wiki:GroupActivities/Meetings/Retreat20190527

ORCHIDEE retreat 2019

Monday 9H45 - 12H00: block 1

Subject: ORCHIDEE sub-grid scale processes: Multi-tiling for Energy / Water / Carbon

Leaders: Catherine Ottlé, Anne Sofie Lansø

Objectives: Define a concept document for the solution of the multi-tiling implementation and its flexibility ? Organisation:

Talks:

  • The new “multi-energy” budget for ORCHIDEE for forest, lakes, glacials and urbanized areas. Present and discuss principles of the approach, first results, question on how to best implement such multi-tiling in ORC (CO/PP/VB/AL) Ottle_MEB_ORCHIDEE_27May2019.pdf

Short summary of the session:

  • The session was initiated with several short presentations showing ideas on how multi-tiling could be implemented in ORCHIDEE and the efforts already done.
  • CO presented different possibilities on how to share soil and water columns for the sweep of PFTs including nobio surface types.
  • AD presented how to include low and highlands within a grid cell which needed separate water budgets, but not necessarily separate energy budgets.
  • PC showed the importance of having individual tiles of soil and hydrology for the ORCHIDEE-MICT crop version and for the peatlands in MICT.
  • Moreover, a short overview of the plans on how to implement multi-tiling in Hydro-JULES was presented (surface layer, sub-surface layer and open waters defined by physical and numerical constraints, and HRU in the soils to describe the lateral exchanges).
  • ASL presented finally the coding issues which appeared in the way the multi-tiling is presently implemented in the various branches (DOFOCO, MEB and LAK).
  • During the discussion, questions were raised as to whether we need the tiling of the surface, when we in the future likely will be able to increase the horizontal resolution of the model? First of all, for the CMIP simulations even in the future it will be highly unlikely that we will have a resolution below 50 -100 km. Moreover, the nobio fraction is currently very poorly described (lakes, glaciers, urban, etc), which from an energy budget point of view is essential to improve even for higher spatial resolutions.
  • Topography affects the hydrology, and it was stressed it would be ideal to take that into account in the tiling making sure they would be laterally connected. AD mentioned that 2 topographic units should be enough to describe the first order effects.
  • To accommodate open water, peatlands , crops etc. we want more flexibility than the trunk. Contrary to JULES, we do most likely not wish to decouple the PFTs and soils and want to account for their spatial correspondence. The vertical discretization of SOM affects soil thermics and hydrology, thus we cannot just distribute the soil mesh randomly.

List of actions / follow up:

  • We decided to settle a working group in order to move forward on the multi-tiling developments in ORCHIDEE, in closed links with the atmospheric coupling with LMDZ and RegIPSL.


Monday 13H30 - 15H30 : block 2

Subject: coupling with Atmosphere?

Leaders: F. Cheruy, P. Peylin

Objectives : Clarify the principle of the coupling with the atmosphere, identify the priorities in the developments both from the atmospheric and Land-surface point of view. required skills.

Talks:

Short summary of the session:

  • The principle of the implicit coupling between the boundary layer and the LSM has been recalled, with some details on how the surface temperature is solved in LMDZOR.
  • The principle of the multi-layer energy budget (Ryder et al. 2015) has been recalled.
  • It is proposed that it is necessary to take into account the details of the canopy
    • to get better results when comparing LSM simulations to site observations (forced mode)
    • to represent the forest canopy climate .
  • Based on the work of Masson and Seity (2009) an attempt to take into account the drag of canopy elements which crosses several vertical levels in the large-scale atmospheric boundary layer has been presented (implemented in LMDZOR)
  • Discussion on how to better account for the turbulence into the canopy and how the coupling with the large-scale atmospheric (including boundary layer turbulence) model can be done
  • in the future it is expected to use the LAM in construction by T. Dubos et al. for regional simulations. In this configuration the coupling could be revisited and if short time step are used in the atmosphere an explicit coupling could be possible.
  • Which subgrid-scale energy budget do we need in particular to better represent the snow impacts in elevated terrains (but not only)? Is it time to re-write pbl_surface in LMDZ (where the dialog between the atmosphere and the underlying surfaces is done).
  • In the future it is expected to use the LAM developed by T. Dubos et al. (ongoing work) for regional simulations. In this configuration the coupling could be revisited and if short time steps are used in the atmosphere an explicit coupling could be possible.

List of actions / follow up:

  • See previous block (joint action with the multi-tiling): a working group will try to propose a ways to improve the current coupling with LMDz.


Monday 16H00 - 18H30: block 3

Subject: ORCHIDEE governance: Issues and way forward

Leaders: P. Peylin, J. Lathière

Objectives: Update, refine and precise the ORCHIDEE project governance

Organisation:

  • PP and JL oragnized beforehand a survey to collect all “issues” linked to the governance of the ORCHIDEE project (see

Discussion of the problems and objective to find a new strategy for: https://docs.google.com/document/d/1O_DINci1jjQI9EZn427TrZvMK8Vq7elubnb0UGJVzys/edit#heading=h.d3ytce359c9b

  • PP and JL used a presentation to discuss different points ORC_Retraite2018_Gouvernance_small.pdf
    • Vision / common objectives
    • Gouvernence / Organisation
    • Practical Operation of the groupe
    • Share of the code and FAIR use
    • specific actions (projects, recruitement, etc)

Short summary of the session:

  • GOVERNANCE
    • Need to have stronger priorities, stronger dates and delais for each action, more defined meetings.
    • PP recognised that he needs to delegate MORE but in the mean time Everyone should be more pro-active in the different action they commit to lead !
    • ==> We need to function more by "major Task / chantier" that are decided by the group with precise timing.
    • Role and Duration of the coordinator: PP expressed that he is still willing to lead the ORCHIDEE project but that we need a more clear governance scheme with:
      • an agreed duration for the madate (a priori 2 year proposed)
      • regular retreat (every 2 years) to review / change the coordinator / functionning / etc
      • Objectives for the overall coordinator / technicall coordinator (Josefine) / "Tasks" leaders
      • ==> To be precised at the next PROJECT meeting
    • Overall recognition of everyone (CDI and CDD): need to be improved with a diagramm on the website / wiki detail the roles/responsabilities of everyone. This is even more important for CDDs; We need to comunicate to the main entities about our structure and the role of everyone. Maybe a meeting every-year with the ORC group and the lab directors + IPSL + ?? should be organised.
    • Scientific advisory board: Proposed for the Labelisation (already some names selected) with 1 meeting every year; Role to be precise at the next meeting and possibility to invite each orther external people for these meeting. Composition can change over time (but likely for a 5 year period).
    • New code/developments integration: They are all supervised by Josefine. The integration steps are allways more complicated and long than expected! We have progressed but we are still not "efficient" enough. To improve we propose to:
      • Anticipate more with the ongoing development
      • Work more in parallele and not only sequentially (as it was mostly)
      • Find ways to have make sure the original developpers are committed to help
      • NOt all dev should end up in the Trunk.
      • Have potentially a more precise/strict "cahier des charges" ?

  • Practical functionning
    • We need to have more information on the work of everyone and the ongoing developments or planed project, shared between all of us.
    • => Thus we need to have more regular "informal meeting" with presentation of ongoing developments/activities. Maybe 1 out of 2 retreat should be a large meeting with developper's presentation.
    • We need to use more the google document that resume the Projects/Developments?/ect: https://docs.google.com/document/d/13R22r1fx0JoYKCojG1k3IM2C18RWuFNvxoYKdtacJEI/edit#heading=h.yubx14b801dp (to be completed by everyone regularly)
    • Weekly meeting: Need to fix in advance the agenda (as today), Need to decide at the begining the priorities in terms of subject to be discussed; Important to keep the length to 1hour unless mentionned before.
    • Project meeting: Once a month. We should try to alternate between Jussieu and LSCE. Next one in July: try to clarify some point of the governance and discuss again the vision document (PP will send a reminder with information)
    • "Task / activities reporting": we lack of reporting on the ongoing tasks/actions. We need to be better organised with reporting every month on the different tasks..
    • Some request from CDD to have speicif meeting (like "LMDZ bocal") on the technical aspect of using/debugging ORCHIDEE. Should we put in place a "forum" online. Suggestion by Martin Menegoz with a specific tool (to be investigated by PP)
  • Share of the code: We mainly discussed the Integration of external collaborators (special case for China's big groups): we discuss potential issues linked to the emergence of new groups in China with potentially many model developpers (Sushi, Wei li, Shilong, Yue, ....); Although the subject is not close we temptatively agreed that:
    • Either the person connect "regularly" or at least once in a while to the Tuesday project meeting to report on the use/dev of ORCHIDEE to maximize synergies and avoid conflict of interest
    • OR the group is tightly connected to one person of the core ORCHIDEE team who become responsible to report regularly on their activity
    • We have to precise a bit more the right and obligation of each group with respect to the Trunk! (to be refined in the coming months)

List of actions / follow up:

  • Project meeting early July 2019: we need to discuss the VISION and the overally governance; PP will propose a little document with that precise roles of coordinator, duration of the mandates, etc.. To be organised in Jussieu for alterance.
  • PP send a reminder to everyone to fill the google docuement with projects/people involvement.
  • After the Summer project meeting (early september): We need then to define organise the main "TASKS / EFFORTS" with some responsible and time line for reporting to the group.
  • Wiki / Site Web to be filled; Create a page dedicated to the Organisation with the role of everyone including CDD, with also the main Task that are ongoing.. Volunteer for that are needed ?
  • Scientific advisory board: in the context of the labelisation a list of people was proposed! We can modify it later and enlarge for the first meeting; We proposed to have yearly meetin; First one to be organised by PP.
  • Renvoyer le lien (PP) pour la mise à jour des projets/personnes autour de ORCHIDEE: a faire pour Juillet.


Tuesday 9H00 - 11H30: block 5

Subject: Photosynthesis and Phenology

Leaders: F. Maignan, N. Viovy

Objectives:

Present recent works to a larger community that may not be aware of them.

Discuss how recents developments links with CN/-CAN, BVOC and can be integrated in the coming year(s).

Discussion on how to merge these developments with CN/-CAN

Organisation:

Short summary of the session:

  • Yuan presented his work on the impact of diffuse light on GPP. This includes both taking into account the quality of the incoming light (direct versus diffuse) and the propagation of the light with the canopy, with sun and shaded leaves. The evaluation is made at a large number of FLUXNET sites, the model correctly reproduces the larger GPP observed under diffuse light.

Discussion: We won’t change the radiative transfer that is implemented in ORCHIDEE-CN-CAN, but there is an interest in introducing sun and shaded leaves. This is also important for BVOC.

  • Marc presented a new phenology model for temperate and boreal evergreen needle-leaf forests, including onset and senescence, based on observations from the RENECOFOR network and different leaf age studies. Marc also presented his work on the coordination theory, where the leaf optimal N is the one where the rubisco-dependent assimilation is equal the electron transport rate-dependent one.
  • Philippe C. presented the work of Xiuzhi Chen, who has introduced a phenology scheme for tropical evergreen forests, evaluated at sites in the Amazon forest and over the whole basin. The idea is to accelerate the shading of the oldest leaf age class and to increase the NPP allocation towards the youngest class, at the onset of the dry season. The favored climatic driver is thus the shortwave incoming radiation. Philippe also presented the work of Devaraju on aerodynamic roughness and proposed a reparametrization of gs based on Lin et al. (2015).
  • Discussion: SL stresses the fact that on the mld-term we should rather go towards a generic phenology model, rather than ever more specific ones. FM says that Natasha expressed the same goal a few years ago (see for example the co-limitation model of Jolly et al. (2005)).

List of actions / follow up:

  • FM has agreed to animate an action whose goal is to introduce sun and shaded leaves in the photosynthesis module of ORCHIDEE-CN-CAN.
  • When? ORCHIDEE-CN-CAN is the new TRUNK and is working reasonably well at global scale from a GPP point of view.


Tuesday 11H30 - 12H30

Subject: Frontier subject (initially open): Urban PFTs (and then continuation of the Governance section)

Objectives:

Short summary and actions (Jan Polcher)

  • Following the presentation of Simone we decided to set-up a group to think about implementing an urban tile in ORCHIDEE.
  • These are still early days and thus we will need to start with some brain storming regarding the best approach to be followed for our model and what would be the first applications which motivate this effort.
  • For the moment we would like you have your expression of interest so that after the summer vacations we can convene a meeting to discuss this evolution in more details. Please send us an e-mail (Simone and Jan) if you wish to be kept informed about this effort and how you would like to participate.


Tuesday 13H30 - 16H30 : block 6

Subject: Enhancing the traceability, documentation, use and estimation of parameters

Leaders: S. Luyssaert, M. McGrath?

Objective: Agree on a plan to integrate parameter traceability into ORCHIDEE (e.g., which variables were taken directly from the literature, which values were tuned and why)

Organisation:

  • Overview slides to show how the different parts of this session fit together (MATT, 5 mins)Mcgrath_ORCHIDEE_retreat_parameters.pdf. Why have we decided to organize the session this way? Logical ties between presentations
  • Current situation of parameter optimization: The needs for a Bayesian calibration to define always physical range of variations and parameter uncertainties. (PHILIPPE, 5 min) Peylin_Data_assimilation.pdf; Problem with creating uncertainty in parameters due to optimizing against only one or two datasets and implicitly incorporating biases (TASH, 5 min) nmacbean_ORCHRetreat_May2019_parameters.pdf
  • Relatively easy enhancements on top of what we have now (30 min + 5 min for urgent questions Outline of a database to enhance the traceability of ORCHIDEE’s model parameters (SEBASTIAAN, 5 min) Luyssaert_Parameters.pdf
  • Changes to speed-up the model to enhance parameterization (20 min + 5 min for urgent questions) How can we speed-up ORCHIDEE? How can we improve parameter understanding with model speed-ups? (MATT, 5 min) Mcgrath_ORCHIDEE_retreat_parameters.pdf

Short summary of the session:

  • The session was divided up into two main sections: presentations and discussion. The presentations were divided up into three main sections: what is currently being done, ideas for short-term items which could improve parameter documentation and transparency, and a few long-term items related to parameters which will require significant effort and funding. The idea was to focus discussion around the short-term items.
  • Philippe, Tash, and Guillaume presented on efforts currently being made regarding parameter optimization, as well as raising questions to be considered when optimizing.
  • Sebastiaan and Jan presented two ideas to improve parameter documentation, touching on two different aspects. Sebastiaan discussed the need for documenting the origins of parameters, developing some kind of database that stores parameter values with metadata indicating the origin and scripts which update parameter values from existing databases as those databases themselves are updated. Such a database could be converted into a run.def through the use of an additional script, and becomes useful when multiple people are attempting to parameterize simultaneously. Jan introduced a method using XML files that could be formatted as run.defs. Such script would allow to do the quality control on the parameters before they enter the model. The approach Jan proposed would be a very nice last step of the traceability Chain presented by SL.
  • Matt introduced thinking about ways parameters could be improved if the model speed was significantly increased. For example, if emulation of certain parts or the entire model include parameters as inputs. Such speed-up would allow for more thorough sampling of parameter space, in part because the nutrient pool spin-ups could be carried out more efficiently.
  • Aude and March showed how some of the model parameters could be replaced by traits which could help to make use of trait-trait relationships and the coordination theory for photosynthesis as well as for phenology.

List of actions / follow up:

  • The presenters had the impression that the ORCHIDEE group does not experience a sense of urgency related to the ORCHIDEE parameters. No immediate actions will be taken.
  • Matt submitted a parameter-related proposal, Aude and Marc are preparing parameter-related proposals. Philippe, Vlad and Tash will continue their work on parameters optimisation.
  • Tash is setting up a sensitivity analysis of the ORCHIDEE model
  • Sebastiaan will look for a way to start developing parts of parameter traceability chain.
Last modified 3 years ago Last modified on 06/17/19 12:33:24

Attachments (23)