I have a 5-layer model, 230 km², uniform 50*50 m cells, 168 stress periods.
LPF, RIV, SFR2, WEL, DRN, ET, UZF (RCH), PCG, MODFLOW-2005
The model is running fine using RCH with time variant and spatially distributed infiltration rates.
Due to the fact that the modeled area has significantly varying groundwater distances within the range from 0 to > 100 m with correponding time series of groundwater heads, I think it is a good idea to use the UZF package.
NUFZTOP=2 (3 does not work, input of array vks is not possible)
IUZFOPT=2 (vertical conductivity is 5E-06 to 5E-05 m/s)
IRUNBND=array (>100 SFR segments, some lakes)
EPS=constant (2.5 to 5.0)
FINF= 168 arrays
After 15 SP modflow crashes.
Examination of the water budget shows that in a few single cells a real huge UZF-RCH is formed (> 10 m³/s !), although there is no significant UZF-INF in the SP's before (the same FINF arrays were used successfully with the RCH package), no other possible sources. The water is generated "from nowhere" within the unsat zone.
The problem cells do not have a thick unsat zone, only a few meters.
Varying THTS, THTI, EPS does not help, all parameters are very sensitive, modflow crashes earlier.
Lowering NTRAIL2/NSET2 as well.
Assigning IUZFBND of these problems cells to zero leads to a shift of the problem to te next right model cells (and so on).
I have attached a text file with selected lines of the volumetric budget of the model from the output of the model (regard SP 15) and a screenshot of the flow budget of a selected problem cell.
Does anyone have an idea what is going on there?
Thanks a lot, any help is really appreciated.