MODPATH for USGS GSFLOW simulation: problem with budget file

[porterma]

I am trying to run MODPATH6 on a USGS GSFLOW simulation model

So my first question, is MODPATH6 compatible with GSFLOW?

I assumed it was so I gave it a try. MODPATH6 crashes with this error in command prompt

Processing basic data ...
Checking head file ...
Checking budget file and building index ...
forrtl: severe (170): Program Exception - stack overflow
Image PC Routine Line Source
mp6.exe 004B0787 Unknown Unknown Unknown
mp6.exe 00428174 Unknown Unknown Unknown
mp6.exe 00420C79 Unknown Unknown Unknown
mp6.exe 00445E4A Unknown Unknown Unknown
mp6.exe 004CC213 Unknown Unknown Unknown
mp6.exe 004B0C59 Unknown Unknown Unknown
KERNEL32.DLL 7671B028 Unknown Unknown Unknown

The mplist file looks like this towards the bottom.

THE HEAD FILE FOR GRID 1 IS COMPLETE.

CHECKING AND INDEXING BUDGET FILE FOR GRID 1
Index position = 1
READ FLOW PACKAGE BUDGET DATA FOR STRESS PERIOD 1 , TIME STEP 1 ...
Single Precision Binary Budget file
Reading STORAGE for Period 1 Time Step 1
Reading CONSTANT HEAD for Period 1 Time Step 1
Reading FLOW RIGHT FACE for Period 1 Time Step 1
Reading FLOW FRONT FACE for Period 1 Time Step 1
Reading FLOW LOWER FACE for Period 1 Time Step 1
FLOW RATES BETWEEN CELLS HAVE BEEN READ
Reading STREAM LEAKAGE for Period 1 Time Step 1
Reading UZF RECHARGE for Period 1 Time Step 1

So it appears stream leakage is read fine, but not recharge for the uzf package in the budget file.

The portion of the nam package for output looks like this.

data(binary) 50 ../../output/output_nstp/heads.bin
data(binary) 55 ../../output/output_nstp/bud_cbc.bin
data(binary) 49 ../../output/output_nstp/bud_uzf.bin
data 56 ../../output/output_nstp/str_out.txt
data 41 ../../output/output_nstp/hob_out.txt

My beginning of the oc package looks like this and is saving head and budget data for every timestep and stress period..

head print format 0
head save unit 50
COMPACT BUDGET AUXILIARY

period 1 step 1
save head
save budget

period 1 step 2
save head
save budget

Item 1c of the sfr package looks like this and ISTCB1=55

-1969 118 0 0 1.486 0.00001 55 56 1 0

Item 1b of the uzf package looks like this where IUZFCB1=55.

3 1 0 0 55 49 5 20 0 1.0 NUZTOP IUZFOPT IRUNFLG IETFLG IUZFCB1 IUZFCB2 NTRAIL2 NSETS2 NUZGAG SURFDEP

In my well package IWELCB=55.

There is an option in the UZF to set IUZFCB1<0 (e.g.,-55), which saves infiltration and unsaturated ET to binary file. Do I need IUZFCB1=-55?

There is also IUZFCB2 that writes to unformatted file. I currently have that set to a separate ASCII file. Should this be set to 55 and written as binary as well?

Any thoughts would be greatly appreciated. I will test these ideas myself however the simulation take ~1.5 days, so I thought I would ask here for some guidance.

Also, I apologize if this my second discussion ticket regarding this. I wrote one last night and thought I posted, but do not see it on github.

[jlarsen-usgs]

Hi @porterma,

I'm not sure if MODPATH is compatible with GSFLOW. @rniswon should be able to provide more insight into this.

[rniswon]

I am in communication with Mark. MODPATH is not compatible with GSFLOW. However, results from a GSFLOW model can be exported and converted to a MODFLOW only model with equivalent boundary fluxes, and this model can be used with MODPATH. I am working to connect Mark with someone who has done this in the past.

[porterma]

Apologies for the delayed update here.

I was able to get MODPATH running on the GSFLOW budget file. I simply removed the UZF out put from the budget file and the MODPATH works as expected. I realize I am missing the fluxes to and from the UZF package. However, for my purpose of estimating broad particle travel time ranges this is sufficient.

I was not bale to get MODFLOW to run with the the UZF budget data. I started down a path of running MODFLOW alone with the fluxes from PRMS converted to boundary conditions to MODFLOW. However, I did not spend sufficient time on it to come up with a solution.

@rniswon I appreciate your assistance on and advice with this task.