C(z)=Cs-JsZc/D (1-e-z/zc), I want to fit this model to my soil carbon concentrations among different soil depths and obtain Cs, Js/D and Zc parameters. This model first introduced by Wynn et al., 2004. A paper link from Wynn et al.2005 is enclosed.
http://www.sciencedirect.com/science/article/pii/S0016706105000741
You can fit this easily with Excel and the Solver therein.
Please look at the attached file.
In the first column the depth z is given, the second column has a supposed function with the parameters in A1,A2 and A3. By randomization with zufallszahl() you get a dataset with randomized values. This will change every time you change any value in the table.
The values of column C are copied to D and a new function is supposed in column E with new parameters in the cells E1,E2 and E3. The sum of all squares of the differences( from column F) is shown in cell F5.
With the solver you can minimize this value by changing the cells E1,E2,E3.
Hi,
you can use any scientific computing or statistical software. The benefit is that you have also access to statistics on your parameters, the goodness of fit etc.
I attach a small R example using a non-linear least squares fit. In the case of R a huge amount of information is in hte net available.
Dear Professor. Gerhard Herres
I really appreciate your consideration and kindness. Your guidance was definitely useful. Using Solver add-in, fitting the model to any experimental data are easy and precise.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Annealing
- Metallurgy