Numerical instability caused by Q-H boundary?Numerical instability caused by Q-H boundary?https://oss.deltares.nl/c/message_boards/find_thread?p_l_id=1806765&threadId=6697622020-10-21T16:09:35Z2020-10-21T16:09:35ZRE: Numerical instability caused by Q-H boundary?Rudy Schuederhttps://oss.deltares.nl/c/message_boards/find_message?p_l_id=1806765&messageId=6766692015-07-03T12:53:54Z2015-07-03T12:53:54ZHi,<br /><br />It appears that I am able to remove the "wiggle" in water level and discharge when I omit temperature as a modeled constituent. I am not sure why, but I expect it had something to do with the fact that I was specifying temperature in my effluent Q-H boundary condition. Although discharge was always out of the model domain (discharge was always negative), there were observed temperature influences at the effluent boundary in my model output. For instance, heat entered the model from the Q-H boundary at times when the prescribed boundary temperature was higher than the temperature of the adjacent cell within the domain.<br /><br />Does anyone have any ideas as to why a temperature boundary condition would have affected temperature within the domain when water strictly exited through the aforementioned boundary? Perhaps diffusion processes from boundary into the domain?<br /><br />Thank you,<br /><br />RudyRudy Schueder2015-07-03T12:53:54ZNumerical instability caused by Q-H boundary?Rudy Schuederhttps://oss.deltares.nl/c/message_boards/find_message?p_l_id=1806765&messageId=6697612015-06-24T15:50:56Z2015-06-24T15:50:56ZHi all,<br /><br />I have been getting some odd results for WL that manifest about a month into my simulation. My suspicion is that they are the result of some numerical instability because of the simplicity/steady state condition of the model. Please see the attached .pdf for a visualization of what I mean. The first two plots show discharge through (CS effluent) and WL at the effluent boundary (L2). Notice how they begin to go wonky at the beginning of May. The third plot is discharge through the influent boundary (CS influent). It remains constant as prescribed. <br /><br />This is a description of the simplified model that gives me these odd WL readouts:<br />- 1 influent boundary (Pipe): constant total discharge prescribed across one grid cell<br />- 1 effluent boundary (Weir): Q-H relationship describing flow at WL between 0-1 m. WL should never go above 0.06 m. Prescribed at one grid cell<br />- both boundaries have a reflection parameter alpha = 86400 s^2<br />- modeling temperature and pollutants, no wind, waves, or sediment<br />- ts = 0.05 min<br />- I have attached a picture of the grid (cross sections removed for clarity) and the input file for further description<br /><br />Does anyone have any experience with this kind of result? Does it in fact look like a numerical error? I would have expected WL to remain stable since all inputs of water are stable....<br /><br />Any help is much appreciated and thank you in advance,<br /><br />RudyRudy Schueder2015-06-24T15:50:56Z