AMMA icon
  ALMIP-2 Home
  Executive Summary
  Experiment
  Calendar
  Participants
  Input Data
  Output Data
  Results
  FAQ
  Meetings
  Presentations
  Publications
  Contact
  ALMIP (Phase 1)
  ALMIP2 Workshop
  AMMA Related:
  AMMA CNRM-GAME
  AMMA International
  AMMA US
ALMIP title banner

(last updated: Oct. 11, 2013)

FAQ: Frequently Asked Questions

Emails sent to the participants will be posted here. In addition, certain emails from participants which identify problems or common questions will also be posted (anonymously). For such emails, questions posed by a participant are highlighted using Q: while comments are represented by C:, and responses by us will be indicated using A:.

Email Sent (all participants): June 18, 2012

Dear ALMIP Participants,

    First of all, we would like to extend our sincere thanks to all those who have accepted to participate in ALMIP2, your time and energy are greatly appreciated and we hope that you will find this endeavour scientifically stimulating and worthwhile. We have received an excellent response with well over 20 different research groups from the international community involved. We are sending this first global email to convey some information and updates: we will periodically send out emails. If you receive this multiple times or know of anyone else who might like to be on the list (or if you wish to be removed from the list), please let Aaron (aaron.boone@meteo.fr) and/or Christophe (christophe.peugeot@ird.fr ) know.

  • As some of you have noticed, there are no ECOCLIMAP files for 2008. Indeed, originally we planned to end the experiments in 2007. But, as it turns out, most of the AMMA observations have extended beyond the SOP (2005-7), so we decided to extend the experiment. This was not anticipated by the people responsible for ECOCLIMAP: but they have graciously accepted to extend the dataset for ALMIP2. Therefore the ECOCLIMAP files for 2008 are currently being prepared and should be ready by the end of June: we will inform everyone when they are available on the ftp server.

  • We will reformat ECOCLIMAP data to have values interpolated to daily time steps (as opposed to decades as is currently the case). The interpolation is linear. We recommend that modelers use this if possible, but if it poses problems (it likely does for some groups), then modelers can use such data in the usual way for their respective platforms.

  • A few of you have asked about the Donga (high resolution OPTIONAL mesoscale experiment). Indeed, the forcing data has not been completed yet. Again, this experiment is optional, but we are currently seeing if this data can be ready in time for modellers. We'll send an update as soon as we have enough information to make a decision.

  • Local scale forcing and initial data is just about finalized. When it is ready, we'll put it on the ftp server and inform you. Indeed, we felt it most important to prioritize the mesoscale data owing to the more significant technical aspects involved. We will email you when this data is on the ftp server, likely within the next week or so.

  • For those running hydrological models, the needed data (GIS, topographic data, etc...) is now on the ALMIP2 server under data/hydrology/ on the ftp server. See the Readme_hydrology file for information. Note that limited discharge data is available for hydrological models to do some limited calibration: in fairness, we mention this to everyone, including land surface modellers. But note, LSM model runoff will be routed by us for comparison to discharge, so LSM models do not necessarily need to perform such a calibration. But, if a LSM modeller deems this data to be potentially useful, it may be accessed on the ftp server. But we ask that any modeling groups using observed discharge to tune or calibrate their models please inform us. Finally, please note also that the ALMIP Overview document has been updated to include specifics on the hydrological simulations. It is available at

    http://www.cnrm.meteo.fr/amma-moana/amma_surf/almip2/docs/ALMIP2_overview.pdf

  • We are tentatively planning to hold a Workshop in Toulouse in Jan-Feb, 2013. We currently anticipate a full 2 day meeting. We have avoided some of the larger geophysical meetings (AGU for example) and the annual AMS meeting. Currently we are planning for sometime during the week 4-8 February. If someone has a major objection or knows of a conference we have overlooked, please let us know. We will be sending further details on this soon.

  • The 3rd International AMMA Conference will be taking place this July in Toulouse. Some ALMIP1 and 2 related work will be presented. If you'd like more information, visit the AMMA Conference website at http://amma-conf2012.ipsl.fr/home

    Finally, once you upload your mesoscale runs to the server, please let us (Aaron and/or Christophe) know. Also, we are currently developing the sanity checks routines for LSM results.

June 22, 2012: Questions regarding model output


Q: 1) The netCDF template says that times should be reported at the start of the output interval; but the table following says that times should be reported for the end of the current time step. Could you clarify?

A: Indeed, it should be at the end of the timestep


Q: 2) I can't find RootDist in the ALMA tables. Do you mean cumulative root fraction counting from the top soil layer down, or the bottom soil layer up?

A: the cumulative root fraction counting from the top soil layer down


Q: 3) Several variables are described as "average over past output interval". By "past", do you mean that you want the average over the current time step, or over the previous one?

A: Averaged over the *current* time step


Q: 4) Shouldn't DelSoilHeat, DelPond and DelIntercept properly be described as "change over time step" rather than "average over time step"?

A: Yes, It is the change, not the average


Q: How should we be initializing prognostic variables such as soil layer temperatures and liquid water contents? Do you have recommmended initial values for each point, or should we be doing spin-up runs?

A: For the mesoscale runs, we recommend spin up. See the Overview Document for further details. For the local scale, we will provide values (they will be diffused to participants soon)

Email Sent (all participants): July 20, 2012

Dear ALMIP2 Participant,

We have 4 important updates.

1) ECOCLIMAP2 has been extended to 2008. But, the entire chain was updated and now the LAI values are larger than before, and are, in fact, improved. So, data was reprocessed from 2005-8 for all 3 mesoscale sites. This means that for those of you who have done the simulations, we now kindly request that you rerun. We are sorry for the inconvenience, but it is the best way to maintain consistency and again, the new LAI values are improved.

2) Because of (1) and because we know everyone is busy, we have extended the deadline to the end of August. At that time, we'll process results (and do sanity and budget checks, etc.) and iterate with groups if we find problems (as mentioned previously). However, if you already have results, feel free to upload them to the ALMIP2 server and checks will be made starting in mid-August (for those of you who have results before the end of the month).

3) Local scale data is still being finalized (for those of you who have already inquired about this data). Most of the local scale data is ready, however, we prefer to release data for all of the sites at the same time to ensure that the same gap-filling and processing controls have been applied. Since we have extended the deadline, we hope that this is acceptable and we should have ALL of the data by early August.

4) Workshop details will be fourthcoming in Aug/Sept. Tentatively, we are aiming for a 3-day workshop in Toulouse, France, in Feb, 2013.

Email Sent (all participants): Aug. 8, 2012

Dear ALMIP2 Participant(s),

    Some of you are preparing output files, so here is some information concerning a few clarifications, and a few additional output variables.

1) The variable "DelSoilMoist" has been added to the outputs to avoid the need to include the initial time step value for outputs (thanks to the team at ECMWF for inquiring about this variable). This has been updated on the ALMIP2 webpage (and will be visible on Aug 9) at

http://www.cnrm.meteo.fr/amma-moana/amma_surf/almip2/output.html

The corresponding new NetCDF header information is:

    float DelSoilMoist(time, y, x) ;   
        DelSoilMoist:units = "kg/m^2" ;
        DelSoilMoist:long_name = "total soil moisture content change" ;
        DelSoilMoist:associate = "time y x" ;
        DelSoilMoist:missing_value = 1.e+20f ;

And the number of times in the given example has been changed from 17521 to 17520 (for a non-leap year). As defined by the ALMA convention (see http://www.lmd.jussieu.fr/~polcher/ALMA/), outputs for fluxes/tendencies(storage terms) at time "t" correspond to the values averaged/accumulated over the interval from time=t-dt to time t, so for example, at the first time step, this corresponds to from time(t=0) (midnight) to time(t=1)=1800 s. Instantaneous variables (prognostic) are to be reported at time t.

2) Please NOTE that it is assumed that (neglecting snow/ice phase changes):

Evap = ESoil + TVeg + ECanop + EPondInt

If you have another source of evaporation, please let us know and we'll define a convention for adding them to the outputs so that we can close your evaporation budget.

3) Your water budget will be checked using (neglecting snow/ice phase changes):

 DelSoilMoist +  DelPond + DelIntercept = DT*(Rainf - Evap - Qs - Qsb)

However, some of you might include addition storage terms or exchanges which don't cancel and must be accounted for. If this is the case, please let us know and we'll define a convention for adding them to the outputs.

4) SWI or soil wetness index (and similar metrics) would be quite useful for us to examine. If you have not yet written your outputs, we would like you to either

  1. provide an additional file
  2. put the values into your output

which contains the soil porosity, field capacity, and wilting point. To be general, please define them as 3D variables (depth,x,y) as some models can model variable texture profiles. An example addition to the ALMIP2 output template is:


    float SoilSat(lev, y, x) ;        
        SoilSat:units = "m^3/m^3" ;
        SoilSat:long_name = "soil porosity" ;
        SoilSat:associate = "lev y x" ;
        SoilSat:missing_value = 1.e+20f ;
    float SoilWilt(lev, y, x) ;        
        SoilWilt:units = "m^3/m^3" ;
        SoilWilt:long_name = "soil wilting point" ;
        SoilWilt:associate = "lev y x" ;
        SoilWilt:missing_value = 1.e+20f ;
    float SoilFC(lev, y, x) ;        
        SoilFC:units = "m^3/m^3" ;
        SoilFC:long_name = "soil field capacity" ;
        SoilFC:associate = "lev y x" ;
        SoilFC:missing_value = 1.e+20f ;

Note that Field Capacity here corresponds to the minimum volumetric water content value at which the evapotranspiration is not under water stress.

5) Owing to the size of requested outputs, we require tile-based schemes to report grid box average outputs only. However, if they wish to output values for each tile, we ask that they store this on their local servers for now (as such files might overwhelm the ALMIP2 server!): contact us and we can discuss how to analyze this data on a case-by-case basis for now.

6) The output file naming convention is currently:

SVAT_DOMAIN_PRECIPITATION-FORCING_XXXX_YEAR.nc

where SVAT=the scheme acronym, DOMAIN=Niger, Mali or Oueme (note that Oueme is used in place of Benin since both the Oueme and Donga basins are in Benin, and runs for the Donga sub catchment might still be optionally done), PRECIPITATION-FORCING=T (Theissen), L (LaGrangian-krigged) or R (radar-based: potential optional run), and YEAR=2005 to 2008. Note that an additional tag, XXXX, can be defined by the participant if they do additional runs with different options, etc. For example, the control run output file for ISBA using LaGrangian rainfall input for Oueme (Benin) would be named:

ISBA_Oueme_L_2007.nc

And an additional run using the multilayer soil scheme could be named

ISBA_Oueme_L_DIF-Soil_2007.nc

etc. If you have already uploaded your files,we'll rename them for you.

    A final reminder: you do NOT need to report variables which your scheme does not model. But bear in mind that we will check your energy and water balances, so you must report the minimum number of variables for us to do these computations. We'll be doing these checks starting in the latter part of this month, into Sept (as results arrive). And again, all of the above additions will be visible on the ALMIP2 output webpage starting on Aug. 9.



Aug. 9, 2012: Comment regarding model output

C: CLM4 has a groundwater component so that there is variable called "WA= groundwater storage".

A: Indeed, soil moisture in most models refers to the near surface part of the soil column. In order to account for separate deep water or aquifer reservoirs, the variable added is DelGroundWat (kg m2). So, the water budget balance Eq is now:

 DelSoilMoist +  DelPond + DelIntercept + DelGroundWat = DT*(Rainf - Evap - Qs - Qsb)

Note that for schemes where the difference between DelSoilMoist and DelGroundWat is ambiguous, they can lump all column water storage changes into DelSoilMoist and not report DelGroundWat.




Aug. 20, 2012: Comment regarding the vegetation parameter Rsmin

C: It has been brought to our attention (by the CLASS group: thank you!) that in one of the meso-domains (the northern part of Mali) Rsmin (minimum stomatal resistance) ECOCLIMAP values drop below 30 s m-1...which is not realistic. In fact, the problem seems to be where Rsmin is essentially "undefined", there is (virtually) no vegetation cover (veg==>0, veg=vegetation cover fraction, LAI==>0 for the pixels where Rsmin < 30) So, after a closer inspection of the files, we propose that you use:

Rsmin*(lon,lat) = max[30, Rsmin(lon,lat)]

Since the surrounding values (and minimum values in ECOCLIMAP) are around 30 s m-1. But please NOTE, this "fix" should have virtually no effect on the simulated Evapotranspiration (in theory) since veg and LAI are nearly 0, so the value of Rsmin should be somewhat arbitrary. But indeed, by limiting Rsmin as Rsmin* (above), we can avoid any spurious numerical effects which could occur in some models.



Email Sent (all participants): Aug. 21, 2012: Modification to Output NetCDF dimensions

Dear ALMIP2 Participant,

We have just updated the ALMIP2 output template in terms of the *dimensions* definitions. We have done this in order to make the output more standard to facilitate the use of freely available NetCDF processing software. The new example template is attached, and will be visible on the ALMIP2 webpage starting on Aug. 21. Again, the dimensions (lat, lon, and time) are more standard: the rest of the variables are unchanged.

But please note: if you have *already* written your output files, we will accept them using the old template (we will transform the dimensions "in house")...so you do not need to rewrite the outputs in this case.

Aug. 23, 2012: Questions regarding model output variable VegT


Q: There is another variable named VegT and according to the template, it is only one layer. Actually, our model has two layers' temperature for vegetation. So would you please tell me which layer's temperature is preferable to submit, the upper layer (trees), the lower layer (grass) or the average of these two layers?

A: Indeed, there is no accomodation for this. We'll leave this up to you: if you feel it is useful to provide both, then perhaps add a dimension to VegT and report both:

        float VegT(time, lev_veg, y, x) ;

This implies you'd also add a coordinate: let's call it lev_veg so (e.g. for Oueme, Benin for 2005, for a model using 3 soil layers):


dimensions:

        lat = 25 ;
        lon = 28 ;
        time = UNLIMITED ; // (17520 currently)
        lev = 3 ; 
        lev_veg = 2; 

where it could be defined (in a manner consistent with lev) as

        int lev_veg(lev_veg) ;
                lev_veg:long_name = "level in the vegetation canopy" ;
                lev_veg:positive = "down" ;
                lev_veg:units = "level" ;

in your output NetCDF output file. If, however, you feel this is not important, then average the 2. We'll leave this up to you.

Aug. 30, 2012: Uploading Results to the ALMIP2 ftp server at IPSL


C: Please note, the password and login for uploading data are not the same as those for downloading. Please contact us for the logins and passwords.



Email Sent (all participants): Sep. 6, 2012: Update

Dear ALMIP2 Participant,

We have just a few quick updates (as several of you have asked some related questions recently):

  1. Local scale forcing and input parameters will be ready by the end of September. We would like you to report local scale results by December 1, thus, you will have 2 months. We will also provide additional information concerning the local scale runs at the end of Sept.

  2. A participant has found some errors in the Carbon variable units in the ALMIP2 output template. The corrected template will be visible on the ALMIP2 website starting tomorrow, but the corrections are also added here as an attachment.

  3. The upload login and password are different from those used for the download. The upload login is alimp2in For security reasons, please contact us directly for the password when you are ready to upload your results.

  4. We are working on finalizing the updated ECOCLIMAP2 files. Again, the runs with the new ECOCLIMAP2 inputs are not mandatory, but we we strongly encourage you to rerun using them. They cover the period from 2005-2008 (whereas the default parameters were only from 2005-2007). We hope to have the parameters ready next week: we will email you when they are ready, and the deadline for these results.

  5. Please let us know if you are doing additional runs using dynamic vegetation. Several groups will run their models using dynamic vegetation as an option: if you have this option, we would like to encourage you to do these runs (for the local and meso scales).

If you have any questions, please don"t hesitate to ask. Thank you...

Sincerest regards, The ALMIP2 Working Group

Email Sent (all participants): Sep. 18, 2012: Update-Output and Documentation

Dear ALMIP2 Participant,

We are now receiving results and beginning to do so-called sanity checks and analysis. We have 2 items requiring attention as we start to receive results:

1) We would like to emphasize that we would like you to use the ALMIP2 NetCDF template for model outputs (names, sign conventions, grid etc...)...we have a great deal of data to go through so your cooperation on this point is quite important for us to make a timely analysis. If you have any difficulties, please let us know. If your results don't respect the template (save for small inconsistencies which are easily fixed), we will likely ask you to reformat your outputs.

Documentation: We would like to request a short summary of your ALMIP2 model configuration. Of special interest, are parameter values or options which you modified or had to create owing to a lack of specification from us (compared to your default configuration) for ALMIP2. We are really asking for a short informal document containing key references. You can also include the most recent/relevant paper(s) or technical note(s)...but please limit this to the most important 1 or 2 if possible. For this, we ask that you go to the ALMIP2 ftp site, and under

/documents/Models

Please make a directory with your model acronym and upload the relevant documents. If you could do this within the next week or two, it would help us immensely.

If you have any questions, comments, concerns..please don't hesitate to contact us!

Email Sent (all participants): Oct. 05, 2012: ECOCLIMAP2_v2 runs:Update

Dear ALMIP2 Participant,

First, we wish to thank those of you who have uploaded results. We are starting the sanity checks and will begin to send reports to participants within the next week or so.

And, as mentioned in a previous email, we have re-processed the ECOCLIMAP2 data in order to include 2008. But, because the processing is different, we have made a new version of ECOCLIMAP2 data for 2005-7 also. We would like to ask participants who have the time and resources, to do an *additional* set of runs (note, not mandatory) using the new ECOCLIMAP2 files. This simply means repeating the runs using the exact same methodology as was done for the mandatory mesoscale runs, except using the new ECOCLIMAP2 files and running for an additional year (2008). Experiment details are given at the end of this email. The data is now available on the ALMIP2 web server. The file directory and name convention:


data/forcing/meso/

CASE/ALMIP2_ECOCLIMAP2_v2_YY_CASE.nc

YY   = 05, 06, 07, 08
CASE = Benin, Niger, Mali

Please note that we strongly encourage you to do these runs because it means we have an additional year, however, we realize that some of you will not be able to do this because of manpower, time, etc issues, so this is why we are making these runs *optional*. We will still use the *mandatory* runs which you have/are done/doing as the baseline results. We are operating under the assumption that once you have done the hard work of setting up your model to use the mesoscale forcings and to process your results, it *should be* straightforward to replace the vegetation parameters and rerun. But, since this might not be the case, we understand if you can't do these additional runs.

The deadline for this second set of mesoscale runs is Dec. 1. If you have any questions or problems, please don't hesitate to contact us.

As a final note, information and data for the local scale runs is forthcoming and we will email as soon as the data are ready, and we'll provide more instructions on how to proceed with these experiments at that time.

=============================================================

Experiment Details - ECOCLIMAP2_v2 mesoscale runs

  • Benin: Rerun the mesoscale runs for Benin from 2005 to (now) 2008 using the LaGrangian and the Thiessen rainfall inputs, for a total 8 yearly output files
  • Niger: (as above)
  • Mali: Rerun using the Thiessen rainfall from 2006 to (now) 2008. But note, LaGrangian rainfall only exists for 2008 for this meso domain, therefore we ask that you initialize the 2008 LaGrangian run for Mali using the state variables from the *Thiessen* run on Dec 31, 2007 at 24h. The reasoning: we wish to avoid treating 2008 as a spin up year. For Mali, most of the rain falls during a short period in the middle to end of summer, so there should not be any kind of dramatic discontinuity owing to the change in input rainfall at the start of 2008. This corresponds to 4 additional yearly output files.

So, the total is 20 additional yearly files for all 3 meso domains.

Email Sent (all participants): Oct. 15, 2012: ftp servers offline during the week of Oct. 22, 2012

Dear ALMIP2 Participant,

Significant refurbishing of the room which contains the Climserv servers and the re-deployment of the network is to take place during the week of October 22. As a result, all of the disks and servers will be stopped during the aforementioned period (more precisely from Monday, Oct. 22 at 8h to Friday, Oct. 26 at 18h Paris time)...i.e. the ftp and storage services will be stopped/offline. So while the server is down, please refrain from uploading your ALMIP2 results. You may download and/or upload data again after the services have resumed at 19h Paris time on Oct 28. We apologize for any inconvenience.

Email Sent (all participants): Jan. 9, 2013: Updates, Workshop Date

Dear ALMIP2 Participant,

First of all, we hope that you have had a pleasant holiday and wish you a happy new year!

The purpose of this email is to give a quick update on the ALMIP2 Workshop (since some of you are possibly planning to attend the EGU, and also we imagine everyone's agendas are rapidly filling in). The dates are

April 15-17 (Monday-Wednesday)

Which is the week after the EGU in Vienna. Please let us know if you plan to attend the ALMIP2 Workshop. We will make an agenda once we get a certain number of responses from you. The basic plan is to set up 3 sessions:

  1. We'd like each group to present their model (concise overview of key features) and possibly some ALMIP2 sensitivity results (optional). Also, any groups who have developed any additional products we are using for evaluation will also be invited to present an overview of their work/methodology.

  2. Then, the working group will present preliminary analysis.

  3. Finally, additional experiments, final analysis and publication strategies will be discussed. Groups are welcome to propose additional multi-model analysis at this point also.

We will be contacting groups individually with mesoscale sanity check results over the next 2 weeks or so. There will be a window provided for those who need to rerun.

So me groups have yet to submit results...a few of you have already contacted us. Please let us know where you stand/if you still plan to submit results and when it's possible for you.

Finally, thank you again for all your hard work...we know that this project demands a significant effort on your part and we are appreciative, and hope that you will find it worth your while once it's completed.

Email Sent (all participants): Feb. 19, 2013: Updates, Workshop Details

Dear ALMIP2 participant,

    We have begun organizing the ALMIP2 Workshop (Monday-Wednesday, April 15-17). Although we have already received commitments from some of you to attend, we would like to ask you now to please confirm your participation at the workshop. This will help us organize the agenda and prepare other logistics for the meeting in a timely fashion. It would be greatly appreciated if you could respond before the end of this week.

    A conference web page is being set up (it will be accessible starting on Wednesday, Feb. 20), and it contains information on hotels and logistics:

http://www.cnrm.meteo.fr/amma-moana/amma_surf/almip2/workshop.html

We have a very limited amount of travel assistance we can offer: if this is necessary for your participation, please let us know as soon as possible, and we can try to see what can be done.

    The tentative agenda will be:


Mon, April 15
  9:15-10:00 Welcome at the CIC at Météo-France, coffee will be offered
 10:10-12:30 Overview of ALMIP2 Objectives, presentations of the sites, data, key processes
 12:30-14:00 Lunch
 14:00-18:30 Presentations by the Participants (we will give more details soon, and the length of the 
             presentation will depend on the number of participants) a presentation of your model, any 
             difficulties encountered in doing the ALMIP2 runs, explanation of any additional sensitivity 
             tests, etc...

Tues., April 16
  9:15-18:30: Presentation of preliminary results (by sites and across sites, by variable, by process,....) 
              - Lunch from 12:30-14:00
 19:00-20:30 Apéritif offered at the CIC

Wed., April 17
 9:15-12:30 Discussion of reruns, Additional simulations, Supplemental Analysis
 12:30-14:00 Lunch
 14:00-16:30 Discussion of follow-on actions (calendar), Publication timetable and strategy

    You may purchase lunch at the Meteo-France restaurant. We will provide coffee break refreshments. Again, we will finalize the agenda once we have heard back from you. If you have any comments, suggestions, questions, please don't hesitate to ask!

Sincerest regards,

The ALMIP2 Workshop organizing committee

Email Sent (all participants): Apr. 25, 2013: Updates, Brief (Post) Workshop Summary

Dear ALMIP2 participant,

    This is just a quick update after the ALMIP2 Workshop (April 15-17). We will be sending a more detailed summary of the Workshop and planned actions in approximately 2 weeks. Some highlights :

  • At the workshop it was decided that models will be given the opportunity to rerun (if errors were found in their simulations or in the input data: see the next item). If a group reruns, the new simulations will replace the original simulations if the same filenames are used. If not, reruns will be considered as additional runs (sensitivity tests, etc...).

  • There was some confusion about forcing levels (for the input meteorological variables) : in the NetCDF, the levels for all variables (wind, temperature, relative humidity...) is given as 10 m. However, on the ALMIP2 web page (Input Data), the forcing level was erroneously stated to be 2m for air temperature and specific humidity. This has since been corrected. If you used 2m in place of 10m, please rerun. We are sorry for any inconvenience this might cause, but the impact of this change is likely significant.

  • The ordering of ECOCLIMAP2-v2 data was found to be reversed in the latitudinal direction compared to the original ECOCLIMAP2 files. If one used the lat-lon coordinates, then this caused no problem. But if one used the fortran index, then the vegetation and soil parameters were potentially reversed (which is incorrect). We have uploaded new ECOCLIMAP2-v2 files and they are now available on the ALMIP2 server (in the same location as previously). The new filenames begin with "new_". Please rerun if your parameters were reversed. In addition, it was found that the patch fraction sum exceeded unity at some pixels : it seems that the participants using patch information already corrected for this : if not, the new ECOCLIMAP2-v2 files have renormalized patch fractions at the pixels in question.

  • The experimental protocol for the local scale runs was discussed. A short description of this will be made available in May, along with the needed input data. We'll provide information at the next update (in about 2 weeks).

  • In terms of conferences, we will likely submit ALMIP2 papers to AGU, but it is a bit soon to have a devoted ALMIP2 session. However, we are tentatively planning to request a fully devoted ALMIP2 session for next years EGU. Again, more information will be provided at the next update.

  • In terms of publication, we are now thinking of an open special issue : again, details will be provided in another 2 weeks or so.

    We would like to thank again all of those who participated at the workshop, and those who have (or will) contribute results ! And, we'll be providing a more detailed update on the local experiments, publication, possible additional OPTIONAL sensitivity runs soon.

Email Sent (all participants): May. 27, 2013: Updates, Workshop Summary Done

Dear ALMIP2 Participant,

    First off all, we wish to thank those of you who attended the ALMIP2 Workshop in Toulouse: it was great to have good attendence (despite recent budget problems and travel restrictions). We feel we made progress on several issues. The Workshop summary report is currently located at

ftp://cnrm-ftp/pub-moana/boone/almip2/Summary_ALMIP2_workshop.pdf

It will also be uploaded to the documents directory of the ALMIP2 server. Note that the location of the conference presentations is indicated (for those who are interested). We suggest that if you did not attend, you check your model performance in the mesoscale overview presentations (you might decide to rerun) . Note that our policy is that if you wish to rerun your model, please use the same name convention when uploading your files as this will over-write your old results (we will only use the most recent resultsin the final analysis). Of course, if you use an alternate name convention, we will treat your simulations as additional or sensivity runs.

    We now have an updated calender (see the attached workshop report). We have included deadlines: KEY deadlines are:

  1. mesoscale reruns and sensitivity tests due

  2. the local scale runs (2 steps: "blind" runs and "tuned" runs after feedback from us and a distribution of observations for certain time periods) due. The local scale data (for performing these runs) and a short experimental document will be released in early to mid June.

    If you have a desire to lead a model intercomparison or evaluation study, please don't hesitate to contact us (but please do so soon).

    A publication and presentation strategy is also discussed in the attached report: feel free to comment to us.

    Finally, we are preparing a Carbon document for those who wish to run their models using interative vegetation and/or Carbon stores. These will consistute a set of optional experiments. This will be released soon, along with a timetable for results uploading, analysis, etc.

Thanks for your ongoing efforts!

Email Sent (all participants): Aug. 13, 2013: ALMIP2 Update: Local Scale Experiments Ready to Launch

Dear ALMIP2 Participant,

    The local scale NetCDF atmospheric forcing and soil-vegetation files are now available for download (they will be visible on the ALMIP2 data server starting tomorrow, Aug 14.). In addition, there is i) an overview PDF file for each meso-site which gives additional information which may be of use for specific models, along with possibly additional short PDF files with ancillary information (so we encourage you to also look carefully at the PDF files) and ii) ASCII files with soil temperature and moisture initial conditions and ancillary information (additional details which could be useful for certain models). PLEASE take a few moments to look at the PDF and ASCII files (along with the NetCDF files which contain the more standard parameters). For example, very detailed soil profile information is given for Mali in the ancillary ASCII files which could prove useful for certain models (which can simulate multiple heterogeneous layers for example).

    In the NetCDF files, please note that some variables have a "_z" extension: this implies they have a vertical dimension. Those soil parameters without this extension correspond to "bulk" value (or for models which must impose a heterogeneous soil property profile). Also please not that if a particular parameter is FLAGGED (as missing), this means that you are to make your best estimate of this parameter (meaning it was not measured, and or it could be model specific). Also, some vegetation parameters use a daily time step: if your model requires monthly or decade (or weekly, etc..) inputs, please average accordingly. For certain vegetation parameters (eg. LAI), some files also provide addition overstory and understory values (certain Benin and Mali files) using the convention (e.g.)

LAI

LAI_over

LAI_under

    Again, models using a bulk description of the vegetation should use "LAI" in this case, while so-called 2-source models should use LAI_over and LAI_under (this also applies to a few other vegetation parameters).

    Finally, please document how you determine or modify (and possibly interpolate in time) any soil and vegetation parameters (and submit this information as a .doc or .pdf file with your results).

At ftp.bddamma.ipsl.polytechnique.fr on the *download* ftp site, please find:

Atmospheric Forcing files:

/data/forcing/local/$meso_domain

Input soil-vegetation files (ASCII and NetCDF):

/data/forcing/local/$meso_domain/TXT

/data/forcing/local/$meso_domain/NetCDF

Corresponding PDF files (with ancillary information):

/data/forcing/local/$meso_domain/Doc

As a reminder, additional general information is in

/documents/Local_Simulations

    We now request that you begin your "blind" simulations (i.e. using your best estimates for any options or parameters not listed etc.). You may submit multiple runs (using several options or parameter values). Please use the same output template as for your mesoscale results (in NetCDF format), except obviously for a single point per site (using the dimension in space (1,1), similar to the forcing). Please report 30 minute outputs as before.

Practical;

    Over the next several weeks, some of us will be out of the office on annual leave: so please have patience if you have any questions: they will be answered. Please address any questions to the list below. Also, we will collect questions, problems (please do not hesitate to contact us if you find a potential bug, error or missing bit of information: we appreciate your help), comments etc.. and post them on the ALMIP2 web site at:

http://www.cnrm.meteo.fr/amma-moana/amma_surf/almip2/emails.html

    We plan to give everyone first round feedback by Dec. 15, so we would like you to upload your results by Oct. 31. At that time (Dec 15), we will release a subset (time period) of observed variables which can be used to evaluate and potentially tune models. This step is optional, but highly recommended. Tuned results are tentatively due on March 15.

    Please don't hesitate to contact us with questions, problems, comments... Thank you again for your effort!

Sincerest regards

Aaron Boone, Christophe Peugeot, Manuela Grippa Jerome Demarty, Bernard Cappelaere, Laurent Kergoat and the rest of the ALMIP organizing team

For Comments/Questions/Problems: please send email to

aaron.a.boone@gmail.com, christophe.peugeot@ird.fr, jerome.demarty@univ-montp2.fr, laurent.kergoat@get.obs-mip.fr, manuela.grippa@gmail.com, bernard.cappelaere@univ-montp2.fr

Email Sent (all participants): Oct. 11, 2013: ALMIP2 Update-analysis

Dear ALMIP2 Participant,

We have some updates concerning a few points (below).

Mesoscale Reruns:

    We are now post-processing the mesoscale *reruns*. If you intend to perform reruns but have not yet done so, please contact us as soon as possible: otherwise we will consider your most recent simulations as your default runs. Several models which had some problems which were identified at the Workshop have yet to do reruns. Also, if you did reruns, please drop us email to let us know so that we can make sure that we have processed your most recent results. Also, for US participants, we can be flexible on any of the deadlines if needed: please contact us. We will send you new diagnostics (sanity checks) once we have re-processed your results (as before). Once we have all the mesoscale results finalized, we'll proceed with the full analysis and keep you updated.

Local Data corrections:

Please note that a few problems have been identified in the meteorological and/or soil vegetation input NetCDF files. Here's a summary:

Mali

The meteorological reference heights for the 3 sites should be:

Agoufou
Wind : 3m
RH, Tair: 2m

Kelma
Wind: 13 m (= 6 m above the cover)
RH, Tair, 8 m (= 1 m above the cover)

Eguerit:
Wind, RH, Tair: 2.5 m

Benin

For Bellefoungou, just a clarification (this wasn't clear since a few questions have arisen):

Two layers of vegetation are documented (for dual-source models). The first one (Hveg in the files) is the height of herbaceous from 0 to 1.4m The second is the height of trees (15m) (i.e. the total vegetationn height). The displacement height of the forest is 7.6 m and the roughness length, z0, is 1.2m which is an annual average value.

Niger

There are 2 items:

1) The Hveg (for the Millet site) has been corrected, and following soil/vegetation files have been updated:

soildata_Niger_Millet_2005.nc  
soildata_Niger_Millet_2006.nc  
soildata_Niger_Millet_2007.nc  
soildata_Niger_Millet_2008.nc

the files have been updated at

ftp.bddamma.ipsl.polytechnique.fr

This update has a 1 day delay, so the data are also currently located at

ftp cnrm-ftp.meteo.fr
(anonymous login)

/pub-moana/boone/almip2/Niger.tar

2) Please note that for the Niger, do NOT re-initialize your models each year (even though we provided this data). Rather, please run 2006-8 for both sites without re-initializing.

Conferences:

    We mentioned the EGU as a possible forum for a special ALMIP2 session: we have decided not to do this as the calender for this request is not very coherent with our analysis timeframe (indeed, the deadline was in mid-September to make such a request, which was a bit early, especially since the local runs have yet to be analysed). We are currently considering a venue at a later date: your input on this topic is welcome. We will likely have a presence at the upcoming 7th International Conference on the Global Energy and Water Cycle, in the Hague, Netherlands, this upcoming July. Here is a list of some upcoming presentations (if you have additions, please let us know so we can update the website):

Grippa M., Kergoat L., Boone A. Peugeot C., J. Demarty, B. Cappelaere and the ALMIP2 Working group. Evaluation of water and energy cycle simulated by different land surface models over the Gourma Sahelian site in Mali (ALMIP2 project).

Peugeot et al, (and the ALMIP working group) : "How well do land surface models reproduce the water and energy cycles in the West African monsoon system ? Evaluation over the Upper Ouémé basin, Benin (ALMIP2 project) " AGU

Peugeot et al, (and the ALMIP working group), Model, satellite and ground-based estimates of evapotranspiration. A comparison in sub-humid tropical West Africa (Benin) within the framework of the ALMIP2 project. AMS Feb 2014

Boone, A. et al (everyone as coauthors)...likely an ALMIP2 overview abstract (proposal) at the 7th International Conference on the Global Energy and Water Cycle.

Email Sent (all participants): Mar. 16, 2014: ALMIP2 Update

Dear ALMIP2 Participant,

We have updates on 2 items.

    First, we are currently working on finalizing the mesoscale analysis. Our goal is to have a report with updated results/analysis (compared to the Workshop last year) ready by the end of April. Indeed, many of you have resubmitted results, and we thank you for your continued efforts. Some of you have submitted results more recently, and results sent in by April 7 can still be included in the report. This report will be for participants, and will form the foundation for several papers we hope to publish next year (an overview paper, plus a mesoscale analysis paper by site). Additional papers will also be done (and can be proposed by you). Once everyone has seen the report and given any feedback, we can then make a publication calender and make final decisions on the journal, list of papers etc. We will send out an email when the report is ready.

    Second, some of you have performed the local scale simulations, and we thank you. But indeed, we have not been as reactive on this point since we have focused on advancement of the mesoscale runs, so with this email, we would like to re-launch the local scale phase. Can you please let us know if you still plan to participate in this phase? From a technical aspect, this is a much less demanding phase (in terms of volumes of data, post-processing etc.). Also, this phase is more process-oriented, thus it pertains to one of the core goals of ALMIP2 (a better understanding and representation of key surface processes in this region). So for those of you who have not submitted results yet, we hope that you still plan to participate: please let us know and as a function of the responses, we'll make a revised calender for local scale results and analysis.

Email Sent (all participants): Jul. 2, 2014: ALMIP2 Update

Dear ALMIP2 Participant,

    The mesoscale model results were finalized this Spring and we have been analyzing the results (re-running the analysis performed for the ALMIP2 Workshop and adding new evaluations). So, we are preparing a list of papers which we hope will form the basis of either a special issue or special section of a journal. Before we contact the journal, we would like to get a list of possible contributions to this issue. Currently, we propose an overview paper (I), 3 site specific overviews (II) and additional contributions (III). The proposed papers are:

I) OVERVIEW PAPER:

1 An evaluation of surface and hydrological processes along the West African Eco-climatic gradient: AMMA Land surface Model Intercomparison Project phase 2. Boone, A., C. Peugeot, B. Cappelaere B., J. Demarty, M. Grippa, L. Kergoat, S. Ait-Mesbah, M. Anderson, G. Balsamo, E. Chan, J-M Cohard, A. Ducharne, E. Dutra, F. Favot, A. Getirana, S. Galle, T. Gascon, M. Gosset, Y. Gusev, C Hain, P. Harris, X. He, P. Hiernaux, A. Kaptue, Y. Kazuaki, H. Kim, S. Kumar, S. Lafont, M-H Lo, C. Magand, S. Mahanama, F. Maignan, O. Mamadou, E. Mougin, O. Nasonova, C. Ottlé, F. Pappenberger, T. Pellarin, C. Pierre, A. Richard, J-L Roujean, J. Polcher, L. Séguis, A. Shmakin, K. Shunji, V. Sokratov, K. Tanaka, D. Turkov, C. Velluet, D. Verseghy, T. Vischel

II) SITE SPECIFIC PAPERS:

    Please note, for the 3 site papers, we have provided a core group of authors + the ALMIP2 Working GRoup (defined below) If you wish to contact the lead author and make a more significant active contribution to the paper, please do so (and then you will become an "explicit co-author".). By default, everyone is included explicitly in the overview paper author list.

2 Analysis of surface-atmosphere exchanges simulated by land surface models over the AMMA-CATCH central Sahel mesosite (SW Niger). Demarty, J., B. Cappelaere, Boone A., Peugeot C., Grippa M., Kergoat L., C. Velluet, + field campaigne group + (those in the ALMIP2 group who wish to manifest/aid specifically) and the ALMIP2 Working Group

3 Meso-scale evaluations of the Water cycle in state-of-the-art land surface models in the sub-humid west African tropics. Peugeot C., Boone A., Cappelaere B., Demarty J., Grippa M., Kergoat L., M. Anderson, Awessou B., Cohard J.M., Favot F., Galle S., T. Gascon, Getirana A., Gosset M., Hector B., Mamadou O., D. Miralles, T. Pellarin, Quantin G., A. Richard, Séguis L., Seghieri J., Vischel T., + (those in the ALMIP2 group who wish to manifest/aid specifically) and the ALMIP2 working group

4 Modelling surface runoff and water fluxes over contrasted soils in pastoral Sahel: evaluation of the ALMIP2 land surface models over the Gourma region in Mali. Grippa M., Kergoat L., Boone A. Peugeot C., J. Demarty, B. Cappelaere, P. Hiernaux, E. Mougin, L. Gal, M. Anderson, D. Miralles, + (those in the ALMIP2 group who wish to manifest/aid specifically) and the ALMIP2 Working Group

III) ADDITIONAL STUDIES/PAPERS:

Additional paper(s) confirmed thus far with "preliminary" author list:

5 ALMIP2 simulations of dynamic Leaf Area Index and CO2 fluxes: quality of the simulations and importance for surface water and energy budget in West Africa. Kergoat, L., Demarty J., Grippa M., B. Capelaere, A. Boone, J.-M. Cohard, Peugeot C., + one author per model with CO2 or dynamic LAI (Orchidee, Tessel, CLM, Jules, SarraH, STEP...) + other data contributors (TBD)

6 An evaluation of river flow simulated by the ALMIP2 ensemble of land surface models over a West African basin. Getirana, A., A. Boone, C. Peugeot

7 Uncertainty propagation in distributed land surface modelling from point rainfall in the Sahel. B. Cappelaere, D. Feurer, T. Vischel, J. Demarty, C. Ottlé, et al. (TBD)

8...+?

ALMIP2 Working Group: S. Ait-Mesbah, M. Anderson, G. Balsamo, E. Chan, J-M Cohard, A. Ducharne, E. Dutra, F. Favot, A. Getirana, S. Galle, T. Gascon, M. Gosset, Y. Gusev, C Hain, P. Harris, X. He, P. Hiernaux, A. Kaptue, Y. Kazuaki, H. Kim, S. Kumar, S. Lafont, M-H Lo, C. Magand, S. Mahanama, F. Maignan, O. Mamadou, E. Mougin, O. Nasonova, C. Ottlé, F. Pappenberger, T. Pellarin, C. Pierre, A. Richard, J-L Roujean, J. Polcher, L. Séguis, A. Shmakin, K. Shunji, V. Sokratov, K. Tanaka, D. Turkov, C. Velluet, D. Verseghy, T. Vischel

    At this time, we would like to know if wish to contribute a paper based on your work under the ALMIP2 theme (analyzing the multi-model results, using the ALMIP2 data in a study, model modifications/improvements owing to ALMIP2 etc..). Please contact us to let us know the proposed title and potential list of co-authors. Once we get some feedback, we'll re-circulate the list with you before submitting to a journal. Currently, we are hoping to submit to the Journal of Hydrometeorology. Note, that they currently have an "Open" special issue format, so that there is a window in which papers can be submitted.

    Also, please check the co-author lists for the above papers: if you have been left out (or you see an omission or you prefer to not be a co-author of a particular paper), please let us know (and we'll make the correction). The list of authors for each paper is a first cut so please accept our apologies if someone was inadvertently left out.

    We would like to get feedback from you on a potential title and author list within the next week or so. Once we have the list and have contaced the journal, we should be able to contact you with the deadlines.

Local Scale:

Because of the longer-than-expected time required to process and finalize the mesoscale results, we have put the local scale runs/analysis on hold until autumn (so for those of you who have done the results, we still expect to work on the local scale runs). In autumn, we will likely recontact you to see who wishes to still participate (or withdraw from) the local scale phase.

GEWEX Conference at the Hague:

Manuela Grippa has an oral presentation covering the most recent overview of ALMIP2 results at the GEWEX Conference in 2 weeks. Aaron will also be present, so please don't hesitate to contact us to discuss the project or meet us for a coffee at the meeting!

Sincerest regards,

Aaron Boone, Christophe Peugeot, Jerome Demarty, Manuela Grippa, Laurent Kergoat

Email Sent (all participants): Sep. 23, 2014: ALMIP2 Update

Dear ALMIP2 Participant,

    Last call for papers...

    We are currently finalising the list of publications for a "Special Collection" proposal to the Journal of Hydrometeorology. Indeed, we have already iterated this list with you: but as we are about to send it in, please contact us if you wish to add a proposal to the list (with a title, and a tentative list of co-authors....). Also, if you feel you have been left out of (or erroneously added to) the list of authors of a particular paper, please contact us. After the first few "major contributing" authors for each paper, we will resort to putting names in alphabetical order (this has not been done yet as the lists are still a bit in a state of flux). Also note that we have been contacted by a few of you (expressing interest), but now we need a solid proposal if you wish to be a part of the special collection proposal. As a final note, the special collection submission will be open for an extended period (a few months at least) likely. Please note, we are about to submit the proposal: if it is not accepted, we will let you know as soon as possible and come up with an alternate journal or strategy. Concerning a possible paper proposal, please respond to us by Friday.

Email Sent (all participants): March. 17, 2015: ALMIP2 Update

Dear ALMIP2 Participant,

    We have an update on the project publications. We received a positive response from the J of Hydrometeorology in late autumn to our proposition for a special collection of ALMIP2 papers. The submission portal is now open: apparently such collections are open-ended in terms of submission dates, as they are published together in the virtual sense. So, we are focused on trying to begin submitting papers in the summer/early fall time-frame. This means that some of the lead authors will likely be sending you draft papers in the upcoming months (for your potential input). The current list of papers is below. Thanks again for your hard work and patience, and we'll be in touch in upcoming months. Please don't hesitate to contact us with questions, concerns, etc....

Sincerest regards,

Jerome Demarty, Christophe Peugeot, Manuela Grippa, Laurent Kergoat, Bernard Cappelaere and Aaron Boone



The proposed title of the special collection is : The AMMA Land surface Intercomparison Project phase 2 (ALMIP2): Improved modelling of key land surface and hydrological processes over West-Africa from semi-arid to forested regions

The following papers are being prepared for a special collection in the J. of Hydrometeor. (papers are to be submitted in 2015) We will keep updates on the page:

http://www.cnrm.meteo.fr/amma-moana/amma_surf/almip2/pubs.html

Boone, A., C. Peugeot, B. Cappelaere B., J. Demarty, M. Grippa, L. Kergoat, S. Ait-Mesbah, M. Anderson, G. Balsamo, S. Bousetta, E. Chan, J-M Cohard, A. Ducharne, E. Dutra, F. Favot, A. Getirana, S. Galle, T. Gascon, M. Gosset, Y. Gusev, C Hain, P. Harris, X. He, P. Hiernaux, A. Kaptue, Y. Kazuaki, H. Kim, S. Kumar, S. Lafont, M-H Lo, C. Magand, S. Mahanama, F. Maignan, O. Mamadou, E. Mougin, O. Nasonova, T. Oki, C. Ottlé, F. Pappenberger, T. Pellarin, C. Pierre, A. Richard, J-L Roujean, J. Polcher, L. Séguis, A. Shmakin, K. Shunji, V. Sokratov, K. Tanaka,D. Turkov, C. Velluet, D. Verseghy, T. Vischel, T. Lebel, 2015: An evaluation of surface and hydrological processes along the West African Eco-climatic gradient: AMMA Land surface Model Intercomparison Project phase 2.

Cappelaere, B., D. Feurer, T. Vischel, J. Demarty, C. Ottlé, (+ other contributors, TBD), 2015: Uncertainty propagation in distributed land surface modelling from point rainfall in the Sahel.

Cohard J. M., Mamadou O., Galle S., Peugeot C., Awessou B., Seghieri J., Séguis L., Boone A., Cappelaere B., Demarty J., Grippa M., Kergoat L., and the ALMIP2 working group, 2015: Evaluation of surface energy budgets from ensemble land surface model simulations over the sub-humid west African tropics.

Demarty, J., B. Cappelaere, Boone A., Peugeot C., Grippa M., Kergoat L., C. Velluet, + field campaign group and the ALMIP2 Working Group, 2015: Analysis of surface-atmosphere exchanges simulated by land surface models over the AMMA- CATCH central Sahel mesosite (SW Niger).

Getirana, A., A. Boone, C. Peugeot, and the ALMIP2 Working Group, 2015: An evaluation of river flow simulated by the ALMIP2 ensemble of land surface models over a West African basin.

Grippa M., Kergoat L., Boone A. Peugeot C., J. Demarty, B. Cappelaere, P. Hiernaux, E. Mougin, L. Gal, M. Anderson, D. Miralles, and the ALMIP2 Working Group, 2015: Modelling surface runoff and water fluxes over contrasted soils in pastoral Sahel: evaluation of the ALMIP2 land surface models over the Gourma region in Mali.

Kergoat, L., Demarty J., Grippa M., B. Cappelaere, A. Boone, J.-M. Cohard, Peugeot C., (+ other contributors, TBD), 2015: ALMIP2 simulations of dynamic Leaf Area Index and CO2 fluxes: quality of the simulations and importance for surface water and energy budget in West Africa.

Lo, M.-H., A. Ducharne, H. Kim, X. He, T. Oki, P. Yeh, A. Boone, C. Peugeot, (+ other contributors, TBD), 2015: Evaluation of groundwater table simulations from ALMIP2 project

Peugeot C., Boone A., Cappelaere B., Demarty J., Grippa M., Kergoat L., M. Anderson, Awessou B., Cohard J.M., Favot F., Galle S., T. Gascon, Getirana A., Gosset M., Hector B., Mamadou O., D. Miralles, T. Pellarin, Quantin G., A. Richard, Séguis L., Seghieri J., Vischel T., and the ALMIP2 working group, 2015: Meso-scale evaluations of the Water cycle in state-of-the-art land surface models in the sub- humid west African tropics.

Yilmaz, M. Y., M. C. Anderson, C. R. Hain, (+ other contributors, TBD), 2015: Comparison of satellite and land surface model-based energy balance estimates over flux sites in West Africa with a focus on moving from local to pixel scales.

The ALMIP2 Working Group: S. Ait-Mesbah, M. Anderson, G. Balsamo, S. Bousetta, E. Chan, J-M Cohard, A. Ducharne, E. Dutra, F. Favot, A. Getirana, S. Galle, T. Gascon, M. Gosset, Y. Gusev, C Hain, P. Harris, X. He, P. Hiernaux, A. Kaptue, Y. Kazuaki, H. Kim, S. Kumar, S. Lafont, M-H Lo, C. Magand, S. Mahanama, F. Maignan, O. Mamadou, E. Mougin, O. Nasonova, T. Oki, C. Ottlé, F. Pappenberger, T. Pellarin, C. Pierre, A. Richard, J-L Roujean, J. Polcher, L. Séguis, A. Shmakin, K. Shunji, V. Sokratov, K. Tanaka, D. Turkov, C. Velluet, D. Verseghy, T. Vischel