Welcome, Guest
Username: Password: Remember me
Forum on HARMONIE Surface development

TOPIC: Glaciers plans and work

Glaciers plans and work 3 years 7 months ago #1522

Here is link to the glacier related wiki:

hirlam.org/trac/wiki/HarmonieWorkingWeek...201604/Glacier_plans

Glaciers plans and work 3 years 7 months ago #1530

Hi!

The email below was just sent to the Météo-France surface team to ask for guidance and advice on how to proceed with glacier development in SURFEX:

From: Samuelsson Patrick
Sent: 29 April 2016 11:08
To: Patrick Le Moigne ( This email address is being protected from spambots. You need JavaScript enabled to view it. ); Aaron Boone ( This email address is being protected from spambots. You need JavaScript enabled to view it. ); FAROUX STEPHANIE ( This email address is being protected from spambots. You need JavaScript enabled to view it. ); DECHARME Bertrand < This email address is being protected from spambots. You need JavaScript enabled to view it. > ( This email address is being protected from spambots. You need JavaScript enabled to view it. ); ' This email address is being protected from spambots. You need JavaScript enabled to view it. '
cc: This email address is being protected from spambots. You need JavaScript enabled to view it. ; This email address is being protected from spambots. You need JavaScript enabled to view it. ; Kristian Pagh Nielsen ( This email address is being protected from spambots. You need JavaScript enabled to view it. )
Subject: Glaciers in SURFEX

Hi Météo-France SURFEX friends (cc Ruth, Emily, Kristian)

HIRLAM had a surface workshop in Oslo recently and there we discussed the ideas on glacier development that were mentioned at the last SURFEX steering committee. We have compiled a wiki page dedicated for our glacier development discussion:
hirlam.org/trac/wiki/HarmonieWorkingWeek...201604/Glacier_plans
Until now Ruth Mottram (DMI), Emily Gleeson (Met Éireann) and Kristian Pagh Nielsen (DMI) have been running the glacier studies in our operational NWP environment (with SURFEX7.2).

Now we have a few questions to you guys on how we may proceed:

There are two different complexity levels of glacier models discussed:
- One is fairly advanced and would fulfil the needs requested by regional climate modelling. I don’t know any details so we need Ruth Mottram at DMI to explain the details.
- One is fairly simple to fulfil the needs requested by NWP (current method to initialize permanent snow areas with huge snow layers does not work well over e.g. Iceland and Greenland).

The NWP idea has again two alternative ideas discussed:
(i) To utilize the ES snow scheme (or Crocus) by defining very dense snow layers (ice) at the bottom of the snow pack which then should represent the glacier ice and that the upper layers are more “normal” snow. But I’m not sure how it would work… We should be able to simulate the situation when seasonal snow pack disappears and an ice surface appears which has quite different albedo (given e.g. dust accumulated with time)….
(ii) Utilize the sea-ice model a la Yurii as now implemented in a HIRLAM version of SURFEX7.3. According to Yurii this should be pretty straight forward. Then the Explicit snow scheme would be used for the snow on ice. Here the characteristics of the ice itself can be specified and the interface snow/ice is well defined.

Independent on above ideas, I guess whatever glacier scheme we call, it should be done from isba.F90 when some glacier namelist option is activated, right? We should not implement a separate tile, right? For grid boxes with 0 or 100% permanent snow it is not a problem but for other grid boxes (given NPATCH<12 or 19) it would be very strange to lump the permanent snow (glacier) characteristics with soil/veg characteristics since we look for realistic ice-melting conditions. In our NWP setup we are now testing 2 patches but with glaciers we would need 3 patches to keep the ice patch physically realistic. But that would mean a redefinition of the case NPATCH=3 into (i) permanent snow (ii) forest (iii) remaining open land. Hmhm, or how can we proceed…? Please give your opinion and ideas!!

On behalf of HIRLAM glacier team,

Patrick

Glaciers plans and work 3 years 7 months ago #1532

Here I paste in a response from Patrick Le Moigne on the email to the SURFEX team.

/Patrick Samuelsson


Hi Patrick,

We met this morning with Stephanie, Marie, Bertrand and Aaron. We went through the different topics and proposals raised in your wiki document. Here are the main conclusions and proposals from our side.

The use of ISBA-ES model based on Surfex v8 should solve most of your problems, both for NWP and the Regional Climate Model. It would be of great interest to redo the study conducted by Ruth with Surfex v8 and ISBA-ES. (The use of D95 doesn't work properly probably because the LGLACIER switch was not activated. Its goal being to ensure a sufficiently high albedo in time and a large snowpack). You could do tests with LGLACIER activated or not and also use 12 patches (to be sure to have the glaciers separated from the others). Note that using ISBA-ES in Arpege Climate model works well, particularly over Greenland or Antarctica.

The use of Sea-Ice model combined with ISBA-ES is clearly not a suitable solution since the Sea-Ice model solves the heat transfer by diffusion for a soil having the ice characteristics: the proper solution is to use ISBA-ES, a more physical scheme for snow and ice that solves the heat equation in a continuous way (for both snow and ice, infact this currently done in Arpege-Climat). Second, such an approach (putting ISBA-ES on top of the ice model) will never be used on our side and this would limit the scientific exchanges between all of us on that specific topic.

There are a few important comments that we would like to add:
• there's a need to take care of the water from melted snow to avoid a mass loss (currently melt "runs off"). Eric Brun had worked on a simple solution to retain this water somehow, we're digging into the past to check what he did
• when snow density becomes greater than 850kg/m³ then snow albedo for each spectral band should be set to ice albedo (this is what is currently done in Crocus)
• accounting for more than one layer for the albedo computation is especially important when snow thickness is small since the albedo of a thin layer of snow over dark ice is affected by the ice albedo. Currently, ISBA-ES albedo computation is based on one layer only, this has to be modified.
• on our side we're going to start working on these 3 points

Concerning tiles and patches: yes you're right, please do not add any new tile! Then to ensure a 100% fraction of "permanent snow" patch, this can be done in the setup by as you said defining 3 patches: low veg, high veg and glaciers. There's a need of redefinition of aggregation rules for that but it's probably the best solution. Anyway you can start with 12 patches for testing.

As far as ECOCLIMAP is concerned, if there are improvements on the HIRLAM side for the definition of glaciers extent (or any other surface type characteristic), we would of course be happy to benefit from these improvements.

Finally, perhaps the best way to collaborate efficiently on the glacier modelling could be that one of your HIRLAM colleague come to Toulouse to work with us (perhaps Ruth?).

I'll try to join the hangouts meeting on Thursday...

Best regards.

Patrick

Glaciers plans and work 3 years 7 months ago #1533

Hi,

Ok, so if the ISBA-ES model based on Surfex v8 can solve the glacier problems it is great. I just like to hear what Ruth has to say since she has experience from the glacier model used in HIRHAM.

Bolli, as you describe under "Current IMO setup" on the wiki you initialise your NWP system (cold start) with 10000 kg/m2 of snow in autumn over Greenland glaciers but then the snow disappears next summer, at the edge of the ice sheet, exposing a bare soil surface, right? Did you try the LGLACIER option in SURFEX? Or it does the same thing?

As Patrick Le Moigne says, the snow albedo can be set to ice albedo when the density of a snow layer exceeds 850 kg/m3. So what is this ice albedo? Rontu et al. reported dirty ice surface with albedo as low as 20-30%. Is exposed glacier ice always so "dark" or?

How should a snow/glacier(ES snow)/soil column in SURFEX be initialised? 12 layers with the bottom layers (5?) set to density 900 kg/m3 and how deep? Any ideas we can start to test with?

Discussion continuous Thursday May 12 with a Google Hangouts session. See wiki here: hirlam.org/trac/wiki/HarmonieWorkingWeek..._plans#HangoutsMay12

Cheers,

Patrick

Glaciers plans and work 3 years 7 months ago #1534

Hi Patrick, Patrick & Co.!

At DMI we are interested in forecasting the time when the snow melts away. For this purpose the LGLACIER option is no good, since it adds a permanent large snow pack on top of the ice. This is why we want to fix the ice atmosphere coupling in the D95 scheme.

I suppose that the ISBA-ES scheme will first be available in cy43h, which will be ready by 2018 or so. Thus, we need an intermediate solution.

Regarding the ice albedo and the thin layer of snow on ice issues, these can be simulated with a special version of the DISORT model that I have worked with.

For the ice albedo several ice conditions could be considered:
- Solid ice;
- Porous ice with air bubbles;
- Porous ice that is water saturated;
- Ice that is covered with water (lakes or supraglacial rivers);
- Dirty ice with basal glacier debris;
- Dirty ice due to aeolian dust covering;
- Etc.

Due to all of these complications, it is a promising prospect to be able to use albedo maps and albedo assimilation in future SURFEX versions. For instance dirty ice with basal glacier debris can be mapped.

Best regards,

Kristian.

Glaciers plans and work 3 years 7 months ago #1535

Hi all,

Few points from IMO:

We currently setup the model in autum by adding a lot of snow on top of the glaciers (D95 scheme). Reasons for doing this:
  • Snow does not melt away during next summer melt season
  • Gives us a reasonable behavior over the glaciers, Ts, LE, constant melt etc. during the next summer.
  • Fixes big errors in TS/T2m/LE over glacier when previous setup melted snow away over the glacier areas.

But this is a quick fix! We do not get correct albedo over the dirty glacier ice exposed during summer season for example.

Regarding the albedo, not all glacier ice is dirty and dark. It's valid for most of the ice close to the edge, but you can have seasonal snow fully melt further away from the edge and then a much brighter glacier ice is exposed. So I support Kristians idea of using albedo maps in the future to correct for this.

The PGD/ECOCLIMAP update for Greenland is mainly to fix the ice sheet extent and change some glacier areas into rocks.

Regards,
Bolli

Glaciers plans and work 3 years 6 months ago #1538

Hi,

Please find memory notes after today's Google Hangouts meeting on the glacier wiki page:
hirlam.org/trac/wiki/HarmonieWorkingWeek..._plans#HangoutsMay12

Those who participated, please check/correct/add.

Cheers,

Patrick

Glaciers plans and work 3 years 6 months ago #1539

Hi Bolli, Emily et al.,

Sorry for being confused but please remind me again. Did any of you use/try the option LGLACIER=TRUE in your setups.

Cheers,

Patrick

Glaciers plans and work 3 years 6 months ago #1540

Hi Patrick,

Yes I tried it once with the d95 scheme in HARMONIE, but I got a crash. Some of my modifications in routine prep_perm_snow.F90 are very similiar to the LGLACIER option.

Regards,
Bolli

Glaciers plans and work 3 years 6 months ago #1541

Hi Patrick, Bolli,

I tried it with D95 in MUSC but it did not work for me and right now, I cannot remember the details of the fail.

All the best,
Emily

Glaciers plans and work 2 years 3 months ago #1927

  • Caroline Webb
  • Caroline Webb's Avatar
Glaciers start to procedure when snow vestiges in the similar part year-round, where sufficient snow collects to convert into ice. Very informative information providing here on glacier related plan and work. Since I am UK based personal statement writer and happily helping students by providing result-driven personal statement writing services at Personal Statement Folks and I found very valuable information from here which is help me to provide best guidance on glaciers to my students. Thanks.
Time to create page: 0.106 seconds