#======================================================================
# L A D C P P R O C . D E F A U L T S
# doc: Fri Sep 17 09:44:21 2010
# dlm: Tue May 20 09:07:05 2014
# (c) 2010 A.M. Thurnherr
# uE-Info: 278 0 NIL 0 0 72 0 2 4 NIL ofnI
#======================================================================
# default parameters for [LADCPproc]
# NOTES:
# - defaults are taken:
# 1) from my current UH processing merge control files
# 2) from the defaults set in Eric's [merge.c]
# - the default version in the ANTS bin dir is always loaded
# - if there is a version in the current processing directory it is loaded
# afterwards
# - per-station parameters can be chosed based on $LADCP_file and $CTD_file
# - for additional notes, see [LADCPproc]
# HISTORY:
# Sep 17, 2010: - created
# Dec 9, 2010: - added doc for ASCII CTD file support
# Jun 15, 2011: - added Svbin_start, Svbin_end
# Jul 7, 2011: - added BT processing parameters
# Jul 12, 2011: - added $PPI_editing_enabled
# Apr 11, 2012: - cosmetics
# May 16, 2012: - BUG: comment var name typo
# Jan 8, 2013: - added CTD_ASCII_header_lines
# Apr 22, 2013: - BUG: BT_begin_search_above value of 300m did not make sense
# Jun 5, 2013: - added $bad_beam
# Sep 6, 2013: - BUG: BT_begin_search_above value of 300m was correct;
# the original bug was in the documentation
# Sep 19, 2013: - added support for $BT_range_method
# Feb 22, 2014: - added $LADCP_max_gap
#----------------------------------------------------------------------
# Data editing
#----------------------------------------------------------------------
# Sometimes, an ADCP beam degrades to the point where its data contaminate
# the 4-beam solutions. Set the following variable to the number of the
# bad beam to discard its data. This variable can also be used for crude
# wake editing in cases where only a single beam is affected, which can
# happen in particular with CTD platforms that tend to "weather vane".
# $bad_beam = 1;
# Some ADCPs report occasional spurious velocity data while on deck.
# The software can detect and remove many of these bad data based on
# long gaps between subsequent valid ensembles. To turn on this heuristic,
# reduce the value of the following varible (time in seconds) to the longest
# LADCP gap in the data set (bottom bottle stop, large instrument tilt, etc).
$LADCP_max_gap = 9999;
#----------------------------------------------------------------------
# ASCII CTD file support
#----------------------------------------------------------------------
# By default, [LADCPproc] reads a standard seabird .cnv file. Alternatively,
# the CTD data can be supplied in a plain ASCII file. To do so, the following
# variables must be defined in the setup file (-s). Each variable defines
# the field (column) number of the corresponding data field. Fields are
# numbered beginning with 1.
# $CTD_ASCII_sampfreq = 1; ## in Hz
# $CTD_ASCII_press_field = 1;
# $CTD_ASCII_temp_field = 2;
# $CTD_ASCII_salin_field = 3;
# $CTD_ASCII_lat_field = 4;
# $CTD_ASCII_lon_field = 5;
# The following variables are optional:
# $CTD_ASCII_header_lines = 4; # if this variable is defined, #-comments are disallowed
# $CTD_ASCII_badval = 999;
#----------------------------------------------------------------------
# Time Lagging
#----------------------------------------------------------------------
# Calculation of the LADCP time series of depth is re-started
# after a gap exceeding $LADCP_max_gap seconds. This is useful
# for dealing with double dips, data from instruments that
# return occasional velocities while on deck, etc. By default,
# all gaps are ignored.
$LADCP_max_gap = 9999;
#----------------------------------------------------------------------
# Backscatter Profile Parameters
#----------------------------------------------------------------------
# The output profile of volume-scattering coefficient is calculated
# from a subset of the bins only. This is based on the observation,
# from on a single P403 profile, that the volume-scattering coefficient
# correction of Deines (IEEE 1999) yield similar results only for
# bins 3-9.
# The seabed-finding routine, on the other hand, uses acoustic
# backscatter from all bins, as it is possible that the seabed is only
# seen in the farthest bins.
$Svbin_start = 3;
$Svbin_end = 9;
#----------------------------------------------------------------------
# Bottom Track Parameters
#----------------------------------------------------------------------
# First bin to consider when looking for seabed to calculated post-
# processed BT. NB: For consistency with EF's code, bin 1 is the
# first bin.
$BT_bin_start = 1;
# Apparently valid BT data collected at distances greater than 300m
# from the seabed are discarded.
$BT_begin_search_above = 300;
# Maximum allowed spread of bin number where max echo is found. Large values
# imply sloping seabed and/or large instrument tilt. The default value of
# 3 is probably too tight. The range spread calculation should be modified
# to take instrument tilt into consideration.
$BT_max_bin_spread = 3;
# For tricky BT cases, the acoustic backscatter from the seabed can be plotted
# and the bottom range set manually. This has been found to be very powerful
# in case of 2011_IWISE yoyo profile 160, where the bottom was around 150m
# from the lower turning points and where the method of Deines [1999] clearly
# does not work.
#$BT_min_depth = xxx;
#$BT_max_depth = xxx;
# Maximum difference between water depth and average distance of echo max.,
# if $BT_min_depth and $BT_max_depth are not set. The stddev of the detected
# water depth is added to this number.
$BT_max_depth_error = 20;
# Maximum allowed difference between reference-layer w and BT w. Note
# that this must be relaxed in regions of significant near-bottom
# vertical velocity, e.g. due to cross-slope currents.
$BT_max_w_difference = 0.03;
# When the BT velocity is taken from the wrong bin, the measurement is not
# taken from the main acoustic beam but, rather, from a side lobe. As a
# result, the BT measurement is biased, causing biased near-bottom LADCP
# velocities when the instrument is moving horizontally.
# Methods:
# 0: Take BT velocity from bin with max(Sv). Not recommended.
# 1: Chose either bin at max(Sv) or one of its neighbors,
# depending which shows the smallest discrepancy between w_BT and w_reflr.
# Causes underestimation of instrument speed in P403 tow-yos.
# 2: Visbeck (2002): use median from 3 bins. Requires $BT_range_Visbeck_center
# to define where the 3 bins should be centered wrt. the max(Sv) bin.
# Visbeck suggests +1. Based on P403 tow-yo 030 it looks more like -2.
$BT_range_method = 1;
#----------------------------------------------------------------------
# Shear Processing Parameters
#----------------------------------------------------------------------
## u_bin0, u_bin1, w_bin0, w_bin1: These set the first and
## last bin indices used when integrating horizontal and
## vertical velocity, respectively. The indices start
## from 1 and are inclusive. These parameters do not
## affect the calculation of shear. They should specify
## a good reference layer (for the calculation of the
## barotropic component) such as bins 1 to 4.
$wbin_start = 1; ## These parameters start from 1, not zero
$wbin_end = 5;
$ubin_start = 1;
$ubin_end = 5;
## sh_bin0, sh_bin1: These are like the above except that
## they affect the shear calculation only. They can
## normally be left at the default range of 1 to 128
## (disabled).
$shbin_start = 1;
$shbin_end = $LADCP{N_BINS};
## w_ref_bin, w_dif: These control one of the editing
## criteria recommended by Fischer and Visbeck, or rather
## a modification of it. All velocity data are rejected
## below the point at which the vertical velocity
## estimate is w_dif larger or smaller than the estimate
## in w_ref_bin. I have not found this criterion
## helpful. 94/12/22: changed, so that only those points
## where w actually deviates from the mean from 0 to w_ref_bin
## are flagged.
$w_ref_bin = 10;
$w_dif = 0.05;
## wake_hd_dif, wake_ang_min, min_wake_w, n_wake_bins: These control
## editing based upon the direction and inclination of the package wake,
## calculated from reference layer U,V, and W. wake_hd_dif
## sets how close to the heading of the wake must be to
## the heading of any beam for interference (in degrees).
## wake_ang_min sets the minimum wake angle from the
## vertical for interference (in degrees). min_wake_w sets
## the minimum package speed required for wake interference
## (although upward speed is measured negative, this parameter
## is input as a positive value). All three criteria
## must be satisfied for the wake flag to be set. n_wake_bins
## determines how many bins from the top are removed from a
## flagged profile, the default is 1 (top bin only).
## Additionally, if the previous ensemble met all
## interference criteria, the present ensemble
## will be flagged even if the criteria are not met.
## Default values are wake_hd_dif=0.0, wake_ang_min=90.0,
## and min_wake_w=0.1, which results in no wake editing.
$wake_hd_dif = 0.0; ## set wake editing defaults so that no wake editing
$wake_ang_min = 90.0; ## occurs
$min_wake_w = 0.1;
$n_wake_bins = 1; ## wake editing default - top bin only
## e_max is the maximum error velocity. An error velocity
## greater than e_max will flag the other velocity
## components. For the BB with 2-ping ensembles, 0.01
## m/s looks about right for this parameter. It
## knocks out the incorrect ambiguity glitches as well
## as some smaller but still significant glitches.
## Note that for the BB, the PG criteria operate on
## percent 4-beam solutions, so PG combined with the
## e_max provides a reliable filter against big
## glitches.
## For the NB with 18-ping ensembles, it is not yet
## clear whether there is any benefit in using e_max
## at all. To be effective, the value will have to be
## small, something like 0.015 or 0.02. Note also
## that for the NB, the pg array holds pg counting
## both 4-beam and 3-beam solutions, so 3-beam
## glitches will slip through the e_max net.
## Default is e_max= 10.0, which effectively disables
## it.
$e_max = 0.1;
## min_correlation (BB only) is the minimum correlation, in
## counts, for each beam in a given bin.
$min_cor = 70;
## Shear editing: This requires 2 passes through the
## database, one using the option
## binned_shear_time_range:
## to set the time range for a pass during which the
## shear statistics will be calculated on a relatively
## coarsely grid by binning rather than interpolation,
## then using the usual option
## time_range:
## for the normal pass in which the results of the
## previous pass can be used to flag bad velocity values
## based on anomalous shears. The first pass can include
## both up and down casts if desired, in which case it
## would be followed by a second pass for the up and
## downcasts separately. The binned shear statistics are
## saved until they are explicitly recalculated with the
## "binned_shear_time_range:" option.
## The shear editing is controlled by these parameters:
## shear_dev_max= x.x, where x.x is a floating point
## number giving a threshold in standard
## deviations. Any shear component deviating from
## the binned mean by more than this times the
## local standard deviation will raise a flag.
## Suitable values for this parameter are probably
## in the range 3-5. Less than 3 is likely to
## start trimming too many valid samples, more than
## 4 or 5 is likely to do nothing at all.
## shear_sum_dev_max= x.x gives a threshold used in
## detecting isolated bad velocity points. If the
## sum of two successive bad shears (based on
## shear_dev_max), divided by the
## standard deviation, is LESS THAN this amount,
## then the common velocity point is considered to
## be an isolated glitch, and it alone is flagged.
## Otherwise, both velocity samples contributing to
## a flagged shear will be flagged. A reasonable
## value for this parameter is probably around 1-2,
## but I don't yet have enough experience to be
## sure.
## Warning: don't forget to reset the first_x=,
## first_y=, and first_z= parameters after pass 1
## (with binned_shear_time_range:). If you don't, and
## the depths are not in the database, then the depths
## will be completely wrong in pass 2 (with
## time_range:). You may also want to use a different
## output file name for pass 1, so that the binned
## shear statistics file will not be overwritten
## during pass 2.
$max_shdev = 3.5; ## to disable, set to nan
$max_shdev_sum = 1.5;
## previous ping bottom bounce interference editing is controlled by:
## clip_margin 0.0 turns off this editing; otherwise, it is the margin in
## meters (on each side) by which the calculated range of the
## interference is expanded before clipping. A reasonable value would
## be something like 32 (2 depth bins).
undef($PPI_editing_enabled); # set to 1 to enable PPI editing
# NOTES:
# - default value (90m) taken from comment in merge control files
# - clipping is disabled if water_depth cannot be determined
$clip_margin = 90; ## default: no previous ping bottom bounce editing
$first_clip_bin = 1; ## default: apply previous ping clipping to all bins */
# tilt editing as in Visbeck's code
$max_tilt = 22; # max allowed angle from vertical
$max_delta_tilt = 4; # max allowed ping-to-ping tilt difference
# On DIMES US2 stations in Drake Passage it was found that the ambiguity velocity had been
# set too low. It appears, however, that aliased velocities are easy to detect because aliasing
# causes a large positive velocity to appear as a large negative one, and vice versa.
# The following defines the maximum allowed discrepancy between the vertical velocity from CTD
# pressure and the LADCP reference-layer w. This parameter should probably be set to something
# similar as the ambiguity velocity.
$w_max_err = 2.5;
1;