mvco_optics

Routines, notes, and results for review paper on particle optics.

Objective

The immediate goal is to extract some example vertical profiles from the MVCO experimental data. In this case, the raw data consists of bin-averaged times, elevations, and instrument responses for instruments on the profiling arm. Ancillary data of time series of flow characteristics are in other data streams.

Data files

This is a list of key, most recent datafiles from the profiler. These reside in the data_proc directory, and will be copied into this repo if they are needed for the processing scripts stored her.

• advpr_mvco11_05-Jan-2012.mat - Processed ADV pair data, with velocities, wave parameters, and stresses
• puv_proc_green_30-Dec-2011.mat - Current, wave, and Madsen 94 model results from green ADV
• _pfa.mat files are time series of profile fit parameters (not saved in this repo)
• suspsed_ba20_20rstrim_crs_cen.mat - Profile data structure for all instruments after bin averaging. This is the starting dataset for this repo, and has two structures (ba for bin-averaged, and va for variance) with fields like this:

ba = 
  struct with fields of arrays with dimensions [vertical bins x time]:

          ADVzc: [12×2441 double]
          ADVtc: [12×2441 double]
          OBSzc: [12×2441 double]
          OBStc: [12×2441 double]
        transzc: [12×2441 double]
        transtc: [12×2441 double]
         ADVagc: [12×2441 double]
            OBS: [12×2441 double]
       tranattn: [12×2441 double]
              u: [12×2441 double]
              v: [12×2441 double]
          ADVdn: [12×2441 double]
          OBSdn: [12×2441 double]
         trandn: [12×2441 double]
           ADVz: [12×2441 double]
           OBSz: [12×2441 double]
          tranz: [12×2441 double]
          ADVbe: [1×13 double]
      ADVfields: {13×1 cell}
      LISSTattn: [12×2442 double]
      LISSTD16v: [12×2442 double]
      LISSTD50v: [12×2442 double]
      LISSTD84v: [12×2442 double]
       LISSTsvv: [12×2442 double]
     LISSTSzmnv: [12×2442 double]
      LISSTvtot: [12×2442 double]
      LISSTD16a: [12×2442 double]
      LISSTD50a: [12×2442 double]
      LISSTD84a: [12×2442 double]
       LISSTsva: [12×2442 double]
     LISSTSzmna: [12×2442 double]
      LISSTatot: [12×2442 double]
    LISSTfinesv: [12×2442 double]
    LISSTmicrov: [12×2442 double]
    LISSTmacrov: [12×2442 double]
    LISSTfinesa: [12×2442 double]
    LISSTmicroa: [12×2442 double]
    LISSTmacroa: [12×2442 double]
     LISSTvconc: [12×2442×32 double]
     LISSTaconc: [12×2442×32 double]
        LISSTsz: [32×1 double]
        LISSTdn: [12×2442 double]
         LISSTz: [12×2442 double]
        LISSTzc: [12×2442 double]
        LISSTtc: [12×2442 double]
        LISSTbe: [1×13 double]
    LISSTfields: {28×1 cell}
    
    ...and many more

Processing scripts

Important processing scripts that have not been copied into this repo are:

• bin_ave_var_suspsed_insts.m - Bin averages raw data (as posted in data release) to make profile .mat files (e.g., suspsed_ba_20_20rstrim_crs_cen.mat). The number (e.g., 20) indicates the minutes for bin-average...20 = 1 profile, 40 = 2 profiles, 80 = 3 profiles. The var here indicates that variance results are also produced. cen in the output file name means that bin centers are included for some (but not all) of the instruments. (Why not all? Might need to look back at this aspect.) The product of this script does not include any of the finest LISST sizes...the first 13 bins are all zeros, as is the input data...so it is not a problem with the script, and may be a result of the LISST data processing choices.

This script was slightly modified to process a different LISST output file, producing a new bin-averaged datafile called suspsed_ba_20rstrim_crs_cen_june_25_2018 which does have information in the first LISST bins. The input file from the STG server was ..\Stg-proc\data\MVCO_11\910moving_arm\9109armlisst\Processing\Spherical\MVCO_lstats.mat, dated 8/15/2012. This looks like the most-recently produced LISST output file, and (I think) it was processed using LISST SOP 5.

• plot_pfa_ts.m - prepares data for fdyn.m

  • Loads the time series data ustar_av.mat created by plot_ustar.m
  • Loads LISST size bins from lisst_av
  • Loads a bunch of *_pfa.mat files
  • Interpolates the results onto the u* time base, then onto two-hour time base

• fdyn.m - various time series plots of floc parameters

Scripts included in this repo are:

• ws_fit_mo.m - Fits Rouse profiles to profile data (e.g., suspsed_ba20_20rstrim_crs_cen.mat) using both linear and non-linear fitting routines. The mo suffix indicates this file has been derived from un-prefixed version in data_proc directory. This was a preliminary start on plotting, and can probably be removed from the repo.

• p_fit_times_mo.m - Derivative of ws_fit_mo.m restructured to produce profiles of many instruments for one time period. This is still a work in progress. It calls fitting routines pfit.m and pfit_nlp.m, which require lsfit.m, and the plots are made with a `plot_snippet.m'.

• p_fit_times_mo_new.m (and plot_snippet_new.m) - These are versions of the the script described above revised by Emmanuel to a) plot a preferred selection of profiles and b) change the temporal interpolation of ustar.

Preliminary Results

This is an overview time series of the OASIS measurements at the USGS profiling tripod.

We have identified several periods with distinctive conditions. A "typical" profile plot for each one appears below.