Experimental Galveston Bay/Houston Ship Channel Nowcasting/Forecasting System

TRANSECT PROGRAM Processing Instructions:

TRANSECT version 4.0 was obtained via email from Paul Devine, RD Instruments. In the email file save the attachments and run .exe file in directory in which you wish to run TRANSECT. TRANSECT should be run in DOS. A WINDOWS-NT version is under development.

To run TRANSECT:

1. transect.exe /b (Note /b fixes the screen problems)
2. Select 5 Playback from main menu
3. F3: Load in tamu.cfg (configuration file)
4. F3: Load in series name of raw data fileset (HSC001R)
5. F4: Pause
6. ALT-W: rewind
7. ALT-A: Select ASCII out filename and filesize
8. F4: Resume
(Repeat 4-8 for each raw data fileset (HSC002R-HSC006R)
9. Select Exit from main menu

The above steps are executed to get ASCII out files with position data included from navigation files. NAVMERGE is no longer necessary to run. The default file size is 300 kbytes, which may result in a large number of files. A file size of 1 megabyte is recommended.

There are several things one can do while producing the ASCII out files. Use ALT-L to show side lobes. See pages 8-2 and 8-3 of TRANSECT Manual.

PROGRAM ADCP_SCAN

ASCII output files, from the RD Instruments TRANSECT Program, were then scanned and edited such that for each CTD cast a separate file was created. With the file, the first point was selected nearest the west channel marker and the last point nearest the east channel marker. All points considered were within +/- 6 minutes of the CTD cast time. If the nearest points to both the west and east channel markers were greater than 100m away or if the start and stop times of the start and end points did not bracket the CTD cast time, the file was discarded from further analysis. Four of the 55 files were discarded based on these criteria.

For files selected for subsequent analysis points were discarded that met any of the following criteria:

1) average of the four beam depths were less than 10m
2) the bearing of the point from the start point was greater than or less than the transect bearing by more than 20 degrees
3) the distance of the point from the start point was greater than 300 m
4) the distance of the point from the last selected point was less than 10m

Based on these criteria, the points for each CTD cast were along transects and within the navigation channel of width order 130m. Typical weak flood and strong ebb CTD profiles exhibit substantial stratification order 8 PSU and diurnal heating/cooling order +/- 0.5 degrees C.

PROGRAM ADCP_TRANSPORT

Adcp_transport.f was written by Phil Richardson and Dick Schmalz to estimate the transport through a given transect based upon towed ADCP current velocity data. The ADCP current velocity data is generated by the RDI program Transect. The ascii output from Transect is in a standard format which is documented.

For each record, adcp_transport first reads time information including year, month, day, hour, minute, and second. The program then reads the four beam depths and calculates an average beam depth.

Adcp_transport then reads latitude and longitude, and records the latitude and longitude for the first record of the transect. For each record after the first, the distance between records and the distance from the first record is calculated. Also, the ship's bearing is determined. The current velocity values of speed, direction, U component, V component, and the vertical component are read by bin. The vertical component of velocity does not enter into the transport calculation. Each bin represents a particular depth range. Also read is the percentage of good pings. Any bin with a percentage less than 100 is considered to be bad data.

A bin of bad data can be "fixed" with a call to subroutine binfix. Bad values are replaced by taking the U and V component values at the next higher and lower bins and determining the average value. For a situation where two or more consecutive bins are bad, the routine will return an interpolated value for U and V. Adcp_transport has several criteria for rejecting a record. For a record to be considered good, there must be at least six bins of data. Also, the bottom three bins of data must be good.

In order to calculate transport through a transect, it is necessary to determine the direction of the transect from the first record to the last record. The direction normal to the transect is this direction plus isign times ninety degrees. Orientation is taken such that flood direction is positive. In this case, isign = -1. Subroutine uv2norm is given the U and V components of velocity, the direction normal to the transect, then returns a value for the current speed normal to the transect.

Each transect is divided into segments. A segment is considered to be the distance between two consecutive records. To calculate the estimated transport through one segment for one bin of data, the normal current speed is multiplied by the bin depth (deltabin_depth = 1.0 meters) and one half the distance of the segment. For each given segment, the current velocity values from the first record are used to calculate the transport through the first half of the segment. The velocity values from the second record are used to calculate the transport through the second half of the segment. The transport through the first half is added to the transport through the second half. The same procedure is carried out for all bins from surface to bottom. We calculate transports for half segments because records can change in depth from one to the next. For example, record 1 may have eight bins of data, while record 2 may increase to nine bins of data. The mid level transport is determined by summing all bin transports.

A separate calculation is done for the surface layer transport. In the case of the 1999 Houston Ship Channel Survey data, bin 1 is centered at 2.95 meters. The top edge of bin one is therefore 2.45 meters deep. The surface layer transport is obtained by multiplying the normal current speed from bin 1 with the depth value of 2.45 and one half the distance of the segment. A separate calculation is also done to obtain the bottom layer transport. In this case, deltabin_depth, the bin thickness, is the difference between the average beam depth and the low end of the bottom bin depth. The bottom layer transport is obtained by multiplying the normal current speed of the bottom bin with deltabin_depth, the bottom thickness, and one half the segment distance. The total transport through one segment is the sum of the surface layer, the bottom layer, and the mid level transports. After the transport has been determined for one segment, the transport is summed over all segments of the transect.

The normal current speeds are used not only in the transport calculation, but as input for the ADCP current speed contour plots. Adcp_transport generates an output file we call 'adcp.contour'. This file includes the normal speeds by bin and depth, along with segment distances, transect distance, and record depths. Plot_adcptrans.pro is a program written by Phil Richardson, in the IDL programming language, which generates the ADCP current speed contour plots. Typical weak flood and strong ebb contours exhibit significant shear and a strong core region, respectively.

Adcp_transport contains a plot option. If desired, the program will plot record locations. The plot portion of the program makes use of NCAR graphics. In the case of the 1999 survey data, the transects are quite short. So a number of X's appear to be almost on top of each other.

Adcp_transport has eight debug options. They are as follows :
idebug = 1, write all ADCP current data, as it is read, to the output file
idebug = 2, bad data
idebug = 3, U and V components by bin
idebug = 4, transect information; record number, time, position, distances
idebug = 5, beam depths
idebug = 6, bearings
idebug = 7, consecutive records which differ in number of bins
idebug = 8, plot option

PROGRAM ADCP_MOD

The resulting files for each transect were processed for comparison with nowcast/forecast results. Within each file, at each point bins were considered bad if 1) the percent good pings were less than 100 or 2) the error velocity was greater than 10 cm/s. Points were discarded based on whether any of the following criteria were met:

1) the top two surface bins were bad.
2) the three bottom bins were bad.
3) there were less than six good bins

It was also necessary to set the bin depth to the depth of the last good bin. For points with bad bins, a linear interpolation from the nearest above and below good bins was used to fill in the bad bin value.

Once all bad bins had been replaced with interpolated data, bin averages of the two horizontal and vertical velocity components were formed. Based on the horizontal component averages and average transect normal velocity component was computed. The transect average normal velocity and average vertical velocity were used to compare with the appropriate Houston Ship Channel Model grid cell values over the nowcast period corresponding to the survey.

HSC SURVEY DATA INFORMATION

To request survey data, please send your inquiries to
Mr. Philip H. Richardson
NOAA/NOS/Coast Survey Development Laboratory
Phone: (301) 713-2809 ext. 115
Fax: (301) 713-4501

Last updated on January 26, 2001