nobio and conservation

[aducharne]

To prepare the cleaning of ORCHIDEE with respect to the nobio fractions, we could start by answering the following two questions:
1) what are the active processes and related variables in the nobio fraction?
(in terms of water, C and energy fluxes)
2) what was the reason to add the nobio fraction to soiltile(1) in hydrol?
(soiltile(1) is the soiltile holding the bare soil "PFT")

Anyone with answer elements is invited to report them in this ticket.
Thanks.

[nicolasviovy]

Replying to aducharne:

To prepare the cleaning of ORCHIDEE with respect to the nobio fractions, we could start by answering the following two questions:
1) what are the active processes and related variables in the nobio fraction?
(in terms of water, C and energy fluxes)

Nothing is related to nobio in the C (stomate) part. It is just used in stomate.f90 to renormalize the veget & veget_max in such a way that sum(veget_max)=1

2) what was the reason to add the nobio fraction to soiltile(1) in hydrol?
(soiltile(1) is the soiltile holding the bare soil "PFT")

Anyone with answer elements is invited to report them in this ticket.
Thanks.

[jpolcher]

1) nobio has to do exchanges with the atmosphere but as there is no biology on it this is not needed in STOMATE.

2) The interactions of nobio with hydrol depends on which is considered. For ice one can consider that the soil moisture below is saturated by the meltwater. When nobio represents lakes, then there is no point in doing hydrology there. So we have to better define the nobio classes we wish to take into account and how they should be considered in hydrol.

[luyssaert]

Question 1: for the moment their is no interaction between no_bio and stomate. This view is overly simplistic:

• Urban areas are included in no_bio. Urbanization is one of the most common land cover changes outside the tropics. When urbanization occurs we should be able to move the soil carbon to an (inert?) urban soil carbon pool. Note that in most sub-urban areas there is considerable biological activity (from gardens and urban green). Urbanization is currently not accounted for and the land_cover_change module cannot correctly deal with changes in the fraction of no_bio.
• Underneath lakes there is soil with considerable biogeochemistry processes.

In short, the current distinction between bio and no_bio is too simplistic and should be revised in this light. See also the answer of Jan for Question 2.

[nvuilsce]

NOBIO fraction into hydrol routine

[nvuilsce]

About nobio and water conservation, non-closure of the water budget has been reported from pre-CMIP6 simulations. It has been diagnosed that this non-closure was related to the evapnu (bare soil evaporation) flux. Attached is a brief synthesis (in French) of my understanding of the hydrol code concerning the NOBIO fraction, compared to hydrolc (Choisnel).

To summarize, we propose to remove the NOBIO fraction from the soiltile 1. Consequently, the soiltiles will now represent only BIO fraction. To work correctly (and to conserve water into hydrol), the soiltile fractions have to be normalized, in such a way that sum(soiltiles(i))=1 for i from 1 to 3.

This modification has been commited in the trunk (r3588)

[aducharne]

Small corrections to r3588 in r3593 and r3594

Prepared on July 17, by Nicolas & Agnès:

• We checked all averages involving soiltile in hydrol, and made sure we had averages over vegtot or over the grid-cell depending on the use of the average. For output of storage and fluxes with xios, we need grid-scale averages, to transmitt to stomate, which also works on the bio fraction only, we need vegtot averages.
• For the variables transmitted to thermosoil (mc_layh, etc., shumdiag_perma), we chose grid-scale averages, assuming the thermal properties of the nobio are not explicitely accounted for.

Prepared on November 21, 2016, by Nicolas and Agnès:

• hydrol_tmc_update has not been corrected for the new meaning of soiltile

The two sequences of changes above are explained in ​README_comments_nobio for committ in r3969.

Still to be done:

• In the split of input variables to the soil tiles has been simplified in hydrol_split_soil. The case of ae_ns has been left out for the moment, for dedicated test.
• Check how the nobio must be treated for the thermal properties (mc_lay, and the termla parameters)

[aducharne]

The last two questions are now addressed by specific tickets:
#325 Hydrol_split for ae_ns
#324 Which thermal properties on the nobio fraction