AI-Translated Page
This page was translated using artificial intelligence (AI). The content may contain inaccuracies or misinterpretations. For critical or official use, please consult the site administrator to confirm the information.
November 4, 2021
MINUTES 004/2021 - Revision 004
Meeting of the Scientific Committee of the Community Model of the Unified Earth System. At 9:00 am on November 4, 2021, representatives of INPE (National Institute for Space Research), INMET (National Institute of Meteorology), UFCG (Federal University of Campina Grande), CENSIPAM (Management and Operational Center of the Amazon Protection System), ITA (Technological Institute of Aeronautics), INPA (National Institute for Amazon Research), UFSM (Federal University of Santa Maria), USP (University of São Paulo), LNCC (National Laboratory for Scientific Computing), UFMS (Federal University of Mato Grosso do Sul) and UFRJ (Federal University of Rio de Janeiro) met virtually, with the aim of continuing the work of the Scientific Committee of the Community Model of the Unified Earth System (MCSTU) 1 . These minutes record the meeting and gather the information entered in the chat, such as links and information relevant to the discussions held. Following the opening of the meeting, carried out by Saulo Freitas and Pedro Dias, this document is oriented according to the guidelines established by Saulo Freitas for conducting the meeting.
Opening
Saulo Freitas begins the meeting by welcoming the participants and then presents the topics to be discussed:
- Brief report on the activities of the Scientific Committee (CC) coordination;
- The process of choosing the name of the Unified Earth System Community Model (presentation of the report by Saulo Freitas/INPE and Pedro Dias/USP);
- The future of Data Assimilation (report presentation by João Gerd/INPE);
- Report of subcommittee meetings:
- Continental Surface and Soils;
- Data assimilation;
- Communication from CC member Ronald Buss/INPE;
- Next actions.
Agenda 1 - Brief report on the activities of the Scientific Committee (CC) coordination
Saulo Freitas reports that two Project Opening Terms (TAP) were opened within INPE/MCTI: one TAP for MCSTU, with the prospect of R$50 million for 10 years; and the other TAP associated with the purchase of a supercomputer and contracting of development services (via the National Fund for Scientific and Technological Development - FNDCT ): with the prospect of R$30 million (for scholarships, services and costs) for 4 years and R$170 million (including the new supercomputer, a data center and energy and cooling infrastructure). He reports that at this moment, the Ministry of Science, Technology and Innovation ( MCTI ) is preparing a budget execution term via ( FINEP ), but there is uncertainty regarding the Ministry of Economy ( ME ). There was also the production of reports and meetings with the National System Operator ( ONS ) about the water crisis. The following arrangements were made with the General Coordination of Earth Sciences ( CGCT/INPE ): 50% of the capacity of the EGEON cluster will be reserved exclusively for the development of the MCSTU (2 thousand cores); 50% of the workload of the High Performance Processing (PAD) group, which involves 3 employees of the Numerical Modeling Division of the Earth System - DIMNT /INPE) will be reserved for the development of the MCSTU; scholarships from the Institutional Training Program (PCI) of INPE at the master's level for the development of the components of Physics, Dynamics, Surface, Data Assimilation, PAD and Model Evaluation. Saulo Freitas also reports that actions are being taken with the National Laboratory of Scientific Computing ( LNCC ) for access to PAD services. He mentions the presentation of the MCSTU project at the 51st anniversary event of INPE in Cachoeira Paulista, which was attended by the representative of the Ministry of Science, Technology and Innovation ( MCTI ) and the director of INPE.
Agenda 2 - The process of choosing the name of the Unified Earth System Community Model
Saulo Freitas presents the results of the process of choosing the name of the community model. He mentions that 9 names were suggested and 24 people voted, and informs that the name "BRACES" received the most votes. He comments that the new name for the community model is not yet suitable. He adds that, in order to attract international partners, the name should not allude to the country of origin, and that it is not a tradition to include the names of countries in the names of models because this inhibits the attraction of partners, especially from Latin America. Pedro Dias complements this by citing the case of the model Brazilian developments on the Regional Atmospheric Modeling System ( BRAMS ). He comments that he thought about the meaning of the word braces in English, which can have different meanings and that, for this reason, he questions whether this name is suitable for a numerical model. He comments on the meaning of the name "MONAM", which indicates a relationship with indigenous gods from South America. He mentions that the name "COMMUTES" is interesting and that, in English, it shows connections with the "spirit" of modeling. He comments that the name "BRUMA" refers more to the environmental aspect. He concludes that the names "COMMUTES" and "MONAM" are more appropriate for the name of a numerical model. After the results are released, Saulo Freitas and Pedro Dias begin the discussion with the members of the CC.
Enio Souza comments that, given the arguments presented, one should choose between the names "COMMUTES" and "MONAM". He reveals that he voted for the name "MONAM". He maintains that his choice is justified by the reasoning behind the name not making explicit reference to any specific aspect of the model, in addition to being easy to pronounce. He comments that the explanation of the name of the model contains a positive invocation (whose context is indigenous and environmental) and that this is important, especially for the coming decades.
Marcia Yamasoe, in the chat, comments on the name "MONAN", showing concern due to possible associations with folkloric figures. Ronald Buss comments that he sent a suggestion that was not considered because he did not send it in time. In his view, he comments that the names of the models cannot refer only to the meteorological part. He agrees that the name should not contain the name of the country, but that it should be simple and indicate what the model does. He mentions the Brazilian Earth System Model ( BESM ) and that his suggestion for the name is Earth System Model (ESM, without the letter B). He comments that the name "MONAM" does not represent our cultural alignment with indigenous peoples. He adds that it should be assumed that the name reflects in a basic and direct way its essential functionality. Comparing it with the names of other models, it is noted that it is unnecessary for the name of the model to reflect the community, unified and Brazilian aspects. It concludes that the acronym can later be linked to the country or institution of origin, as is the case with several models from the Climate Model Intercomparison Project ( CMIP ).
Haroldo Fraga comments that the choice of name is very important. He cites the example of physicists, who adopted specific names for certain theories of physics and that this took away the merit of the authors of these theories. He adds that, for this reason, it is worth dedicating more time to defining the name of the model.
Marcia Yamasoe adds to her initial statement by saying that she found the proposal for the name "MONAM" interesting, but that the association with folkloric figures may not be positive. She adds that instead of "BRACES", it could be Latin America Climate Model of the Earth System (LACES). Saulo Freitas comments that it is a good idea, but that so far, there are no international partners. Haroldo Fraga agrees with Saulo Freitas. Luiz Cândido, via chat, suggests that, given the establishment of these rules, it would be better to hold a new round for the proposal and the choice of the name. Saulo Freitas agrees and reinforces that it is a good idea to hold a new round of proposals for the name of the community model. Haroldo Fraga also agrees with Luiz Cândido's proposal.
Antônio Manzi suggests that, for now, the current acronym can be kept and that an option that better represents the community model should be considered. He suggests that, in addition to the name of the model not making any reference to Brazil, one of the rules should be that the name contains Earth System Model (ESM).
Ronald Buss, via chat, agrees with Antônio Manzi. Haroldo Fraga agrees with Pedro Dias' initial statement, that the name should be well chosen and that the meaning in English should also be well thought out, to avoid bad connotations. He agrees with a second round of name suggestions. He comments that he has no restrictions regarding indigenous references.
Luiz Flávio comments that choosing a name is especially important for the computing team, since it is from this choice that the foundations of the model will be established in the repositories. He adds that it is important that the name of the community model be defined soon so that workers can begin and avoid future difficulties.
Ronald Buss cites Figure 3 of Casagrande et al. (2021) with a list of 20 Earth system model acronyms. Saulo requests that the list be sent to Fabielle Alves for dissemination among CC members.
Pedro Peixoto, via chat, shares a link with a list of acronyms that the American Meteorological Society ( AMS ) recommends for its publications and indicates the "Climatic, meteorological, oceanographic, and other models" section of the list.
Enio Souza defends the name "MONAM" and comments that the folkloric personalities remembered by Marcia Yamasoe are part of our culture and that this is not a relevant issue.
Joaquim Costa, via chat, suggests the name "MODEST".
Saulo Freitas concludes by saying that a new survey will be requested to choose the new name for the community model.
João Gerd, via chat, suggests that each person can vote for more than one name.
Agenda 3 - Future of Data Assimilation
João Gerd begins his presentation by discussing data assimilation, with the aim of introducing the main concepts to all CC members. He then discusses the use of the Joint Effort for Data Assimilation Integration ( JEDI ), which is a community effort to assimilate data for use with various numerical models. He mentions that for different components, in general, there are data assimilation systems that are distinct from each other and that with JEDI, the aim is to overcome the differences between these systems. He also mentions that modern data assimilation systems are too complex for a single person to master alone. He adds that software modernization increases the speed of future developments, in addition to facilitating maintenance and increasing portability and efficiency. In addition, software modernization reduces duplication of efforts among the partners of the Joint Center for Satellite Data Assimilation ( JCSDA ), especially in the inclusion of new observations and the implementation of new algorithms. He adds that the methodologies employed by JEDI allow all components of the Earth system to be treated within the same data assimilation system, which facilitates the transition between research and operations (Operations to Research to Operations - O2R2O ). He comments on the Object-Oriented Prediction System ( OOPS ) component, which is the basis of JEDI. Originally developed by the European Centre for Medium-range Weather Forecasts ( ECMWF ), OOPS is an abstraction layer of JEDI, in which some classes are implemented that are applied in different types of algorithms and numerical models, something called separation of concepts. He mentions the main differences between Gridpoint Statistical Interpolation ( GSI ) and JEDI regarding the observation operator, the difficulties and the cost involved in working with each system: for GSI, it is necessary to modify the data assimilation system as a whole, so that it works with a new numerical model. This is not necessary in JEDI, which has generic implementations and makes the user just submit the numerical model predictions in the standard JEDI format.
Pedro Dias intervenes and mentions the importance of the observation operator. He comments that the subject seems simple, but in reality, it is not. He argues that it is necessary to have a very good understanding of the model in order to transform the predictions into radiances, which is what the satellite measures. He adds that this requires a team that has a very good understanding of the model and the data assimilation process. He mentions that this is a factor that has delayed the implementation of the data assimilation system in Brazil and that what was missing was the interaction between these two teams (numerical modeling and data assimilation). He concludes that this should serve as an example for the development of the MCSTU.
Haroldo Fraga mentions the mistakes made in the past, and states that data assimilation is the process that unites everything and everyone in the modeling process. For assimilation, observation data is necessary. Sometimes the data is available, but often the quality of the data is unknown and therefore it is not assimilated. He comments that this needs to be done automatically and that this is a point that needs to be evaluated. This is a fundamental issue that has never been addressed. Ronald Buss, via chat, agrees with Haroldo Fraga and comments that this is one of the major problems for oceanographic data.
João Gerd agrees that these points are important, and that the quality control process for observation data is automated because someone programmed it. He adds that satellite data is the most abundant and that someone in this area (a satellite data specialist) is needed to work with it. He comments on the observation operators that are already implemented in JEDI and that JEDI is being developed by a satellite data agency, the JCSDA, and that their interest is for the data to be used in numerical modeling.
Pedro Dias asks if data that is not in the Global Telecommunication System ( GTS ) can also be assimilated by JEDI.
João Gerd comments on the Interface for Observation Data Access ( IODA ) and its generic interfaces, which can be used to include other types of data and on the implementation of quality control procedures.
Haroldo Fraga questions whether anyone has already mastered this system, arguing that it is necessary to know the system in order to develop the ability to assimilate observation data about Brazil that are not available through the GTS.
João Gerd comments that the JCSDA has been offering courses and training on JEDI ( JEDI Academy ), where all JEDI components are explained and trained. He mentions that the training is independent of the model, highlights the system's open development methodology and that it offers support for the necessary developments. He concludes that these JEDI development approaches are important to them, since users also contribute to code development.
Haroldo Fraga asks if there is someone assigned to prepare the data for the data assimilation process, and mentions that it is necessary to have people specialized in this subject to give it due attention. Saulo Freitas agrees with the point raised by Haroldo Fraga. Carlos Bastarz, via chat, adds a link and comments that we should inform ourselves about how the system works and that the JCSDA/JEDI can guide us on specific issues.
João Gerd argues about the need for specialists in observation data and cites the history of this issue in CPTEC data assimilation and the greater focus given to the numerical modeling aspect of the center, in contrast to the needs regarding the treatment of satellite observations for use in data assimilation.
Pedro Dias intervenes and mentions that in the history of CPTEC, there was a data analysis group that contributed to the construction of quality control systems. Over time, this function evolved into other activities and the initial objective was left in the background. He adds that the modeling personnel became more concerned with the models themselves and not with the interfaces between the model and the data assimilation system, which is as important as the model.
João Gerd comments on the quality control process of observations and mentions the buddy check process, in reference to Pedro Dias' comment on the need to improve these processes. Carlos Bastarz, via chat, indicates an article in which there is a discussion on the improvement of the quality control processes of observations.
Antônio Manzi intervenes and comments on the need for quality control of satellite data for use in data assimilation. He suggests the creation of an intermediary data group in the Satellite and Meteorological Sensors Division (DISSM) 2 , so that its current activities can also continue, in the development of new products from new satellites. Another aspect highlighted by Antônio Manzi is the role of a forecasting group that can monitor numerical forecasts to assist in the development of the model. For both proposals, Antônio Manzi emphasizes the need for discussions.
João Gerd mentions that the use of synthetic observations is an important tool for quantifying the impact of observation systems on the quality of analyses. He adds that DISSM is the largest group of satellite experts and that it is necessary to take advantage of the knowledge of this group for activities related to satellite data in data assimilation.
Haroldo Fraga intervenes and argues that it is necessary to present the demand for this activity, in order to request the data that the satellite group has and perform the necessary tests for its assimilation. He comments that, although this initiative is necessary, it requires coordination in order to be effective in the operational sphere. Pedro Dias comments on the opportunity for numerical modeling to also contribute to satellite product development activities. He mentions that modeling can indicate what the demands are in terms of remote sensing products that can assist in numerical modeling. He argues that data assimilation is as important as numerical modeling and that it is essential to develop satellite products that provide useful information for improving forecasts.
João Gerd comments on other JEDI components, such as the Interface for Observational Data ( IODA ), OOPS and Unified Forward Operator ( UFO ). He talks about the portability of the system, which can be run in containers on systems such as workstations, laptops, in the cloud and in High Performance Computing (HPC) environments.
Pedro Dias comments that João Gerd's presentation was fundamental for understanding the complexity and importance of data assimilation. He adds that this topic is within the spirit of the group, which is the idea of using the tools that are available, such as JEDI.
Saulo Freitas reinforces his commitment to working with the data assimilation group and that without the appropriate initial conditions, it will not be possible to achieve the MCSTU objectives.
Other Matters
Haroldo Fraga comments that there is a version of the BRAMS model available in container form .
Ronald Buss reports on the activities of the oceans and cryosphere subcommittee. He comments that, according to the group's discussions, there is a consensus that the priority should be resolving the meteorological time scale. From a climate perspective, this will be slower for an Earth system model. From an oceanographic perspective, data assimilation is not only resolved with satellite data. The Oceanographic Observation and Modeling Network ( REMO ) is an example of this, due to the way it handles observations. He cites the examples of applications of the HYbrid Coordinate Ocean Model ( HYCOM ) and Regional Ocean Modeling System ( ROMS ) models for coupled ocean-atmosphere studies on the time scale, which are not the same as those applied to studies of ocean processes on the climate scale, such as the application of the Modular Ocean Model 6 ( MOM6 ) in the context of the development of the Brazilian Earth System Model ( BESM ). He comments that these two scales should be prioritized and that the application of these models should still be discussed within the subcommittee. He adds that there are few groups that have worked with regional ocean models, but that the members of the subcommittee have interacted with them. In addition, there are other groups that have worked with the coupling between ocean, atmosphere and sea ice processes and these processes are more important for the representation of seasonal climate scale phenomena, which influence the climate of Brazil. He adds that in data assimilation, it is important to identify the processes that should be mastered so that the relevant data can be assimilated.
Saulo Freitas appreciates the report presented by Ronald Buss and comments that each subcommittee must clearly define its demands, since it is necessary to bring resources to the MCSTU. All information provided should be considered to take advantage of opportunities to prospect resources. He invites Ronald Buss to detail his report at the next CC meeting.
Ronald Buss comments that the Secretariat of the Interministerial Commission for Marine Resources ( SECIRM ) of the Brazilian Navy ( MB ) announced an agreement with the Coordination for the Improvement of Higher Education Personnel ( CAPES ) for projects in the area of marine sciences (for up to 10 projects financed with up to R$1 million each). He comments that this is an opportunity to carry out a direct action so that part of these resources can be used in coupled ocean-atmosphere modeling projects, which may be of interest to SECIRM since it is coordinated by the MB.
Saulo Freitas thanks everyone for their presence and closes the meeting.
Actions for the Next Meeting
- Discussion with the São Paulo Research Foundation ( FAPESP );
- Internal meetings with the Postgraduate Program in Meteorology ( PGMET ) and DIMNT of INPE on the role of these bodies in the development of MCSTU;
- Meeting of the atmosphere subcommittee;
- Oceans/Cryosphere Subcommittee Meeting;
- Demands from the Brazilian Navy (
MB
) and the National Institute of Meteorology (
INMET
) for the MCSTU (what are the requirements that the MCSTU must
have):
- INMET: Francisco Quixaba;
- Brazilian Navy: Flávia Rodrigues.