RCEMIP Updates

November 5, 2024

So far, 9 models have contribute their output to the DKRZ swiftbrowser, which is great. But many still have not yet uploaded data.

Originally I was hoping for the simulations to be completed by September, which has obviously come and gone. We are all busy people and I completely understand that delays occur (indeed….my own simulations, while complete, are not uploaded yet, oops!). But, as we’d like to start analyzing data, I’d really like the simulations completed and data uploaded by the end of December. If you need a little bit more time, let me know, and of course also get in touch if you have any questions about the simulations or the output.

As a reminder, here are the relevant instructions:

  1. Protocol paper at https://gmd.copernicus.org/articles/17/6195/2024/
  2. Compress data where possible, use single precision if possible.
  3. Upload the last 100 days of 3D data, but store all 200 days locally. (GCMs, keep 200 days locally and upload the last 100 days. GCRMs, keep 3D output local, do not upload yet).
  4. Instructions for uploading your data to the DKRZ swiftbrowser can be found here.
  5. Along with your model data, please fill out and upload this model documentation form.
  6. Complete your simulations and upload the data by the end of December.

    7. Thank you for your contributions to making this project a success!

    August 23, 2024

    Greetings from the RV Meteor at 5.7 N, 24.7 W in the eastern tropical Atlantic. I'm here for the next month on the ORCESTRA field campaign, so apologies if my replies to your emails are delayed or brief. But I did want to provide a few updates!

    First, the protocol paper was final published in GMD. You can find it here: https://gmd.copernicus.org/articles/17/6195/2024/.

    Second, DKRZ generously increased our storage quota on the swiftbrowser, so now we have space for the RCEMIP-II simulations. But, we still have a limited amount of space, so please ensure the following when providing your data:

            
    1. Compress the data where possible
      2.       
    2. Use single precision to write the output if possible
      3.       
    3. Please only upload the last 100 days of 3D data. If you have already uploaded the full 200 days, that is fine, but for those that have not uploaded yet (most of you), please provide only the last 100 days. You should still store the full 200 days locally. For GCMs, keep the last 200 days locally and upload the last 100 days. For GCRMs, keep the 3D output locally and do not upload it at this time).

      4. Thank you all for your contributions, and I look forward to seeing more of the models uploaded in the coming weeks.

      July 15, 2024

      Here is a quick update on RCEMIP-II. The protocol paper has been accepted (finally!) by GMD pending some technical corrections (which I’ve submitted). Nothing in terms of the protocol itself has changed since the last version I sent you, but the final version is here for your reference.

      I know some of you are working on completing your simulations and getting ready to upload the data. We are having a bit of a space crunch on the DKRZ swiftbrowser, and while we work on figuring out how to resolve it, here are a couple things to consider:

      1. We recommend that you compress your data. This may not provide a huge reduction in data, but some reduction in especially the 3D data would help. Files like those for qcl, qci, qpl, and qpi that have a lot of zeros compress well. At this point I’m not suggesting a specific compression approach, but the resulting netcdf file should still be readable by any software. If there are particular compression strategies that people have found work well for the RCEMIP data, please let us know.
      2. We may need to reconsider the volume of the 3D data request. We had requested that the 6-hourly 3D data be output for the entirety of the simulation, so that the evolution of the aggregation and circulations as they develop could be examined. However the storage cost may be prohibitive. So I’m considering reducing the request to just the last 100 days of the simulation. I don’t think we can get away with a shorter time given some of the low-frequency oscillations that emerge.
      3. GCRMs should not upload 3D data to the swiftbrowser, but please store them locally and we can discuss other options for sharing that data as needed.

      As a reminder, we are hoping for simulations to be completed by September. As always, please get in touch with any questions!

      June 2024

      The revised RCEMIP-II protocol paper remains under review in GMD. It can be found here.

      The protocol has not changed from the initial submission in December, but there are a few additional output requests, so please review the appendix. These additional requests came from the community, to be able to perform analyses that weren’t possible with the Phase I output. Regarding the output, for GCMs one of the requested variables (new compared to Phase I) was clisccp, the ISCCP histogram. This is listed in the Table with the 3D output, but really it should be a histogram (of cloud fraction as a function of cloud top pressure and optical depth) for each horizontal grid point and time (hourly). ISCCP simulator data is not required for CRMs, though CRMs are welcome to run the ISCCP simulator and generate one histogram for the whole domain if your model has that capability.

      Instructions for uploading your data to the DKRZ swiftbrowser can be found here. For technical questions about uploading or downloading data to/from the DKRZ cloud, contact Karsten Peters (peters at dkrz.de). Along with your model data, please fill out and upload this model documentation form. We still hope to have simulations completed by the end of the summer. Please let me know if you have any questions or concerns as you begin to set up and run the simulations.

      December 18, 2023

      I am pleased to finally formally announce Phase II of RCEMIP!

      The Radiative-Convective Equilibrium (RCE) Model Intercomparison Project (RCEMIP) leveraged the simplicity of RCE to focus attention on moist convective processes and their interactions with radiation and circulation across a wide range of model types including cloud-resolving models (CRMs), general circulation models (GCMs), single column models, global cloud-resolving models (GCRMs), and large eddy simulations. While several robust results emerged across the spectrum of models that participated in the first phase of RCEMIP (RCEMIP-I), gaining a deeper understanding of convective aggregation and tropical climate will require reducing the degrees of freedom with which convection can vary. As discussed at the CFMIP-GASS Meeting this past July in Paris, we propose a Phase II of RCEMIP (RCEMIP-II) that utilizes a prescribed sinusoidal sea surface temperature (SST) pattern to provide a constraint on the structure of convection and move one critical step up the model hierarchy.

      The protocol for RCEMP-II is described in a paper recently submitted to GMD, which will be available for discussion in GMDD until February 12, 2024. Information is also provided on the RCEMIP website. Thank you to those who have already indicated your intent to participate in RCEMIP-II. We welcome additional participation! Please register here to participate. I’d also like to emphasize that participation in the Phase I is not necessary to participate in RCEMIP-II, but RCEMIP-II is limited to CRMs, GCRMs, and GCMs.

      We would like the RCEMIP-II simulations to be completed by June 2024, though this timeline may change depending on the outcome of the review process for the protocol paper. If you wish to begin performing the simulations now, you may, but there is the risk that something in the protocol could change during the review process (though I hope not!).

      Finally, even if you do not intend to contribute simulations, please sign-up to receive updates at the registration form, as I won’t send further updates out to this broader list. Please feel free to forward this message to other individuals who may be interested in participating.

      Thank you, and happy holidays!

      p.s. A reminder that the RCEMIP-I data is publicly available at http://hdl.handle.net/21.14101/d4beee8e-6996-453e-bbd1-ff53b6874c0e. It has been great to see this data being used by the community. If you have used the RCEMIP data in a paper, please fill out this form so that we can advertise your work! An incomplete list of publications using RCEMIP simulations can be found on the RCEMIP website.

      March 13, 2023

      Here are a couple quick updates about Phase II of RCEMIP and an upcoming conference of interest to this community.

      CFMIP-GASS Joint Meeting
      The 2023 CFMIP-GASS Meeting on Clouds, Precipitation, Circulation & Climate Sensitivity will be held July 9-13, 2023 in Paris, France. Abstract are solicited on the topics of convective organization, cloud processes, cloud-circulation coupling, and climate feedbacks and sensitivity and the deadline has been extended to March 30. There will also be opportunities for breakout sessions on RCEMIP and other topics, so please consider attending!

      Phase II
      We are still working on testing the Phase II Mock-Walker protocol, which has been delayed somewhat by trying to figure out some low-frequency oscillations that emerge in some settings. A protocol paper is in preparation and some preliminary information about Phase II and the proposed simulations is on the RCEMIP website. You may now register to confirm your participation in Phase II, but we don’t expect to finalize the protocol and begin simulations for another few months.

      AGU Special Collection
      The deadline for our special collection across AGU journals Using Radiative-Convective Equilibrium to Understand Convective Organization, Clouds, and Tropical Climate was extended to June 30, 2023.

      Thank you for your continued interest and participation in this project, and as always, feel to reach out to me with any questions. I hope to see many of you at the CFMIP-GASS meeting in Paris!

      November 18, 2022

      It is hard for me to believe but it has been five years since RCEMIP was officially announced and the protocol paper published (six years since the idea was seeded at the first Model Hierarchies Workshop). In that time, 41 of you contributed simulations with 30 different models and over a dozen papers have been published using the RCEMIP dataset (publicly released with the overview paper a bit more than two years ago). The wide range of topics covered in these papers and the many others in review and in preparation about convective organization, clouds, and tropical climate highlight the utility of RCE for studying many fundamental questions in climate science and this community’s enthusiasm for the project.

      AGU Special Collection
      I would like to remind everyone that our special collection across AGU journals on Using Radiative-Convective Equilibrium to Understand Convective Organization, Clouds, and Tropical Climate is currently set to close on December 31, 2022. So if you are planning on submitting a paper soon on anything using RCE (*not* just RCEMIP simulations), please consider submitting to the special collection before it closes.

      Phase II
      As I announced at meetings this past summer (CFMIP and the second Model Hierarchies Workshop), and have discussed with many of you over the last year and a half, Phase II will consist of “Mock Walker” simulations that will use the same set-up as the large domain Phase I simulations but with a prescribed SST pattern. These simulations are designed to build on the success of Phase I while providing a constraint on convection and circulation (and a hopefully narrower intermodel spread), and thus the motivating science questions are similar to Phase I, geared towards further investigation of cloud-circulation coupling, convective aggregation and climate. More information on Phase II and the proposed simulations is on the RCEMIP website, which will be updated in the coming weeks as we continue to test and optimize the simulation design. I hope to finalize the Phase II protocol by the end of the calendar year, so stay tuned for updates on that and a solicitation for commitments to contribute Phase II simulations!

      Other opportunities
      Separate from the official RCEMIP Phase II simulations, I also wanted to forward information about other potential opportunities to utilize the RCEMIP set-up to investigate other phenomena: (1) tropical cyclones; and (2) cloud-aerosol interactions.

      1. Rotating RCE configurations have been shown to be useful tools for exploring the impact of surface temperature on tropical cyclone genesis and development. If you are interested in adapting your RCEMIP Phase I simulations with the simple addition of planetary rotation, please contact Kevin Reed at kevin.reed@stonybrook.edu.
      2. Guy Dagan (Hebrew University) has proposed examining the sensitivity of RCE to cloud-aerosol interactions and examine these interactions under a range of (equilibrium) environmental conditions. This would be used to bridge between RCEMIP and the GAP initiative global CRM simulations with aerosol perturbation which are currently underway (led by Philip Stier and Sue van den Heever). If you are interested in running your RCEMIP Phase I cloud-resolving simulations again but with different aerosol (or cloud droplet) number concentrations that affect only the microphysics, not directly the radiation, please contact Guy Dagan at guy.dagan@mail.huji.ac.il.

      Thank you for your continued interest and participation in this project, and as always, feel to reach out to me with any questions. I’m looking forward to Phase II!

      July 22, 2022

      Check out the update on RCEMIP from the CFMIP Meeting, held 19-22 July 2022 in Seattle, WA:

      May 11, 2022

      I’m writing to give you an update on Phase II of RCEMIP. If you recall, last summer I reached out to see if there was interest in performing “Mock Walker” simulations as Phase II of RCEMIP, to build on the success of Phase I while providing a constraint on convection. The motivating science questions would be similar to RCEMIP Phase I, geared towards further investigation of convective aggregation and climate and hopefully a narrower intermodel spread.

      Thank you for your interest and willingness to participate in Phase II! I’m happy to say that my NSF proposal, which I submitted last summer regarding this, was funded. So at this time, I’d like to move forward with planning for Phase II. This means finalizing the simulation design and output request and then moving towards performing the simulations.

      The idea was that the simulations would use the same set-up as the Phase I simulations but with a prescribed analytic SST boundary condition. Levi has been testing some configurations in CAM, and I plan to do some testing in SAM. But I know several of you have done similar simulations previously, so now would be a great to time share any ideas or suggestions that you have. Making the set-up as consistent as possible between CRMs and GCMs is one aspect that we need to work to implement.

      There are also several upcoming meetings that could provide good venues to discuss this at, if any of you will be in attendance. The CFMIP meeting will be in Seattle in July, and the Model Hierarchies Workshop will be at Stanford University in CA at the end of August (abstract deadline May 15). The original RCEMIP arose from discussions at the first Model Hierarchies Workshop, so it would bring things full circle to be discussing RCEMIP-II at the second.

      My proposed broad timeline is as follows:

      • Summer-Fall 2022: Optimize and finalize simulation design
      • Fall 2022: Solicit commitments to submit simulations and begin performing them
      • Spring 2023: Deadline for submitting simulations

      That is all for now! Thank you for your continued interest in participating in this collaborative project, and I hope to see you at an upcoming meeting soon.

      June 10, 2021

      One of the things that I’m working on is thinking about a potential phase II of RCEMIP. When we initially planned RCEMIP, there were a variety of ideas floated around for a phase II, including ways of constraining the models to be more similar (by using simplified physics, or imposed convective aggregation), among other ideas. In considering the results we have found so far and the discussions about them when I have presented at various conferences and seminars, the two things that stand out the most to me are (1) how strikingly large the spread in simulated RCE states is, and (2) how strong an imprint self-aggregation has on the climate sensitivity, yet how wide a range there is in the degree of self-aggregation and its temperature dependence. Given these two points, my hope for a phase II of RCEMIP is an additional suite of simulations that would provide some sort of constraint on convection, that would narrow the intermodel spread, but still allow for further investigation of aggregation & climate sensitivity. After discussing with some of the core RCEMIP participants, I would like to propose four “Mock Walker” simulations as Phase II of RCEMIP, with a prescribed SST analytic boundary condition.

      My hypothesis is that having an SST gradient will provide a partial constraint on the convective aggregation, reducing the intermodel spread, but still allowing for some “unforced” aggregation to emerge. By varying the strength of the SST gradient with warming, we’d be able to attribute the impact of aggregation on climate sensitivity to the “forced” aggregation that we are controlling with the SST gradient, and any remaining “unforced” aggregation intrinsic to the model that emerges. An additional motivation for these simulations as a Phase II, rather than a different type of simulation, is that they use the same set-up as the Phase I RCEMIP simulations, with just one easy to implement difference (an analytic SST boundary condition). I hope this will make it easier to analyze Phase I and Phase II together as well as easier for everyone to complete these new simulations.

      So far, I’ve discussed this with 5 of the groups who participated in RCEMIP, from whom I received positive feedback and who will collectively contribute 12 models. With this email, I am reaching out to the rest of the participants who previously submitted RCE_large simulations. The simulations would not actually need to be completed until ~a year from now or longer, and the details of the proposed simulations can be discussed and iterated on with feedback from the RCEMIP community during the next year or so, and they could certainly change somewhat from what I propose above. What do you think about this idea? Would you be willing to agree to perform these new simulations with the model you ran for RCEMIP?

      April 1, 2021

      After recovering from the big push to get the RCEMIP overview paper published last summer, here at FSU we have been continuing our analysis of the RCEMIP simulations with a particular focus at the moment on the clouds, controls on cloud amount, and cloud feedbacks. As you may recall, one of the results in the overview paper was that there is quite a wide spread in cloud fraction across the models. One of the things we are doing as we are digging into the cloud properties and their variability across models more is to re-calculate the cloud fraction. This is in part due to a suggestion of one of the reviewers of the RCEMIP overview paper.

      We are using the 3D data, for those models for which it is available, to re-calculate cloud fraction using a new definition, as well as a re-calculation using the RCEMIP definition to confirm how the sparser sampling from the 3D 6-hourly snapshots compares to the 1D hourly averages. One thing we noticed when we did this was that there was sometimes quite a discrepancy between a given model’s 1D cloud fraction variable and our attempt to re-calculate it form the 3D data.

      Given that, I’d like to solicit some more information from you all about *exactly* how you defined a cloud in computing the 1D cloud fraction variable that you uploaded. At this point, it is not so important exactly what you did, just that we know what you did.

      Please fix out this quick, one-question survey as soon as possible but by Thursday April 8 at the latest: https://forms.gle/aczheSjUYrnzMMsp7

      Thank you very much for your attention to this, and for your continued participation in RCEMIP! Please let me know if you have any questions.

      September 17, 2020

      Check out the update on RCEMIP from the virtual CFMIP meeting:

      September 3, 2020

      I have several brief updates:

      1) Thanks to DKRZ, the RCEMIP dataset now has a permanent identifier! hdl:21.14101/d4beee8e-6996-453e-bbd1-ff53b6874c0e Please use this when linking to or citing the data.

      2) The RCEMIP overview paper has been published in JAMES! https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2020MS002138 Thank you to my 40 co-authors for your contributions. The final formatted version is expected to be posted very soon.

      3) The AGU special collection on “Using radiative-convective equilibrium to understand convective organization, clouds, and tropical climate” is open and already has 4 papers in it! Don’t forget that all RCE papers (not just RCEMIP) are welcome and encouraged for inclusion.

      4) Many of you may be interested in the 2020 Virtual CFMIP Meeting on Clouds, Precipitation, Circulation, & Climate Sensitivity, to be held virtually 14-17 September 2020 at 8 AM MDT (10 AM EDT/2 PM UTC). The registration deadline is September 7 (next Monday). There are great sessions planned on 1) Updates on CFMIP and the WCRP Climate Sensitivity Assessment 2) Does convective organization matter for climate? 3) Does atmosphere-ocean coupling matter for climate? And 4) Do extratropical cloud feedbacks matter for climate?. Speakers include Steven Sherwood, Mark Webb, Yen-Ting Hwang, Sandrine Bony, Natalie Burls, Kyle Armour, Jennifer Kay, Johannes Mulmenstadt and myself. I will be sharing a few RCEMIP results in my talk!

      May 10, 2020

      I am pleased to announce that the RCEMIP data is now publicly available! The standardized RCEMIP output is hosted by the German Climate Computing Center (DKRZ) and is publicly available at https://swiftbrowser.dkrz.de/public/dkrz_70a517a8-039d-4a1b-a30d-841923f8bc7a/RCEMIP/. RCEMIP is a collection of more than 30 different models including GCMs, SCMs, CRMs, GCRMs, and LES, configured in radiative-convective equilibrium. More information about the project can be found here, more information about the protocol can be found in the RCEMIP protocol paper (Wing et al. 2018), and more information about scientific results can be found here. The RCEMIP overview paper is currently in review at JAMES.

      While the RCEMIP organizers and model contributors have various analyses in progress and planned, there are many questions that can be investigated using the RCEMIP data so we strongly encourage others to make use of this unique dataset! We'd like to keep track of who is using the data for what, both to try to avoid duplication of research efforts and help promote work that uses the RCEMIP data. We will list your paper/presentation on the RCEMIP website, which will also help advertise it. Therefore, when you download RCEMIP data we ask that you fill out this brief form to keep us in the loop.

      I’d also like to announce a special collection across several AGU journals, including JAMES, GRL, ESS, and JGR Atmospheres titled Using radiative-convective equilibrium to understand convective organization, clouds, and tropical climate. Submissions are open now through 31 December 2021. Submissions that use the RCEMIP data and/or protocol are particulary encouraged but all RCE studies are eligible for inclusion. The call for papers is as follows:

      Radiative-convective equilibrium (RCE) is a time-honored idealization of the tropical atmosphere suitable for studying basic questions in climate science and accessible to a wide range of model types. This special collection consists of studies that employ RCE frameworks to better understand convective organization, clouds, and tropical climate. This includes questions about the role of convective organization in modulating climate sensitivity, changes in cloudiness with warming, controls on convective organization, and the representation of the equilibrium climate. Studies using results from the RCE model intercomparison project (RCEMIP), an ensemble of more than 30 models of many types following a standard protocol, or that employ the RCEMIP protocol, are especially encouraged.

      Please feel free to forward this information to any others that may be interested in contributing to the special collection or using the RCEMIP data.

      September 20, 2019

      Dear RCEMIP model contributors,

      Would you please fill out the following form to clarify what quantity you uploaded for the cloud water - related variables in RCEMIP - clw_avg, clwvi_avg, clwvi, and clw? CLICK HERE FOR FORM

      Technically, the 1D and 3D variables clw_avg and clw should have been cloud liquid water, while the 0D and 2D variables clwvi_avg and clwvi should have been condensed water path - that is, the column integral of the sum of cloud liquid water and cloud ice. This was described in the CMIP6 documentation for the variables clw and clwvi (from which I based the RCEMIP output specification). However, I realize that it was potentially confusing that clw means something different than clwvi, and want to make sure I have accurate information about what was uploaded for each variable. At this time, it is perfectly fine whether you uploaded cloud liquid water or cloud liquid + cloud ice for clwvi, I just need to know which you did!

      Therefore, I've created the form to get information from each of you about what quantity you uploaded under each variable name, to ensure that we have accurate information and can interpret the output unambiguously. Thank you for your quick response!

      July 25, 2019

      I recently returned from Italy, where we had a very successful summer school on Convective Organization and Climate Sensitivity at the ICTP. During the first week of the summer school, our excellent group of more than 80 students from around the world spent some time analyzing the RCEMIP simulations. The students were divided into small groups, each of who investigated one model. They examined the evolution to RCE, energy balance, and self-aggregation in the simulations. We challenged the students to come up with and apply their own metric for aggregation, and we were impressed with the creative ideas that they had. In the second week, the students presented the results of their analysis, which was very useful for us to get an overview of what the RCEMIP simulations looked like (they also found a few output bugs :-P). Stay tuned in the coming months for more results from the summer school.

      Remember that you can find information about the models who have contributed to RCEMIP, data usage guidelines, and a list of known output bugs on the RCEMIP website http://myweb.fsu.edu/awing/rcemipsims.html.

      I also wanted to remind you that there are several RCEMIP-relevant convection sessions at the AMS Annual Meeting and the AGU Fall Meeting to which I would encourage you to submit abstracts. The session descriptions are below.

      AMS Annual Meeting: “Tropical Convection” Jointly Hosted by the Symposium on Tropical Meteorology and Tropical Cyclones and the Eighth Symposium on the Madden-Julian Oscillation and Sub-Seasonal Monsoon Variability. Abstract deadline: Thursday August 1. Tropical convection, through its control of tropical cloud cover, humidity, and rainfall, is important for both tropical weather and tropical and global climate variability. Organized tropical convection, on a variety of scales, is of particular importance in this regard. Yet important scientific and technical gaps remain in our ability to understand and simulate tropical convection, particularly with regards to the coupling between dynamics and diabatic processes. Contributions are encouraged from theoretical, modeling, and observational studies that address any aspect of tropical shallow and deep convective dynamics, including convective organization, diurnal variations, local circulations (island, sea-breeze, etc..), interactions with the large-scale environment and tropical weather phenomena, and the role of tropical convection in climate.

      AGU Fall Meeting: A013 - Atmospheric Convection: Processes, Dynamics, and Links to Weather and Climate. Abstract deadline: Wednesday July 31. In spite of the important role convection plays in the climate system, many uncertainties remain owing to the wide variety of processes and scales involved. Processes on the smallest scales of turbulence and microphysics connect through dynamics and radiation to both weather and climate, locally through precipitation and hazardous weather, and more broadly through feedbacks on global circulations and the climate system. Improving our understanding of clouds and convection at a process level, their representation in large-scale models, and the ability to forecast related hazards remains an outstanding challenge. This session aims to explore aspects of boundary layer, shallow, and deep convective clouds, considering fundamental processes as well as links to self-aggregation, land surface interactions, and their role in weather and climate. We invite presentations of both observational and modeling studies (including those related to the Radiative-Convective Equilbrium Model Intercomparison Project).

      May 16, 2019

      Just a reminder that the deadline for abstract submission to the CFMIP meeting in Mykonos, Greece (September 30-October 4) is in one week (May 24)! There is an opportunity for us to have an RCEMIP breakout at the meeting and we would like RCEMIP to have a significant presence at CFMIP. I hope to see many of you there!

      In addition, I wanted to bring to your attention that there will be a tropical convection session in the Tropical Meteorology and Tropical Cyclones Symposium at the AMS Annual Meeting in January 2020 in Boston, MA. That session could be a great venue to present RCEMIP-related work for those of you thinking of attending AMS.

      April 5, 2019

      I wanted to bring to your attention the call for abstracts for the CFMIP meeting on Clouds, Precipitation, Circulation, and Climate Sensitivity, to be held September 30 - October 4, 2019 in Mykonos, Greece. As you may know, RCEMIP is an “informal” MIP affiliated with CFMIP and the objectives of RCEMIP are closely aligned with the goals of CFMIP.

      I strongly encourage you to consider submitting an abstract on your RCEMIP-related work to the CFMIP meeting, so that we can maintain the great momentum that RCEMIP has right now and follow-up on the preliminary results that were presented at UCP in Berlin. As noted in the CFMIP call for abstracts below, the role of organization of convection in climate and results from RCEMIP are explicitly highlighted, and we hope that there will be a significant presence of RCEMIP at CFMIP. In addition, there is also the opportunity for us to have a RCEMIP breakout session during one of the evenings of CFMIP to facilitate further discussion and planning.

      For more information, visit conference website. The deadline for abstract submission is May 24, 2019.

      March 14, 2019

      It was great to see many of you in Berlin a few weeks ago, at the UCP conference. In general attendees seemed to be excited and intrigued by the opportunities presented by RCEMIP. We are busy working on analyzing the results, and so I urge those of you who have not yet submitted your simulations to please do so as soon as possible.

      LES Simulations: While in Berlin, a small group of us got together and settled on a configuration for LES simulations to be performed as a subset/branch of RCEMIP. We have commitments from 4 models so far to perform these simulations: SAM (Marty Singh), CM1 (George Bryan), ICON (Tobias Becker), MicroHH (Chiel van Heerwaarden). We welcome participation from other groups - if you are interested in performing LES simulations with your model, please contact me. The configuration guide can be found here and the vertical grid here, and the time frame for completion is by *May*.

      January 5, 2019

      Happy New Year! It was a pleasure seeing many of you at the AGU Fall Meeting in Washington, DC a few weeks ago. We had an excellent session in which some results from RCEMIP and other RCE simulations were presented, and we had an brief meeting to provide updates on the project and coordinate plans for analysis and publication. Those updates, along with a few others, are summarized below. Please forgive the long email.

      1. The deadline for completing RCEMIP simulations has passed (December 31). I am excited that results from 14 models have been uploaded! I have spoken to several others in recent weeks and know that more simulations are expected to be uploaded soon. Even though the deadline has passed, if you are working on your simulations, please continue to do so and upload them as soon as they are complete.

        If you haven’t spoken to me about your timeline for uploading your simulations, please do so. We will continue to accept data after the deadline, but the longer it takes, we cannot guarantee that results from your model will be included in the RCEMIP overview paper. If you are still working on submitting your simulations, please be mindful to follow the format described in the data archive instructions and in my October update, and to also upload the model documentation form.

      2. Publication plans/procedures:
        • Overview paper: We are planning an overview paper of the basic results from all models contributing to RCEMIP, led by myself and the other RCEMIP organizers. This will tentatively include basic analysis of the RCE state across the spectrum of models, the changes in clouds with warming, and the robustness of self-aggregation. My timeline to complete this paper is within the next year. Most of the analysis going into this paper will be completed by myself, my graduate student, and Kevin Reed (we have an NSF grant to support this), but if you would specifically like to contribute analysis to it, let me know. Everyone who has submitted results that are included in the paper will be invited to be co-authors - you do not need to do any analysis to be included, but I will ask you to to at least read the draft of the paper and provide suggestions. I anticipate that an appendix of the paper will include the documentation details of all the contributing models.
        • Subsequent papers using the RCEMIP simulations or a subset of them (by you or anyone else) will need to reference the GMD protocol paper, the overview paper, and the DKRZ for data storage/access. Co-authors should include those who wrote the paper and did analysis; it is not necessary to include those that completed the simulations.
        • Papers you might write about your own simulations, that come out before the overview paper: Please keep me involved and aware of your work, so that we make sure not to overlap. Such papers should be more about the specifics of your model than we would get into in the overview paper.
        • Papers you might write using data from models other than your own, that come out before

          the overview paper: Please consider inviting the contributors of those other model simulations as co-authors on your paper, and keep me involved and aware of your work, so that we make sure not to overlap.

      3. Data access: Currently the DKRZ swift cloud where the RCEMIP simulations are stored is only accessible to those submitting simulations who have requested an account. Once all the simulations that we expect are uploaded, we will create a public url for the cloud server where anyone (regardless of whether you submitted simulations) can browse, download and use data (with proper citation). I probably will wait until the spring to “release” the data to the public.

      4. I have posted a list of the models that have contributed to RCEMIP (and those expected to do so shortly) on the RCEMIP website. If you are still planning on submitting simulations and your name/model is not on this list, please get in touch. I know there are others who expressed interest/willingness in contributing, but I am not sure of your status if your name is not on the list below.

      Thank you all for your hard work and willingness to take part in this project! As I think our AGU session showed, we are already starting to have interesting results and I expect many more to come! As always, please contact me with any questions.

      October 2, 2018

      Thank you to those of you that have completed your model simulations and uploaded them to the DKRZ cloud! We have some output from 4 different models up there now and I’ve started to take a preliminary look at it. Things look very promising, but I have realized that a couple of our data handling specifications were not as clear as they could have been. Here are some clarifications and other updates: (the 4 of you that have already uploaded simulations, don’t worry about changing things)

      1. What we have been calling “1D” output is made of up variables that are only a function of time f(t) and variables that are functions of vertical level and time f(z,t). They were separated by a horizontal line in Table 4 in the paper. From now on, we will refer to the f(t) variables as 0D output, and the f(z,t) variables as 1D output. Accordingly, there should be separate data files written for the 0D and 1D variables, and they should be placed in separate pseudofolders (labeled 0D and 1D) on the DKRZ cloud. The SCALE and dam model outputs on the DKRZ cloud are good examples of what the output structure should look like.
      2. Please remember to fill out the model documentation form. and also upload it to the DKRZ cloud when you upload your data. You can place it in the pseudofolder corresponding to the name of your model (it will then be visible next to the pseudofolders for each experiment).
      3. The time coordinate in your output file should be in units of days since the beginning of the simulation.
      4. If your model does not employ pressure levels (i.e., it uses height levels or a type of hybrid level), please also output the domain- and time-mean values of pressure on your model levels, for approximate plotting purposes. Ideally this would be included in all files, but it is especially useful in the 1D files.
      5. For the cloud fraction variables (1D variable cldfrac_avg and 2D variable cl), please output these as a fraction (between 0 and 1), not a percentage out of 100.
      6. As a reminder, the deadline for submitting output is December 31, 2018.
      7. Our session inspired by RCEMIP at the AGU Fall Meeting ( “A058 Insights on Clouds, Convection, and Climate Sensitivity from Idealized Modeling Studies”) will have poster presentations (A23K, poster numbers 1762-1778) from 1:40-6 pm on Tuesday December 11 in hall A-C and oral presentations (A34F) on Wednesday December 12 from 4-6 pm in room 152A. Please join us!
      8. The data archive instructions and output clarification files have been updated to reflect these clarifications

      August 14, 2018

      I hope you have all had a great summer! I thought it was time to provide a few updates about RCEMIP.

      1. Our nominal deadline for simulation completion is September 2018. I’ve heard from many of you that you are working on configuring your model and completing the RCEMIP simulations, so I know that you are making progress! But I also know that it might take a little bit more time - I myself need to rerun my SAM simulations because I missed a few output variables the first time! So please do your best to get things finished in September, but we will accept contributions through the end of the year (December 31, 2018).
      2. For those of you that have completed your simulations, we have prepared instructions on how to upload your data. Thank you to DKRZ and MPI-M for hosting the RCEMIP data on the DKRZ cloud, where it will be publicly available! The instructions are attached, and are also available on the RCEMIP website (http://myweb.fsu.edu/awing/rcemip.html). Also attached is a form to fill out to document the model simulations that you are submitting (as a fillable pdf and as a Word document) - please fill that out and upload it as a pdf to the DKRK cloud as well (in the pseudofolder for your model). If you have any technical questions about the DKRZ cloud, contact Karsten Peters (peters@dkrz.de).
      3. Regarding the output, there have been a couple of clarifications and corrections, which I have been keeping up to date on the RCEMIP website (current version here) As a reminder, the 1D and 2D output are to be hourly averages that are uploaded for the entire simulation. The 3D output, however, is to be instantaneous 6-hourly snapshots, and only the last 25 days of output are to be uploaded. Please do your best to output all the listed RCEMIP variables, but if there are any that are excessively difficult or impossible, please just let me know.
      4. We have convened a session inspired by RCEMIP at the AGU Fall Meeting (A058: “Insights on Clouds, Convection, and Climate Sensitivity from Idealized Modeling Studies”). Thank you to all that have submitted abstracts, and we encourage you to all attend what we hope will be a great session! Stay tuned for more program details.

      Thank you all for your participation in this exciting project!

      May 8, 2018

      One further clarification: There was an error in the definition of frozen moist static energy in section 4.2.2 (including equations 7 and 8). Please see here for the correct definition - make sure you are using the correct definition when diagnosing the FMSE budget. I apologize if this has caused any confusion - I am not sure how I missed it in the proof stage. The correction is also posted on the RCEMIP website and the official paper will be corrected as soon as possible.

      April 25, 2018

      I wanted to get in touch and provide a few updates about RCEMIP.

      1. As I noted in my previous email, we would ideally like the simulations complete by the fall - September 2018. We understand that it might take some time for everyone to set up and complete their simulations, but that should be your initial goal in terms of a deadline.
      2. To provide some motivation to complete the simulations and encourage groups to perform preliminary analysis, we have proposed a session to be held at the AGU Fall Meeting in December 2018. I’ll send more information about that if the session proposal is accepted, but it would be about “Insights on Clouds, Convection, and Climate Sensitivity from Idealized Modeling Studies”. We hope to have a session at EGU next April as well, and eventually, an RCEMIP workshop.
      3. We will follow-up soon with information on how to upload your simulation output to the data archive.
      4. A couple clarifications on the output specification can be found here.

      February 26, 2018

      I wanted to provide an update on the status of RCEMIP. I am pleased to announce that the protocol paper has been accepted in Geoscientific Model Development (doi:10.5194/gmd-2017-213). The final version has not yet appeared on the GMD website, but can be found here and is attached for your convenience. There have been a few minor changes to the simulation design and specification, so please refer to the final version of the paper when configuring your simulations. You will also find useful information and resources on our website. Our vision of RCEMIP is that the simulations we describe in the paper are only a starting point. One of our goals is to foster a community of RCE modelers and hope that there will be subsequent experimentation, informed by the initial results and the interests of the participants.

      Thank you to those that have filled out the RCEMIP Registration Form. Based on the initial responses, we have about 10 models of various types committed and an additional 10 tentatively set to participate, but we would like there to be more! If you plan to participate in RCEMIP and have not yet filled out the registration form, please do so as soon as possible. If you filled out the form previously with a “maybe” and are now able to commit to participating, please fill it out again. There is no hard deadline for confirming your participation, but we would ideally like the simulations to be complete by Fall 2018 (as a motivation, we plan to convene a RCE session at the AGU Fall Meeting in December).

      *IMPORTANT*: This will be the last email that I send to this broad distribution list, so if you want to continue to receive updates and/or plan to participate in RCEMIP, please fill out the RCEMIP Registration Form

      Again, thank you for your support of this project, and please feel free to contact me (awing at fsu.edu) with any questions you may have.

      September 19, 2017

      I am pleased to announce RCEMIP, which is an intercomparison of models configured in radiative-convective equilibrium (RCE). RCE is an idealization that has long been used to study basic questions in climate science; however, assessing the formulaic sensitivity of simulations of RCE is hindered by the absence of a common baseline. Arising from discussions at the WCRP Model Hierarchies Workshop, we have designed RCEMIP to include a wide variety of models configured in RCE and to address the following three themes:

      1. What is the response of clouds to warming and what is the climate sensitivity of RCE?
      2. What is the dependence of the degree of aggregation of conovection and tropical circulation regimes on temperature?
      3. What is the robustness of the RCE state, cloud feedbacks, and aggregation across the spectrum of models?

      We believe that RCEMIP is poised to make progress on some of the biggest open questions in climate dynamics and the simple premise of RCE will allow the results to be connected to theory, as well as serve as a useful framework for model development and evaluation. RCEMIP distinguishes itself from many intercomparisons because of its ability to involve many model types. Therefore, we ask you to please consider participating in this exciting project! We encourage participation from both modeling centers and individual researchers and welcome contributions of simulations from cloud-resolving models, general circulation models, global cloud-resolving models, and single column models.

      The RCEMIP protocol is described in a paper that is currently in review at Geosci. Model Dev., and is available here, with interactive discussion until 14 Nov 2017. The paper is also attached for your convenience.

      More information can be found on our website, at http://myweb.fsu.edu/awing/rcemip.html. In particular, if you are interested in participating in RCEMIP, please fill out the following form (also linked on the website): RCEMIP Registration Form. You are also welcome to contact me (awing at fsu.edu) or any of the other organizers (listed below and cc’d) with any questions you may have. In addition, please feel free to forward this message to any other individuals who may be interested in participating.

      We thank you in advance for your support and participation in this project!

      RCEMIP Organizers:

      Allison A. Wing (Florida State University), Kevin A. Reed (Stony Brook University), Masaki Satoh (University of Tokyo), Bjorn Stevens (Max Planck Institute for Meteorology), Sandrine Bony (LMD/IPSL/CNRS), Tomoki Ohno (JAMSTEC)

      Questions? Email awing at fsu.edu