mismatched land/sea/lake masks with atm and wave model

[JessicaMeixner-NOAA]

WW3 passes z0 surface roughness to the atmosphere model. When using the global 30 min wave grid, there are some current issues with this:

• The ocean/land mask does not agree with the ocean model's land mask. So for example, in the wave model the Mediterranean is masked out. Currently the atmosphere model now gets 0s as the surface roughness there which is not ideal. This will be improved on when the atmospheric model uses fractional land/sea masks and has a matching land/sea mask with the ocean/ice models. WW3 will then also need updates to use the same tripolar grid as the ocean and ice models, which is ongoing development work for WW3.
• Once the ocean/land mask is consistent between all the models and WW3 is on the tripolar grid, there is still an issue that the lakes should have a surface roughness that is not-zero as it is currently. Again, I believe moving to the fractional land/sea mask and the addition of a lake mask should help with this.
• Once the ocean/land mask is consistent between all the models, we should double check the surface roughness values with partial ice coverage so that if the wave model does not calculate waves because there is too much ice that a reasonable surface roughness gets calculated.

[SMoorthi-emc]

Jessica,
Regarding the second bullet please note that currently wave model only updates Z0 over ocean points; i.e. only when oceanfrac > 0.0. This has no impact on lakes. Wave model can set non-zero values for Z0 over ocean points.

[JessicaMeixner-NOAA]

@SMoorthi-emc This is why I have the second bullet:

This is from an earlier run, so I should get an upto date comparison, but the left is from the fully-coupled and the right is no-wave coupling. This is what FV3 is outputting for surface roughness, so you can see on the left where the land/sea mask mismatch because the wave model masks out the Mediterranean and that the lakes in Africa are now zero and not as they were before.

[junwang-noaa]

So for the second item, from the code, there should be no impact on lakes since atmosphere only takes Z0 when oceanfrac >0. This looks to me is an initialization issue. You should use the oro data with lake fraction defined (1, or 0 for lake points). For non-fractional grid, if lake fraction is defined, oceanfrac will not cover lake, so atm can use its own z0 for lake points. For the first bullet, the only issue it may happen could be that atmos thinks it's ocean while ww3 think it's land. So what values will ww3 set for z0 for land, if we know this information, say, it's undefined value, it can be used in atmos to refine them to a default reasonable value, not sure if we need the whole Mediterranean to be masked in ww3, it could have larger impact than setting z0 on the mismatched points.

…

On Wed, May 13, 2020 at 9:24 AM SMoorthi-emc ***@***.***> wrote: Jessica, Regarding the second bullet please note that currently wave model only updates Z0 over ocean points; i.e. only when oceanfrac > 0.0. This has no impact on lakes. Wave model can set non-zero values for Z0 over ocean points. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI7D6TKSWXW25EESKUX3CCLRRKNP3ANCNFSM4M7WLRGA> .

[JessicaMeixner-NOAA]

The updated wave grid is a big improvement.

@junwang-noaa or @SMoorthi-emc can you point me to the different oro data with lake masks you think we should be using for the benchmarks? Would this have any other effects, such as needing to re-run chgres?

[AvichalMehra-NOAA]

Excellent! Thanks, Avichal.

…

On Thu, May 14, 2020 at 9:44 PM Jessica Meixner ***@***.***> wrote: The updated wave grid is a big improvement. [image: Screen Shot 2020-05-14 at 9 43 42 PM] <https://user-images.githubusercontent.com/22549231/82002527-fe27c200-962b-11ea-97b7-169f6d95815a.png> @junwang-noaa <https://github.com/junwang-noaa> or @SMoorthi-emc <https://github.com/SMoorthi-emc> can you point me to the different oro data with lake masks you think we should be using for the benchmarks? Would this have any other effects, such as needing to re-run chgres? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AJ6ZTFDI2TSAQGXB6S2HTWLRRSNAFANCNFSM4M7WLRGA> .

-- Dr. Avichal Mehra Chief, Dynamics and Coupled Modeling Group Modeling and Data Assimilation Branch NOAA/NWS/NCEP/EMC Ph. 301-683-3746 Fax: 301-683-3703 [email protected]

[SMoorthi-emc]

I have C384 orography from Shan at "/scratch1/NCEPDEV/global/Shrinivas.Moorthi/noscrub/oro_lake/C384"

…

On Thu, May 14, 2020 at 9:44 PM Jessica Meixner ***@***.***> wrote: The updated wave grid is a big improvement. [image: Screen Shot 2020-05-14 at 9 43 42 PM] <https://user-images.githubusercontent.com/22549231/82002527-fe27c200-962b-11ea-97b7-169f6d95815a.png> @junwang-noaa <https://github.com/junwang-noaa> or @SMoorthi-emc <https://github.com/SMoorthi-emc> can you point me to the different oro data with lake masks you think we should be using for the benchmarks? Would this have any other effects, such as needing to re-run chgres? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALLVRYVU4RGSIFQFNTUNMKDRRSNADANCNFSM4M7WLRGA> .

-- Dr. Shrinivas Moorthi Research Meteorologist Modeling and Data Assimilation Branch Environmental Modeling Center / National Centers for Environmental Prediction 5830 University Research Court - (W/NP23), College Park MD 20740 USA Tel: (301)683-3718 e-mail: [email protected] Phone: (301) 683-3718 Fax: (301) 683-3718

[ligiabernardet]

@mzhangw Can you please check if this the same revised oro data that you are working with for introducing Flake in CCPP (with Yihua Wu)?

[JessicaMeixner-NOAA]

Using the orodata from here: /scratch1/NCEPDEV/global/Shrinivas.Moorthi/noscrub/oro_lake/C384
in addition to the addition of lake masks, there are minor differences along coastlines in the oro data:

On the left is the new oro data with lake fraction the right is the old oro data no lake fraction. The surface roughness in the atm model over lakes is back to previous values:

and surface temperature on lakes also have changed:

[SMoorthi-emc]

Now the lake temperature is correct! This is what I have been saying for while now.

…

On Fri, May 15, 2020 at 10:58 AM Jessica Meixner ***@***.***> wrote: Using the orodata from here: /scratch1/NCEPDEV/global/Shrinivas.Moorthi/noscrub/oro_lake/C384 in addition to the addition of lake masks, there are minor differences along coastlines in the oro data: [image: Screen Shot 2020-05-15 at 10 55 57 AM] <https://user-images.githubusercontent.com/22549231/82064367-abd1ba00-969a-11ea-99cb-33d447f67f3e.png> On the left is the new oro data with lake fraction the right is the old oro data no lake fraction. The surface roughness in the atm model over lakes is back to previous values: [image: Screen Shot 2020-05-15 at 10 51 26 AM] <https://user-images.githubusercontent.com/22549231/82064460-cb68e280-969a-11ea-98a3-3d0556a2326c.png> and surface temperature on lakes also have changed: [image: Screen Shot 2020-05-15 at 10 50 14 AM] <https://user-images.githubusercontent.com/22549231/82064434-bf7d2080-969a-11ea-901b-db856e616289.png> — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALLVRYW5MBIFMZKKVRXVGHDRRVKA3ANCNFSM4M7WLRGA> .

-- Dr. Shrinivas Moorthi Research Meteorologist Modeling and Data Assimilation Branch Environmental Modeling Center / National Centers for Environmental Prediction 5830 University Research Court - (W/NP23), College Park MD 20740 USA Tel: (301)683-3718 e-mail: [email protected] Phone: (301) 683-3718 Fax: (301) 683-3718

[junwang-noaa]

Yes, this makes lake temperature correct too as the lake points are not counted in if (ocean fract>0). On Fri, May 15, 2020 at 11:06 AM SMoorthi-emc <[email protected]> wrote:

…

Now the lake temperature is correct! This is what I have been saying for while now. On Fri, May 15, 2020 at 10:58 AM Jessica Meixner ***@***.*** > wrote: > Using the orodata from here: > /scratch1/NCEPDEV/global/Shrinivas.Moorthi/noscrub/oro_lake/C384 > in addition to the addition of lake masks, there are minor differences > along coastlines in the oro data: > [image: Screen Shot 2020-05-15 at 10 55 57 AM] > < https://user-images.githubusercontent.com/22549231/82064367-abd1ba00-969a-11ea-99cb-33d447f67f3e.png > > > On the left is the new oro data with lake fraction the right is the old > oro data no lake fraction. The surface roughness in the atm model over > lakes is back to previous values: > > [image: Screen Shot 2020-05-15 at 10 51 26 AM] > < https://user-images.githubusercontent.com/22549231/82064460-cb68e280-969a-11ea-98a3-3d0556a2326c.png > > > and surface temperature on lakes also have changed: > > [image: Screen Shot 2020-05-15 at 10 50 14 AM] > < https://user-images.githubusercontent.com/22549231/82064434-bf7d2080-969a-11ea-901b-db856e616289.png > > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > < #92 (comment) >, > or unsubscribe > < https://github.com/notifications/unsubscribe-auth/ALLVRYW5MBIFMZKKVRXVGHDRRVKA3ANCNFSM4M7WLRGA > > . > -- Dr. Shrinivas Moorthi Research Meteorologist Modeling and Data Assimilation Branch Environmental Modeling Center / National Centers for Environmental Prediction 5830 University Research Court - (W/NP23), College Park MD 20740 USA Tel: (301)683-3718 e-mail: ***@***.*** Phone: (301) 683-3718 Fax: (301) 683-3718 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#92 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AI7D6TJFBMF36HRLYFSHBSTRRVK6PANCNFSM4M7WLRGA> .