MIM Model

[vss0015@auburn.edu]

Greeting Dr. Jirka,

I am using a MIM model for tracer transport modeling. It worked fine for 4 undisturbed columns but for the next 2 columns, the lower 95% C.I is negative. I was optimizing for dispersivity, Theta immobile, and alpha. An estimate of their initial values was found using CXT-FIT.

My approach was to change the alpha value (to decrease in my case) until the C.I was positive. This led to a very low alpha value of 10^-8 /min. Should I treat it as zero or try something else? This could mean solute didn't have enough time to be exchanged between the two regions or negligible solute was exchanged. (The total time for these 2 columns is very short compared to the rest).

I would appreciate your suggestion on this sir.

With Regards,
Vishawjot Singh Sandhu

[Jirka]

If the optimized parameter value tends to go towards zero, you should fix it at a zero value and not optimize. If the confidence interval spans negative and positive values, then there is not enough information in your data (e.g., it is correlated with some other optimize parameter) to optimize this parameter. You should be able to fix it at any value within this interval and obtain similar results. J.

[vss0015@auburn.edu]

Thank you Dr. Jirka, It is really helpful.

Have a good day sir.