Variable v in CXTFIT

tbann · Post by **tbann** » Fri Oct 26, 2012 8:26 pm

I am using CXTFIT for a long-term problem with a deep soil profile. I'm utilizing CXTFIT because it is very quick with easy data setup and provides a stochastic analysis. The problem I would like to simulate involves not just a variable concentration boundary condition, which CXTFIT can do, but also a variable velocity - a 2-step velocity and concentration input. Is it acceptable to make 2 separate CXTFIT runs for the two different periods of velocity and concentration and add the concentration results at the overlapping times (i.e, superposition)? I'm assuming equilibrium adsorption (linear, instantaneous Kd). Thanks in advance for your help.

ntoride · Post by **ntoride** » Mon Nov 05, 2012 5:13 pm

Sorry for my slow response. I probably do not understand your idea well. Do you want to calculate time-dependent (2-step) input velocity for the CDE? Or do you want to evaluate the mean concentrations for the two stream tubes having different velocities as similar to Fig. 4.1 in the manual? Can you describe your stochastic model a bit more in details?

Nobuo

tbann · Post by **tbann** » Mon Nov 05, 2012 6:34 pm

<blockquote id="quote"><font size="1" face="Verdana, Arial, Helvetica" id="quote">quote:<hr height="1" noshade id="quote"><i>Originally posted by ntoride</i>

Sorry for my slow response. I probably do not understand your idea well. Do you want to calculate time-dependent (2-step) input velocity for the CDE? Or do you want to evaluate the mean concentrations for the two stream tubes having different velocities as similar to Fig. 4.1 in the manual? Can you describe your stochastic model a bit more in details?

Nobuo

<hr height="1" noshade id="quote"></font id="quote"></blockquote id="quote">
Hi Nobuo - thanks for getting back to me. Let me clarify: I'm wanting to simulate a 2-step time-dependent velocity. For example, 365 days at a higher velocity (due to a higher infiltration rate), followed by several years of a much lower velocity. I know that CXTFIT doesn't allow for time-dependent velocities, but I was wondering if I could carry out two different CXTFIT simulations (one for each velocity) and add the resulting calculated concentrations from the two separate simulations according to the superposition principle.

ntoride · Post by **ntoride** » Tue Nov 06, 2012 1:24 am

Strictly speaking, you cannot simply add two solutions. After the 365 days application of a higher velocity, solutes remain in the soil profile when the velocity decreases. You need to consider the remaining solutes as an initial condition for the subsequent lower velocity condition. Although it is not so difficult to conduct two CXTFIT calculations. it would be easier to use the numerical evaluation such as HYDRUS.

Nobuo

tbann · Post by **tbann** » Tue Nov 06, 2012 1:18 pm

Ok. I could use HYDRUS, but it does not have a stochastic module as does CXTFIT. Perhaps I could set HYDRUS up in a batch mode and run a Monte Carlo type simulation based on the mean and standard deviations of the input parameters. Or, run CXTFIT stochastically in two steps: with output of the first step as initial conditions for the second step.

Thanks for your input Nobuo - I appreciate your replies.

Tim