Transient SMBMassBalance Units

DominoAJones

Hi,

I'm forcing a transient simulation with PDD but can't seem to find the units for SMBMassBalance in md.results.TransientSolution?

Cheers!

mathieumorlighem

Hello @DominoAJones
it should be m/yr ice equivalent (same as md.smb.mass_balance if you were to use the default SMB class).
Cheers
Mathieu

DominoAJones

Hi @mathieumorlighem

Thanks! Related, if I have a year's worth of reconstructed temperature and precipitation stored in md.smb - would it repeat the forcing for that year every year if I ran it over a 40 year period?

Cheers!

mathieumorlighem

well, it depends on the field. smb.monthlytemperatures, smb.precipitation, smb.temperatures_presentday and a few others are meant to have 12 columns so these will be repeated (it's the climatology). They do not have a last row with the time for which each column applies. For all other fields that have an extra row with time stamps, they will not be repeated and if you only provide the first year, the last column will be used after one year.
I hope this makes sense!
Mathieu

DominoAJones

Thanks for the clarification!

I've running transient simulations through 40 year periods with the same year's PDD reconstructed temperature and precipitation data repeated again and again (with the time stamp row adapted accordingly). When I plot the average mass balance, it should be the same but I end up with something like this:

Is there some other way I can force the model with PDD to reach a steady state?

mathieumorlighem

Hello
Did you set the lapse rates to 0? Otherwise the temperature is adjusted based on the evolving elevation of the ice. I will forward your question to Josh, he is our PDD expert.
Mathieu

DominoAJones

Hi Mathieu,

Great I adjusted the lapse rate, which gets me something like this:

Definitely better 🙂 Curious to hear what @joshcuzz thinks. Cheers!

DominoAJones

Quick follow up, the above plot is when I run a 40 year simulation with md.transient.ismovingfront=0;. If I keep everything the same but set md.transient.ismovingfront=1;, now the average mass balance is 0.26 m/yr IE for every time step.

Would love some help with this!

Cheers,
Domino

joshcuzz

Hi Domino.

Thanks Mathieu for the tip. The smb will adjust through time based on the lapse rate feedback, which it seems you were able to figure out and adjust. If I understand correctly, it seems once you set the movingfront = 1, your mass balance becomes positive instead of negative? One thought is now that your levelset is turned on, the ice covered grid cells that are very thin (or at your minimum thickness) melt out and become ice free (these thin ice areas before you set movingfront = 1) may have been negative SMB and could have been contributing to the overall negative mass balance.

You can also feel free to email me if its easier to take this offline - adding more and more feedbacks could get tricky to decompose, and there are other variables that can be looked at to get a sense. For one, you could try and output the MaskIceLevelset and see how that changes through your run.

Josh