You are here

Home › Forums › W2 Forum › Water Quality

Reading Downstream concentration (CHD) and temperature (THD) constituents

Mon, 08/27/2007 - 4:10pm — bobolsn
I am attempting to create a "relative difference" model for a small estuary. What are the effects of the reduction in cross sectional area? The model is running, no PreProcessor or run time errors, but it appears that the model is not reading the correct temperatures and TDS concentrations. I started with the DeGRay Reservoir example and modified to be a free flowing model (no structure) with a downstream head boundary. I only have TDS (salinity) as an active constituent. I did not change the upstream Branch Inflow constituents so as not to change the Cin_Br1.npt file. I just changed the TDS in the upstream file for the period I am running to reflect fresh water. I tried both PQC ON and OFF and other variations. I tried various file formats without success. Review of the output files (SNP,CPL and PRF) indicate that the elevations in the EDH file are being correctly read, but the temperatures in the system gradually lower to 0.0 or negative during the course of the simulation. Also, the TDS does not reflect the higher TDS at the downstream boundary. Of course the temperature issue creates havoc in the computations with the different thermal gradients. Several questions: 1. Any suggestions? 2. Do the CDH and cin_Br1 files use the same format (Jday with columns with each acitve constituent or Jday then constituent concentration for each layer)? Ditto for the Tin and TDH files. 3. Do the files need the same input time steps, or is it adequate to span the simulation window and the program will find the appropriate date and interpolate? 4. Does the CDH and TDH files need to have temperature and tds by layer? I noted in one example in the manual(page C-307) that TDS and Temp were in the same file and layered. 5. Any other procedure to check the what is being read? I could not find a flag that would have the program regurgitate the input. Thanks Bob
Forums: 
Water Quality
Tue, 08/28/2007 - 8:55am — Rob

Bob, The Tin and TDH files are different. The Tdh files require a temperature for each active layer at the downstream model segment of the Boundary Condition. See page 307 for an example of the file setup. The sam eis true for the concentration file. The Cin is simply a time series and the cdh file need concentration for each active layer and then each additional water quality constituent is added for the same time before you move on to the next time step in your file. Page C-308 has good overview of the file. The input files do not need to have the same time step. The model will interpolate between whatever frequency of dates you have in each file (as long as you turn on interpolation in the control file). The preprocessor pre.opt file should have statistics on all of your input files and be able to give you a hint as to what is being read in. I suspect that if you are getting zero then you have time series data but not fr each layer in the model grid and hence zero are read in. I hope this helps. Cheers, Rob
Wed, 08/29/2007 - 7:51am — bobolsn

Rob, Thanks for the prompt response. I had been down the road of the file with constituents for each layer (page 307 format), but was confused by several things. First the example on page 307 has temperature and salinity in the same file but the manual said that had to be seperate. Also, the variable in the 307 file is CUH which is an upstream concentration. Lastly, I repeatedly got error messages from the Preprocessor similar to the following: "Julian date 0.000 <= previous date 366.000 in tdh_br1.npt Ending time 0.000 < ending simulation time [TMEND=255.000] in tdh_br1.npt Julian date 0.000 <= previous date 366.000 in cdh_br1.npt Ending time 0.000 < ending simulation time [TMEND=255.000] in cdh_br1.npt" I did review the code and determined that the model was expecting the acitve layer concentrations and temperatures for each time. Though I get the above error messages, the model runs and appears to be using the correct downstream WQ data (temperature and TDS). The above messages are why I questioned the number of input time steps. Again I looked in the preprocessor code and it appears the preprocessor is looking for more input data times (It reads another line with no data, JDAY is therfore 0.0 and ergo the above messages). I only have 4 or 5 Jdays in the downstream files while I have 365 in the upstream files. If I am not correct in my assumptions, please respond. On a results subject, why would the TDS increase at the lower depths. For instance, the maximum TDS that is entered at the downstream boundary is 37 and the results after a 6 week simulation indicate TDS concentrations as high as 51 on the bottom layers. Is it a numerical issue that concentrates the TDS? I looked at the model coefficients to see if there was a TDS release from the sediments but could not find anything. Precipitation and evaporation are OFF as are all other constituents at this time. The system is being "flushed" with 50 M^3/sec with 5 TDS. The fresh water pretty much stays on top with some dispersion/mixing, but the lower layers are "concentrating". I will try various time steps adn look for other potential problems. I am familiar with helping first time users and do appreciate your assistance. Thanks Bob
Wed, 08/29/2007 - 8:45am — Rob

I notice the CUH as well. This is just an error in the manual. Temperature should be seperate from concentration. Also if you are getting Jdays of 0.00 then it means you have too many values in the concentration file for each date. Base don the water level at the downstream end you should have concetration values for only the active layer, i.e. the "wet" layers. So for example, if you have 20 vertical layers (18 in computational grid and 2 layers, 1 top and 1 bottom that are null) and then the water elevation data puts the water surface at layer 8, then you really only need layer 8 to 19 to have values or 12 values. I hope this helps. Cheers, Rob