Dear all,

I am trying to run RegCM-4.3.5.2 for the time period 1989 to 2000 with 50 Km horizontal resolution. For several tries and with different size of dt and dtsrf the model always run for couple of years and then crash saying CFL violation. I just wonder, why the model is selective? if the numerical stability is not satisfied how the model runs for couple of years before crash.

Here are values of dt and dtsrf I have used

1- 150 and 600 (crash somewhere at September 1995)

2- 120 and 600 (crash at 1995-09-24)

3- 60 and 300 (crash at 1992-04-02 )

All the best,

/Abubakr