dotty plots and summary stats

[GULESER SUFRACI]

Hello friends,

After I finish my calibration, I got my results and dotty plots seem confusing to me. Also, in summary stats file, I have got two different results which one is I am going to use? Could you please help me to understand what dotty plots mean?

I am looking forward to hearing from you.

Thanks,

Guleser Sufraci

[pj...@stanford.edu]

Hi Gulecer,

1. The dotty plots are confusing because there isn't much structure to them--the objective function seems pretty random relative to most of your parameters. The one that is clear is SOL_K, which has a better objective function (better match of model to the data) at lower SOL_K. You might want to extend that to the negative to see if the objective function keeps increasing. Also, consider whether you really want to increase SOL_K by 40x or whether you trust the measurements more than that.

2. The first stats are for all of the simulations. The second are only for the "behavioral" simulations--that is, those that meet the objective function criteria you specified (objective function >0.5?). The NSE and R^2 are the same because they only show for the best simulation, while things like the p-factor and r-factor are different because they account for all the simulations.

Hope this helps.

Best,
James

[Abbaspour, Karim]

Dotty plots show objective function values as a function of parameters. It shows the sampling points. It also shows if the objective function is sensitive to the parameter or not. If points are scattered and haphazard, the sensitivity is low. If there is a trend then sensitivity is higher.

Karim

 

-------------------------------------------------
Dr. K.C. Abbaspour
Eawag, Swiss Federal Institute for Aquatic Science and Technology
Ueberlandstr. 133, P.O. Box 611, 8600 Duebendorf, Switzerland
email: abba...@eawag.ch
phone: +41 58 856 5359
fax: +41 58 856 5375
http://www.eawag.ch/index_EN

--
You received this message because you are subscribed to the Google Groups "SWAT-CUP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to swat-cup+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[GULESER SUFRACI]

Dear Dr. Abbaspour,

I know this might be a nonsense question but I have to ask. I know we have to use r (relative) option for soil parameters and CN2. Based on this information can I give the range 0.99 to 99 for sol_k to calibrate the model?? (below table is other parameters I have used for model calibration)

I am looking forward to hearing from you.

Thanks,

Parameter name               Min/Max

r__CN2.mgt                      -0.23 to 0.15

v_GW_DELAY                 10 to 120

v__ESCO.hru                     0.01 to 0.1

R__SOL_AWC.sol            -0.37 to0.25

r_SOL_K.sol                     -0.99 to 50.9

r__SOL_BD.sol                 -0.38 to 0.25

v__CANMX.hru                 2 to 9

r__SLSUBBSN                  -0.5 to 0.5

[GULESER SUFRACI]

Thank you for your explanation. It was helpful.

[Abbaspour, Karim]

r__ mean the existing number times 1 plus the number in the range you provide. If you give 99, it means the existing number times 100. Is that acceptable hydraulic conductivity value? 

Karim

[GULESER SUFRACI]

It seems it is not. Thank you so much again to response very quickly. I will change the paramater range for soil_k.