HEC-DSS for large inflow data input in unsteady analysis.

Dear Hec experts

My unsteady simulation requires an upstream inflow and downstream stage boundary conditions. Unfortunately, both data have more than 10,000 rows which exceed the limit of Hec-Ras. I therefore cannot simply copy and paste my excel table over. I know the user manual suggests HEC-DSS link as the alternative, but I know nothing about HEC-DSS. Please all forgive my ignorance.

So I went to the HEC website to download. There were 5 exe files to download but I had difficulties with the 4th and 5th file (dss1of 2 and dss2of2). Maybe they are very big files.

I already printed the user's manual from the pdf file. There seems to be many utility tools. Can someone kindly point me the procedures I need to convert my excel data table (for the boundary conditions) into HEC-DSS format, so I do not have to lose too much sleep.

Thank you for your kindness.
Yao
Yao.HungLan@GVRD.BC.CA

Yao,

Sounds like you need to use DSSTS for entering time-series data (since your hydrographs are at evenly-spaced time intervals). There is a nice example in the manual. You can set up a batch file by separately arranging your stage and discharge data into a single column and exporting to a text file. Edit the text file to give DSSTS the necessary commands in the same order you would enter them in the interactive mode. Then, when you call dssts, just enter the name of the batch file (make sure you saved it in the correct directory) after you type 'dssts' and viola, you're ready to go. Hope that helps.

Cheers,
Jeremy
FranzJ@AYRESASSOCIATES.COM

Hello Yao;

Use <DSSTS>,a small stand-alone program, to enter manually your Regular Interval Time Series OR to convert an existing ASCII file into a DSSfile.

To convert an ASCII file to a DSS file, 5 lines are to be added to the top of the one- column (ASCII/text) file containing the time series and --- an END is to be placed at the very last line of the file:

1st Line: DSS Filename to be created.
2nd Line: Full Six-part Pathname (A=project name B=site/station
C=the data variable e.g., FLOW , STAGE ,etc.
D=01JAN1955 E= timestep,e.g., 2HOUR F= remarks OBSERVED
3rd Line: Unit of data ( CMS, Meters, mm , MCM, mamsl, etc.)
4th Line: Data Type [could be any of the PER-AVER (average for the
period) or PER-CUM or INST-VAL (instantaneous values) or INST-CUM].
5th Line: Date & Time for the first data value, e.g. 09OCT1995,0600

An END should be typed in after the last line of the data is entered.

In the INTERACTIVE mode the above listed items are asked by the program & the user just types them in.

You need separate (ASCII) files for your flow and stage time series. Here is an example from your Excel file for your first few FLOW data:
------------------------------------
YAO-1.dss
A=YOUR PROJECT B=SITE-8 C=FLOW D=01JAN2000 E=2HOUR F=JUST TESTING
CFS
PER-AVER
11FEB2000,1200
1106 <--- cut and paste your FLOW data here (starting @ 6th line)
1106
8
1106
1106
1106
1106
1106
1106
1106
... <-------- snipped
...
END
------------------------------
Now when finished creating the text file and naming it (e.g., YAO-1.txt) then enter the following command : DSSTS IN=YAO-1.txt , you'll get the DSS file (YAO-1.DSS) you need, as Jeremy explained, to import into your RAS model.

Good Luck!
Mostafa Bozorg-Zadeh
bozorg@KOSAR.NET

Yao:

Refering to what Bozorg wrote; Note that line 2 will be formated as "/A/B/C/D/E/F/".

Also, since you are using a 2 hour time interval, part D of the path should be the first day of the month of the year your time series begins (ie. 01JAN1955 if time series begins anytime in January 1955 or 01FEB2000 if times series begins anytime in Feb 2000 as indicated by Bozorg's example). This is because the DSS Block Length for data in hourly intervals is one month and your Block Start Date, which is part D of the path, should be the first day of the month of the data series start. If the Block Start Date is incorrect, DSS amy still read the file but you output data may be split into more than one file path, which is more cumbersome to work with.

I think your input files should look something like the trucated
examples below. Of couse, the dates will correspond with your data. Also, since I don't use meters, I am not sure if "CMS" and "METERS" are acceptable data types. Finally, DSS is very picky. Pathnames must be exactly correct. An extra space will make file unreadable.

For Upstream Boundary:

YAO-1.DSS
/PROJECT NAME/UPSTREAM/FLOW/01FEB2000/2HOUR/OBS/
CMS
INST-VAL
11FEB2000,1200
1106
1106
8
1106
1106
1106
1106
"
"
END

For Downstream Boundary:

YAO-1.DSS
/PROJECT NAME/DOWNSTREAM/STAGE/01FEB2000/2HOUR/OBS/
METERS
INST-VAL
11FEB2000,1200
438.152
438.152
438.137
438.139
438.141
438.142
438.143
"
"
END

Yao, Good Luck
Jim Bernard, PE
bernardj@pbworld.com

Dear Bernard, Bozorg and Dudley (and other users)

Thank you very much for your detailed guidance in DSS. My Hec-Ras can read in the DSS I prepared now.

However, for an output interval of 2 hr, I can only view stage and flow hydrograph (6th button from the right) for 3 month worth of data (about 1080 intervals). If I run the unsteady simulation any longer, I will not see the stage and flow hydrograph. The error message " Some missing data was detected" (-901).

When I tried to use DSS view to see the stage and flow hydrograph, the message says error loading DSS data to plot/tabulate. OVERFLOW.

I wonder what happen here? I should be able to generate 10 yrs of output when I read 10 yrs of input data via DSS.

Thank you again for your precious time.
Yao Lan

Hello,

Thanks to Jim B for the pointing to the most efficient option for the storage Block size needed declared on part D of the pathname: monthly block is the best for 15-min to 12-hr data. As for the 2nd line of the ASCII file, both formats are equally correct & acceptable.

And Yao, I have no problem at all in 'viewing' your full set of data (each series contains 43801 data values = 43800,2-hr intervals =3650 days ~ a decade) --- I use a quite handy utility called DSSVue for just viewing and/or editing DSS files. If you can not 'view' your full one decade of data (outside RAS) then you might have some problems with setting the correct Start Date & Time and/or with the other parts of the pathname; Jim's cautionary remarks should be taken seriously. And when you finally pass this stage and are sure that your data are correctly transferred into a couple of DSS files, you may then want to take a look at the HEC-Users archive : some colleagues believe that the present version of RAS (v.3.0.1) has serious problems in importing DSS files.

Hope the RAS is kind to you and to your particular case.

Best of Luck!
Mostafa Bozorg-Zadeh (ICE)

Yao,

I have never used HEC-DSS with HEC-RAS. I have used it a lot with HEC-HMS abd UNET. So I may not be the best person to look at your stuff.

I did look at your file "output.dss". I noticed a few things:

First, there is no ouput data in the file, only a stage and flow input hydrographs. Second, you have 10 years worth of data in your file begiing October 1964 and Ending September 1974. Third, your stage hydrograh is units of "METRES". It should probably be "METERS" (English). Fourth, the flow file is in units of "M3/S". In english units, HEC-DSS flows are given as "CFS" not "FT3/S" so I assume the the correct unit may be "CMS".

I tried to run your model using your dss files. I got the following
error

Error - Incomplete unsteady flow data, the following errors were found:
Boundary at River: Columbia River Reach: Arrow Lake Revel RS: 245
Undefined values found in the time series data.
Boundary at River: Columbia River Reach: Arrow Lake Revel RS: 1
Undefined values found in the time series data.

This seems to confirm you are using wrong units in dss.

I suggest you check and revise your DSSTS input files and reload the data into HEC-DSS. HEC-DSS only stores data. It does not interpret it. So you may be able to load dss with incorrect nomenclature. DSS doesn't care. However, when HEC-RAS reads the file, it expects to see it a certain way. If the DSS data is formated with incorrect nomenclature, it will not read the data.Also, it may simplify things if you begin with a shorter input hydrograph, say only one months data.

Also, If you cannot find a way to view your dss data using HEC-RAS, you may try using the DSS program DSPLAY. Assuming you have your DSS programs, including DSPLAY, installed under the c:hecexe and your dss data is in the same directory, you may view data using DSPLAY this way:

1. Open DOS prompt and go to directory c:hecexe. Type DSPLAY.3. The DSS files in the directory will be listed. Enter the DSS file name (YAO-1.DSS) you want to open.
2. You should see "-----DSS---ZOPEN: Exiting File Open, File: YAO-1.DSS. Below that there will be a prompt "D>"
3. At the prompt type the command "CA.N". This command requests DSPLAY to create a new "N" and abbreviated catalog "CA" of DSS pathanames containing data. You should see a numbered list of DSS pathnames. Press enter to scroll though the list. You may recognize the pathnames for your hydrograph input files. The pathnames ending with "COMPUTED" or "SIMULATION" are output data.
4. To view your data on the screen in Tabular format. Type command TA, ##, where "##" is the list number before the DSS pathname you want to see. You may also type the entire pathname, but tha this a lot longer. Don't forget the space between the comma and the list number.
5. When you are done with DSPLAY. Simply type the command "FIN".

Jim Bernard, PE
Parsons Brinckerhoff Quade & Douglas, Inc.

Hi:

I have been trying to follow up the instruction given to create a DSS file from observed data but have not been succesful. I got an error message saying that interval cannot be read. this is part of my input file. If anybody can help me I do really appreciate it.

JOSH.dss
A=SIXPOINTS B=TRANSDUCER3 C=PRECIP-INC D=02FEB2002 E=20MIN
F=TRANSDUCER3
CFS
PER-CUM
02FEB2002, 0240
0.2
0.1
0.1
0
0.1
.
.
.
0.5
0.6
0.1
END

Bill Medina
KCI Technologies

Bill,

if this is a file for incremental precipitation, try this:

JOSH.dss
/SIXPOINTS/TRANSDUCER3/PRECIP-INC/01FEB2002/20MIN/TRANSDUCER3
INCH
PER-CUM
02FEB2002, 0240
0.2
0.1
0.1
0
0.1
.
.
.
0.5
0.6
0.1
END

Hope this works,
Jim Bernard, PE
bernardj@pbworld.com

Sorry, I think you should have a forward slash after the second
Transducer3 of the path

/SIXPOINTS/TRANSDUCER3/PRECIP-INC/01FEB2002/20MIN/TRANSDUCER3/

Also, John Morgan brings up a good concern. Perhaps it would be better
to replace the ". " with "0.0".

Hello Bill;

Use a monthly block to store your 20-min interval data,i.e., the D part of the pathname should start on the very first day of the month (01FEB2002,in your case), just as John explained.

If you're trying to store "incremental" precipitation in a DSS file then you should have the right UNIT on the 3rd line : INCH (or MM) instead of CFS; just as Jim pointed out. And if it indeed is a sort of INCREMENTAL data and not CUMULATIVE then the 4th line should read PER-AVER (i.e., AVERage for the PERiod) rather than the PER-CUM. It would also be wise to avoid using too long identifiers for the pathname parts A, B, and F.

It sure works for you too, Good Luck!

Mostafa Bozorg-Zadeh (ICE)

Bill,

I think your problem is related to your DPART of the pathname. 20 minute data is stored by the month. Your DPART should read

D=01FEB2002

I am not sure how the program handles data where you just have a decimal point without a number.

John G. Morgan
john.morgan@cte-eng.com