I have been attempting to run a model with the entire MCM downloaded from the AtChem archive website (https://mcm.york.ac.uk/MCM/about/archive) and the following environmental, model and solver parameters:
1 TEMP CONSTRAINED
2 PRESS 1000.00
3 RH CONSTRAINED
4 H2O CALC
5 DEC CALC
6 BLHEIGHT NOTUSED
7 DILUTE NOTUSED
8 JFAC 1
9 ROOF OPEN
10 ASA NOTUSED
15 number of steps
300 step size (seconds)
2 species interpolation method (pw constant = 1, pw linear = 2)
2 conditions interpolation method (pw constant = 1, pw linear = 2)
3600 rates output step size - ROPA/RODA (seconds)
300.0 model start time (seconds)
0 jacobian output step size (seconds)
1.3521 latitude (degrees: North is positive, South is negative)
-103.8198 longitude (degrees: West is positive, East is negative)
05 day
11 month
2020 year
1800 reaction rates output step size (seconds)
1.00e-03 atol
1.00e-04 rtol
1.00e-04 delta main
100 lookback
100 maximum solver step size (seconds)
100000 maximum number of steps in solver
3 solver type (1 = spgmr, 2 = spgmr + banded preconditioner, 3 = dense)
750 banded preconditioner upper bandwidth
750 banded preconditioner lower bandwidth
For constraints, I have RH and TEMP constraints and 118 species constraints. When I run the model, it starts with this error:
"Internal t = x and h = y are such that t + h = t on the next step. The solver will continue anyway."
Usually the model is able to get past this but it will eventually will crash with the following error:
"[CVODE ERROR] CVode
At t = x and h = y, the error test failed repeatedly or with |h| = hmin.
ier POST FCVODE()= -3
time = x
SUNDIALS_ERROR: FCVODE() returned ier = -3
Linear Solver returned ier = 0
Note: The following floating-point exceptions are signalling: IEEE_UNDERFLOW_FLAG IEEE_DENORMAL"
I've tried adjusting the error tolerances and some models will progress further than others but all of them will crash before completing. Any ideas for how to get around this? Is the system too constrained/large for the solver to handle?
I have been attempting to run a model with the entire MCM downloaded from the AtChem archive website (https://mcm.york.ac.uk/MCM/about/archive) and the following environmental, model and solver parameters:
1 TEMP CONSTRAINED
2 PRESS 1000.00
3 RH CONSTRAINED
4 H2O CALC
5 DEC CALC
6 BLHEIGHT NOTUSED
7 DILUTE NOTUSED
8 JFAC 1
9 ROOF OPEN
10 ASA NOTUSED
15 number of steps
300 step size (seconds)
2 species interpolation method (pw constant = 1, pw linear = 2)
2 conditions interpolation method (pw constant = 1, pw linear = 2)
3600 rates output step size - ROPA/RODA (seconds)
300.0 model start time (seconds)
0 jacobian output step size (seconds)
1.3521 latitude (degrees: North is positive, South is negative)
-103.8198 longitude (degrees: West is positive, East is negative)
05 day
11 month
2020 year
1800 reaction rates output step size (seconds)
1.00e-03 atol
1.00e-04 rtol
1.00e-04 delta main
100 lookback
100 maximum solver step size (seconds)
100000 maximum number of steps in solver
3 solver type (1 = spgmr, 2 = spgmr + banded preconditioner, 3 = dense)
750 banded preconditioner upper bandwidth
750 banded preconditioner lower bandwidth
For constraints, I have RH and TEMP constraints and 118 species constraints. When I run the model, it starts with this error:
"Internal t = x and h = y are such that t + h = t on the next step. The solver will continue anyway."
Usually the model is able to get past this but it will eventually will crash with the following error:
"[CVODE ERROR] CVode
At t = x and h = y, the error test failed repeatedly or with |h| = hmin.
ier POST FCVODE()= -3
time = x
SUNDIALS_ERROR: FCVODE() returned ier = -3
Linear Solver returned ier = 0
Note: The following floating-point exceptions are signalling: IEEE_UNDERFLOW_FLAG IEEE_DENORMAL"
I've tried adjusting the error tolerances and some models will progress further than others but all of them will crash before completing. Any ideas for how to get around this? Is the system too constrained/large for the solver to handle?