[CF-metadata] Response to ISMIP6 request

Nowicki, Sophie (GSFC-6150) sophie.nowicki at nasa.gov
Sat Feb 11 09:15:29 MST 2017


Dear Alison, Karl and Jonathan,

Thank you so much for all your advise and feedback for the ISMIP6 data requests.

Below are responses/comments for the remaining ISMIP6 names, and thank you, Alison, for having accepted for publication in the standard name tables the other entries.

Kind regards,

Sophie

———
12, 13 & 14: magnitude_of_*_base
———

12. magnitude_of_shear_stress_at_land_ice_base (Pa)
' The magnitude of the shear stress at land ice base'
13. magnitude_of_normal_stress_at_land_ice_base (Pa)
' The magnitude of the normal stress at land ice base'
14. magnitude_of_longitudinal_stress_at_land_ice_base (Pa)
' The magnitude of the longitudinal stress at land ice base'

The syntax of these three names is consistent with existing sea ice names.
We have one existing "magnitude of stress" name,
magnitude_of_surface_downward_stress, which represents the  magnitude of
the x/y or eastward/northward surface stress components and specifies
"downward", i.e. positive when momentum is transferred from upper to lower
medium. Should your quantities be regarded as "downward" (momentum
passing from the ice to the underlying surface) or "upward"? Units of Pa are
fine.

The definitions will of course depend on the sign conventions of the stresses:
' "magnitude_of_X" means magnitude of a vector X. ["Downward" indicates a
vector component which is positive when directed downward (negative
upward).] OR [" Upward" indicates a vector component which is positive when
directed upward (negative downward).]'
Please can you also provide some brief descriptions of ice shear stress,
normal stress and longitudinal stress that I can add into the definitions?
Answer Sophie: we are working of trying to find good definitions, so I will come
back to you for this point.

Answer Sophie to Alison:
I was double checking the final ISMIP6/CMIP6 data request, and we are no longer requesting these names in order to reduce the ISMIP6 data request size. Can I erase the request of creating a standard names for these 3 basal stresses?

————
16. land_ice_mass (kg)
—————

16. land_ice_mass (kg)
'The mass of land ice (glaciers, ice caps, ice sheet and ice shelves), computed
as the ice volume times density.'

The name and units are fine.

We have two existing names, sea_ice_mass (currently undefined) and
sea_water_mass defined as ' The quantity with standard name
"sea_water_mass" is the total mass of liquid seawater in the global oceans,
including enclosed seas.' Is your quantity globally integrated like the sea_water
name or does it refer to the mass in an individual grid cell? We need to be clear
in the definition. I think the method of calculation doesn't belong in the
standard name definition (you could add it elsewhere, e.g., the comment
attribute, for CMIP6 data).
Answer Sophie: land_ice_mass is a quantity that is integrated for the whole ice
sheet (i.e.: a 1 dimensional variable that varies with time), so it would be the
total mass of the ice sheet, and therefore more similar to the
“sea_water_mass"

The name and units are agreed. I suggest the following as the definition:
' "Land ice" means glaciers, ice-caps and ice-sheets resting on bedrock and also includes ice-shelves. The quantity with standard name land_ice_mass is the total mass integrated over an area of land ice. The geographical extent of the ice over which the mass was calculated should be described by providing bounds on the horizontal coordinate variables."

OK?

This name is still under discussion.

Answer sophie to alison:
 “OK", but is there a need to extend the last sentence in a manner similar to what is done for “tendency_of_land_ice_mass_due_to_basal_mass_balance”? The later reads "The horizontal domain over which the quantity is calculated is described by the associated coordinate variables and coordinate bounds or by a coordinate variable or scalar coordinate variable with the standard name of "region" supplied according to section 6.1.1 of the CF conventions”

In the land_ice_mass case, the last sentence would then read: “The geographical extent of the ice over which the mass was calculated should be described by providing bounds on the horizontal coordinate variable or scalar with the standard name of “region” supplied according to section 6.1.1 of the CF convention”.



—————
17. land_ice_mass_not_displacing_sea_water
—————

17. land_ice_mass_not_displacing_sea_water (kg)
'The mass of land ice that does not displace sea water if removed, computed
as the ice volume times density. It excludes ice shelves, and components of the
grounded ice sheet that would not displace sea level (for example, in regions
where the ice rests on a bedrock that sits below sea level).'

The name and units are fine.

As per my comments on proposal 16, is this a globally integrated or local
quantity? Again, I don't think the method of calculation belongs in the standard
name definition but it could be reported elsewhere in the metadata.
Answer Sophie: same answer as proposal 16

The name and units are agreed. I suggest the following as the definition:
' "Land ice not displacing sea water" means land ice that would not alter sea level if removed. It excludes ice shelves and grounded ice-sheets resting on bedrock that is below sea level. It includes glaciers, floating ice caps and ice-sheets resting on bedrock above sea level. The quantity with standard name land_ice_mass_not_displacing_sea_water is the total mass integrated over an area of land ice. The geographical extent of the ice over which the mass was calculated should be described by providing bounds on the horizontal coordinate variables.'

Is the bit about glaciers and ice caps correct? I put it in because the definition of land_ice usually refers to them and I want to be as clear as possible about how this name differs from the others.

This name is still under discussion.
Comment Jonathan was:
Glaciers is correct, but not "floating ice caps". You could say "ice caps".
Glaciologists use "ice cap" to mean a thing like a glacier or an ice sheet
but of intermediate size. (Journalists, confusingly, sometimes use "ice cap"
to refer to Arctic sea-ice.)

Comment Karl was:
In the first sentence does "would not alter sea level if removed" mean
the same thing as "would not alter sea level if melted (assuming the
water runs into the sea)"?  If so, I would rewrite the next two sentence
something like "It excludes ice shelves (and any other sort of floating
ice) and it excludes a fraction of grounded ice-sheet mass equivalent to
the mass of any sea water it displaces.  It includes glaciers and a
portion the portion of grounded ice-sheet mass exceeding the mass of any
sea water displaced."  I think mentioning "ice caps" confuses things.
I think this better accounts for the difference in density of frozen
water and liquid water.  Perhaps that density difference isn't large
enough to matter.  If so, then I think you could say "It excludes ice
shelves (and any other sort of floating ice) and the portion of grounded
ice-sheets that is below sea level.  It includes glaciers and the mass
of grounded ice-sheets existing above sea level."

Answer sophie to Alison, and Karl:
I like the idea of removing “floating” and  "ice caps”. I also like Karl suggestion for rewriting the definition, would the following work?

' "Land ice not displacing sea water" means land ice that would not alter sea level if the ice were converted to water and added to the ocean  It excludes ice shelves (and any other sort of floating ice) and it excludes a fraction of grounded ice-sheet mass equivalent to the mass of any sea water it displaces.  It includes glaciers and a portion of grounded ice-sheet mass exceeding the mass of any sea water displaced. The quantity with standard name land_ice_mass_not_displacing_sea_water is the total mass integrated over an area of land ice. The geographical extent of the ice over which the mass was calculated should be described by providing bounds on the horizontal coordinate variable or scalar with the standard name of “region” supplied according to section 6.1.1 of the CF convention.'

—————
18: grounded_ice_sheet_area
—————
18. grounded_ice_sheet_area (m2)
' Total area of the grounded ice sheet, where grounded indicates that the ice
rests on bedrock, and therefore excludes ice shelves.'

Units of m2 are fine. This name brings to mind a recent discussion we had
about two existing names, sea_ice_area and sea_ice_extent and I think there
are two possible approaches for your name.

The first approach would be to use the name as proposed, in which case it
would need to be defined similarly to other X_area names:
' "X_area" means the horizontal area occupied by X within the grid cell.
"Grounded ice sheet" means glaciers, ice-caps and ice-sheets resting on
bedrock. It does not include ice shelves.'
Taking this route means that the name itself does not represent a total. You
would instead need to include a cell_methods of "area: sum" and coordinate
variables with bounds to delineate the area over which the land ice total area
was calculated.

The second approach would be to have a name of
grounded_ice_sheet_extent. You would need to supply coordinate variables
with bounds to indicate the region of the earth over which the ice extent was
calculated. By analogy with the sea_ice_extent name, this quantity would be
regarded as a total area (so you wouldn't need the cell_methods), but it's
important to note that it would be the total area of the grid cells containing
land ice, rather than  that of the land ice itself. Following the sea_ice_extent
definition, the definition of your quantity would then be something like the
following (you may or may not need the wording about thresholds, depending
on how you identify your grid cells):
' The term grounded_ice_sheet_extent means the total area of all grid cells in
which the grounded ice sheet area fraction equals or exceeds a threshold, often
chosen to be 15 per cent. The threshold must be specified by supplying a
coordinate variable or scalar coordinate variable with the standard name of
grounded_ice_sheet_area_fraction. The horizontal domain over which the
grounded ice sheet extent is calculated is described by the associated
coordinate variables and coordinate bounds or by a coordinate variable or
scalar coordinate variable with the standard name of "region" supplied
according to section 6.1.1 of the CF conventions. "Grounded ice sheet" means
glaciers, ice-caps and ice-sheets resting on bedrock. It does not include ice
shelves.'

Either approach would be OK with me, so the choice of which name to use
really depends on the details of how you calculate the area
Answer Sophie: What we want may be slightly different than what you propose
as options. The quantity that we are after is a scalar value (not spatially
resolved) for the total area covered by grounded ice sheet. The number that we
are after can be calculated by multiplying grounded_ice_sheet_area_fraction
with grid cell area and summing that over the entire grid.


Okay, so the quantity you need is the actual area of the grounded ice sheet. Unfortunately, we can't just call it 'area' because the definition of the existing names refers to the area within a grid cell. I think we would have to call it an integral. You say that it is summed over the entire grid, so is this a truly global quantity? If so, we could perhaps call it global_integral_of_grounded_ice_sheet_area by analogy with the existing global_average_sea_level names. If you want to be able to calculate the quantity for individual ice sheets then we'd have to call it something like integral_of_grounded_ice_sheet_wrt_xy, where xy essentially means the horizontal coordinates. You could then supply bounds on the coordinate variables to show the geographical extent of the area integral. What do you think?

This name is still under discussion.
Comment Jonathan was:
Are you sure we can't call it area? It seems to me it's fine to call it that
if the coordinates indicate it refers to the whole world (regarded as one
grid cell) or to some large region (containing an entire ice sheet).

Response Alison to Jonathan was: (Vol 166, issue 13)
I certainly agree that it seems more natural to call it grounded_ice_sheet_area. The definitions of the four existing "X_area" names say '"X_area" means the horizontal area occupied by X within the grid cell'. When I made my previous comments I was thinking about summing areas across multiple grid cells each of which contains a part of the ice sheet, but actually if the 'grid cell' in this case covers the entire area of the ice then it's not a problem. Perhaps the definition should be modified to emphasize that a bit more. Maybe something like:
' "X_area" means the horizontal area occupied by X within the grid cell. The extent of an individual grid cell is defined by the horizontal coordinates and any associated coordinate bounds or by a string valued auxiliary coordinate variable with a standard name of "region".' Does that sound OK?

Response Jonathan to Alison was:
Yes, I think that would be fine for this and other area-integral quantities
like it. It seems fine to allow "grid cell" to mean any region which can be
defined by horizontal coord bounds.

Response Sophie to Alison & Jonathan is:
I much prefer the last definition that is being proposed, which we can use too for 19, below. Thank you so much!

' "X_area" means the horizontal area occupied by X within the grid cell. The extent of an individual grid cell is defined by the horizontal coordinates and any associated coordinate bounds or by a string valued auxiliary coordinate variable with a standard name of "region”.

————
19. floating_ice_shelf_area
———
19. floating_ice_shelf_area (m2)
'Total area of the floating ice shelves, which is the land ice component that
flows over sea water.'

Units of m2 are fine. As with proposal 18, this name could be the one
proposed, or floating_ice_shelf_extent, with the appropriate definitions and
accompanying coordinates and cell_methods attributes as appropriate. Again
the best choice depends on the details of how the area is calculated.
Answer Sophie: same as with proposal 18.

Alison had said:
I agree that proposals 18 and 19 are very similar. If we can agree (18) then this one will follow the same pattern.

This name is still under discussion.

—————
 23. magnitude_of_land_ice_basal_drag
—————
23. magnitude_of_land_ice_basal_drag (Pa)
'The magnitude of basal drag at land ice base.'

I assume this quantity is essentially a deceleration of the land ice flow
towards the sea? Is it perhaps the sum of the stress terms in proposals 12 - 14?
Please can you provide a bit more detail about what is causing the drag?

Answer Sophie: we are still working on the definitions, and will get back to you
:)

This name is still under discussion.

Response Sophie to Alison:
Is it possible to change magnitude_of_land_ice_basal_drag to land_ice_basal_drag?
with the following definition:

land_ice_basal_drag (Pa)
Basal drag is a resistive stress opposing ice flow at the ice bedrock boundary.

—————
 24. tendency_of_land_ice_mass_due_to_surface_mass_balance
—————
24. tendency_of_land_ice_mass_due_to_surface_mass_balance (kg s-1)
'The total surface mass balance flux over land ice is a spatial integration of
the surface mass balance flux.'

The name and units look sensible to me. You would need to use coordinate
variables with bounds to describe the area over which the mass integration is
calculated. We have the existing name
land_ice_surface_specific_mass_balance_flux and clearly this new quantity is
the spatial integration of that one.  Now I'm wondering what is really meant
here by  'surface' - does it mean the lower boundary of the atmosphere or the
upper boundary of the ice (which may have snow lying on it)? I'm guessing the
latter, so we may need to use something like 'ice_surface' in both the new and
existing names to distinguish from our usual use of 'surface'.

I suggest the following wording for the definition:
' "Land ice" means glaciers, ice-caps and ice-sheets resting on bedrock and
also includes ice-shelves. Mass balance means the net rate at which ice is
accumulated. A negative value means loss of ice. "tendency_of_X" means
derivative of X with respect to time. The specification of a physical process by
the phrase "due_to_" process means that the quantity named is a single term in
a sum of terms which together compose the general quantity named by
omitting the phrase. The tendency in ice mass due to the [ice] surface mass
balance is the spatial integral of the quantity with standard name
land_ice_surface_specific_mass_balance_flux. The horizontal domain over
which the quantity is calculated is described by the associated coordinate
variables and coordinate bounds or by a coordinate variable or scalar
coordinate variable with the standard name of "region" supplied according to
section 6.1.1 of the CF conventions.'
We would also need some words about 'surface' or 'ice_surface' as
appropriate. (The wording should also be updated if we decide to drop 'flux'
from the mass balance name as discussed in proposal 4).

Answer Jonathan (10oct2016)
I think surface may be correct in CF terms. For SMB, the snow and ice-sheet
are not usually distinguished. If we mean exactly the mass added to the
ice-sheet component for ISMIP6, beneath the snow pack, then it should be
_at_top_of_ice_sheet_model, as in the interface temperature.

Answer Sophie: yes, this quantity is a spatial integration of the existing name
land_ice_surface_specific_mass_balance_flux, which was in existence in the
CF world before ISMIP6 (with definition "Land ice" means glaciers, ice-caps and
ice-sheets resting on bedrock and also includes ice-shelves. Specific mass
balance means the net rate at which ice is added per unit area at the land ice
surface. A negative value means loss of ice. In accordance with common usage
in geophysical disciplines, "flux" implies per unit area, called "flux density" in
physics.).

I had assumed that the existing definition for
land_ice_surface_specific_mass_balance_flux was what Jonathan describes
above for SMB (or surface mass balance). If you think that it would be better to
change the name (and perhaps, create a new name for what I thought was
SMB), then please let me know. For consistency, would we then have to also
replace “basal” by “_at_the_bottom_of_ice_sheet_model” in items such as 25
below? Or perhaps, we could rename:
land_ice_surface_specific_mass_balance_flux to
land_ice_specific_surface_mass_balance_flux? Any suggestions more than
welcomed!

Reading through Jonathan and Sophie's comments, it seems that I was trying to be more precise than is really necessary. If snow and ice are not generally distinguished for the purposes of surface mass balance then I think we shouldn't change the existing name land_ice_surface_specific_mass_balance_flux. So for this name we would have

tendency_of_land_ice_mass_due_to_surface_mass_balance (kg s-1)
' "Land ice" means glaciers, ice-caps and ice-sheets resting on bedrock and also includes ice-shelves. "tendency_of_X" means derivative of X with respect to time. The surface called "surface" means the lower boundary of the atmosphere. "Mass balance" means the net rate at which ice is added. A negative value means loss of ice. The tendency in ice mass due to the surface mass balance is the spatial integral of the quantity with standard name land_ice_surface_specific_mass_balance_flux. The horizontal domain over which the quantity is calculated is described by the associated coordinate variables and coordinate bounds or by a coordinate variable or scalar coordinate variable with the standard name of "region" supplied according to section 6.1.1 of the CF conventions. The specification of a physical process by the phrase "due_to_" process means that the quantity named is a single term in a sum of terms which together compose the general quantity named by omitting the phrase.'

OK?

This name is still under discussion.

Answer Sophie to Alison:

 “OK” with the definition given by the  last big paragraph above.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cgd.ucar.edu/pipermail/cf-metadata/attachments/20170211/1bc283a4/attachment.html>


More information about the CF-metadata mailing list