error (72) floating overflow ( H E L P !!!! ) :-(

[chat...@gmail.com]

Dear All,

I am trying to calibrate my swat project (Multi-site calibration), which was run in SWAT 2012. Since I am not successful with my validation (low NE, R2), I made a new project in SWAT-CUP with the same SWAT output and I have added more parameters using the original range permitted by SWAT (absolute_SWAT_values) but unfortunately I am getting:  error (72) floating overflow (See attachment).
Although I have followed the recommendation given in the group: run swat edit, copy the files from backup folder), still I have the same error.

I am so confused, since I already mentioned I am working with the same SWAT model output and if this error is related with SWAT-model run and not with SWAT- CUP, why did it run with the previous SWAT- CUP project? Any suggestion?

Thanking in advance,

KAREN

[Grey Evenson]

Hi Karen,

If I were you, I'd place a debug version of the swat executable in the swatcup txtinout directory and run it outside​ of swatcup (i.e. open a cmd window in the txtinout directory and type swat.exe). Can you do this and observe the error message from the debug executable?

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Abbaspour, Karim]

Adding more parameters may not be the solution. You may be having a strange set of parameters causing swat to crash. 

I suggest following the protocol outlined in the following paper and only adjusting the required parameters instead of a mix of unrelated parameters.

http://www.sciencedirect.com/science/article/pii/S0022169415001985

Also, you should consider improving your model prior to calibration. A good source of information for that is the following paper:

http://www.sciencedirect.com/science/article/pii/S1364815215300499

Karim

To unsubscribe from this group and stop receiving emails from it, send an email to swat-cup+u...@googlegroups.com.

[Rekha Rana]

Dear All ,

I also have same error Flowout float72 error when I calibrated in SWAT Cup . plz tell anyone that has resolve this error and how I can run swat cup properly. Here is the snap shot of mine error. I run it for 11 years ,it run well for 10 years after that it stuck and shoe the error . plz tell anyone.

[Steven Jepsen]

Karen,

I had this problem once. I found the solution by restricting the limits of one parameter at a time until finding the problem, which ended up being the limits on SNO50COV that caused a function to misbehave. Watch out for the extreme limits on parameters, especially those that have a value of zero or one. For example, SNO50COV can cause problems if too close to zero or one.

I'd suggest starting off with just a few parameters. Then add one parameter at a time while checking calibration stats and making sure there are no crashes (floating pt errors, etc.) If met data are okay, I'd expect that you could get an acceptable calibration with a relatively small number of parameters--see refs including those from Karim. 

Steve

[selam haftuu]

I am also facing the same problem if you find the solution to share it with us I will appreciate it. regards. 

[Mamush Tekle]

Dear  Greye, is debug.exe used as SWAT-CUP when I run debug it works successfully.

but in SWAT-CUP I couldnt solve floating overflow forrtl; error (72)

To unsubscribe from this group and stop receiving emails from it, send an email to swat-cup+u...@googlegroups.com.