Does "Perturbations start mode: coldstart" cause issues when Start mode is set to restart for GRACE TWS DA run?

[ThangjamSatrajitSingh]

Hi,
I am currently running the NoahMP4.0.1 for GRACE TWS DA. During the DA run, the model was restarted multiple times using the "restart" option for the Start mode, with corresponding restart files provided, but the perturbations start mode was kept on "coldstart" each time. Here is the lis.config file script snippet for reference:

.
.
.
Start mode:				                   "restart"
.
.
.
#Perturbation options
Perturbations start mode:                  coldstart  #issue?
Perturbations restart output interval:     1mo
Perturbations restart filename:            none
.
.
.

Does this affect the ensemble consistency or DA performance? And what kind of issues should I expect from this mismatch?

[dmocko]

Hi @ThangjamSatrajitSingh

I believe this will not adversely affect your simulation.

The only effect, from my understanding, is that your simulation is not reproducible.

There is a random seed that is used to initialize the perturbations from the coldstart.

There nothing wrong scientifically with the simulation; it's just that you can't exactly re-create it.

[ThangjamSatrajitSingh]

Hi @dmocko,
Thanks a lot for the clarification! I was worried that it might throw off the entire simulation result. It's a relief to know that the accuracy of the simulation result is not affected. Really appreciate your input!