----------------------------------------------------- BS3 LR75 from cage above anchor ----------------------------------------------------- -----from aon11_notes.txt: ----- LongRanger ADCP from BS3 : wake up with BBtalk, send break Clock check: PC 10/13/2011 19:23:31 LongRanger clock: 10/13/2011 19:16:53 -----from aon11_notes.txt: ----- => LongRanger clock late by 6 min 39 seconds or 399 seconds set clock adjustment in read_bc3lr.m, p.drift; check out the syntax that applies it in process_adcp_fb.m: looks good. get microcat record: mc2137_v00.mat, read previously by r_mc.m and converted to the correct format of post-cal data using corr_mc.m ==> NOTE <== although we don't have the post-cal for this mc yet, the changes are expected to be small (if any) adjustments to cond, possibly (even smaller) adjustments to temperature. That brings little change for the usage here. Check cond units: load mc file, run ch_condunits.m: still in S/m; mult by 10 to get mS/cm limit visual edits: t_plots_subsets_z3.m - found E=0.05 m/s a sufficient editing criterion overall - interference by the top sphere is visible, though more subtly so than for BS3_200901_LR75; the latter had 5m bins (by mistake), which likely made it more obvious. But also here: wipe out bin 8 after saving example plots (note: it was bin 16 for 2009 with half the bin length) - find again (as in 2009->2010 LR75) mysterious bin 1 interference; it may not be there throughout, and it does appear to have an E signature. Use strong E criterion (0.04 threshold) for this bin only to edit it (program mark_lots_bad.m) - also, although the surface-affected bins are clearly outlined in the amp profile, and the bins with increased amplitude due to surface interference are taken out, the shallowest bins still exhibits outliers that jibe with increased E. This appears to be the case when roll/pitch are large (see line plot in E pcolor timeseries) Again, use stronger E editing for the shallowest bin (E criterion=0.04 threshold, program mark_lots_bad.m) - no manual edits beyond these! ----------------------------------------------------- BS3 sphere WH300 ----------------------------------------------------- workhorse WH300 ser# ending in "27" red case ADCP time 23:45:23 PC time 23:42:32 => ADCP 2 min 51 sec fast offload data file twice at 115KB 2012/03/01 return; use magnetic declination from "magdec" function that is part of UHDAS on linux. Since I am processing in win, I made a time series of magdec at BS3 for a range of years. mkdir c:\projects\healy10_ta\madcp_process\bs3_201011_wh300, copy raw data BS03A000.000 to here. copy programs from C:\matlab\dtorres_may10\scadcp_fb; ----------------from readme.txt------------------------------------ These programs were used by fb to process the "Claire" mooring off AK in 2010. They are half calling routines for or special versions of Dan Torres' SADCP programs, half further editing etc. I use them here as starting repository for further moored ADCP processing. -----------------end readme.txt------------------------------------ make read_bc3wh.m, first with no editing (PG 0, E 100) to look for overall health, first good prof etc. p.magFLAG=10 to select my special BS3 magdev routine. run that to make nopp11_bs3a_ue_all_raw.mat check for good profiling range: pcolor(squeeze(d.vel(:, :, 1))); shading flat; colorbar looks like first good profile is prf 139; there are none earlier; must have been taken out by instrument editing. Select next one as first good, to be sure: d.vl(140, 11:16)= 2010 9 16 22 0 0 prof 9496 seems to be the last good. d.vl(9496, 11:16)= 2011 10 11 18 0 0 re-rum read_bc3wh with: - editing of PG=30, E=0.05; - clock correction +171 seconds (code says: "subtract to correct clock") decide this is so small that is does not warrent messing up the time variable from integer hours - start/end dates as above will use microcat, even though it has not been returned from SeaBird with postcal. The cal changes tend to be small adjustments to S, which won't matter here. Get the correct microcat file: what format should it be in? It (mc2134_v00.mat) was already on the AON web site, having been made by r_mc.m; however, the structure "mc" only included "dday" (decimal day). I added "yearday" with yearday = dday + 1 Then saved back to c:\projects\aon11\recover\bs3\sphere_mc\mc2134_v00.mat From there, also copy it to c:\projects\healy10_ta\madcp_process\bs3_201011_wh300 ==> issue: found that mc.cond was in S/m, not mS/cm!! factor 10 different! made routine ch_condunits.m to check for conductivity units, and multiply by 10 to convert to mS/cm if needed. continue: run process_adcp_fb.m run t_plot_subsets_z2 -many bins shoot into the air. Introduce bin range limit to process_adcp_fb.m to get better plots - push first good profile back by one to eliminate traces of instrument descent: start now at d.vl(141, 11:16)= 2010 9 16 23 0 0 {note: its d.vl(141.. for the output of the initial RDI->mat format conversion run by read_bc3wh.m, nopp11_bs3a_all_raw.mat; 9501 profiles then its d.vl(1, :) in the output file of process_adcp_fb.m: aon11_bs3a_pro1.mat; 9356 profiles. } manually edit: interrupt (^c; sorry) t_plot_subsets_z2.m; identify via polygon the offending bins with t_polyedit7.m, which saves them as ascii time series in add_flag_fb.asc. It is then applied to the data with app_manfbedits2.m, which saves its output as [input]_edt, so here aon11_bs3a_pro1_edt.mat. Specifically: I take out some questionable surface bins where the shallowest-good bin varied a lot. Possible causes include changing vertical range due to blow-over (a fraction of the ensemble average has surface interference) and ice (water column height changes). once we have all edits applied, rerun t_plot_subsets_z2 to generate time series plots for web. finished by writing the final .mat files, using convert_adcp_pro_fb.m and savepro_adcp.m to convert aon11_bs3a_pro1_edt.mat to aon11_bs3a_Final1.mat plus several additional .mat files, most of which are engineering documentation (e.g., correlation). Others, such as bt.mat for bottom tracking, could be more interesting. On bt.mat: The following structure definition was taken from Dan's Torres' general documentation for d.bt (where the structure "d" is part of the initial .mat files); His notes, which describe the format of other intermediate .mat files as well, are accessable on our site "aon.whoi.edu" via the link "Processing Info", located below the most recent cruise. bt(:,1:4) - % Beam 1-4 BT range bt(:,5) - % East component of bottom velocity bt(:,6) - % North component of bottom velocity bt(:,7) - % Upward component of bottom velocity bt(:,8) - % Error component of bottom velocity bt(:,9:12) - % Beam 1-4 BT correlation bt(:,13:16) - % Beam 1-4 BT correlation from here: enter to web