m D101 1.62 -0.1 0.2 1.67 45m D117 3.47 -1.0 -2.0 1.97 high residual roll error 80m D129 .78 -0.7 -0.2 .33 120m D204 1.24 -0.3 -1.7 .24 Use of the sontek compass calibration would appear to improve the adcp/sontek direction comparison for this deployment. Do final edits and set qualities: /usr/local/nxram/bin/flag_data $bcsraw/curr438a.ram $bcsraw/flag438a.txt makes curr438a.flg /usr/local/nxram/bin/davg $bcsraw/curr438a.flg makes curr438a.davg Next make summary file to be added to ram summary file: $bcprog/make_ram_sum_list < $bcsproc/make_ram_sum_list.in makes $bcsraw/pm438a_sontek.sum copy into ram summary file: pm438a.sum