ERIS Environmental Research Infrastructures Strategy for 2030 Ari Asmi Jacco Konijn Antti Pursula Version: 1.0 Version history: Version Date Author(s) 0.0.1 2013-­‐07-­‐06 A. Asmi (AA) 0.0.2 2013-­‐09-­‐02 AA, Magdalena Brus (MB), Sanna Sorvari (SS) 0.03 2013-­‐11-­‐19 AA 0.04 2013-­‐12-­‐13 AA, Jacco Konijn Iteration by the (JK), Antti Pursula core writing (AP) group, clarified vision 0.045 2014-­‐01-­‐26 AA Added preliminary action descriptions 0.5 2014-­‐04-­‐23 AA, JK, AP, SS Second comments from RIs included 0.6 2014-­‐05-­‐21 AA Minor adjustments on comments from EISCAT-­‐3D and EPOS 1.0 2014-­‐06-­‐01 ENVRI partners Accepted contentwise by the ENVRI RIs Description Initial draft First internal iteration (U. Helsinki) First comments from RIs included Contents Table of Contents Abstract ......................................................................................................................................... 4 Scope and purpose of this document .................................................................................. 5 State of Environmental research infrastructures .................................................. 6 Environmental challenges now and in future .................................................................. 6 Environmental Research Infrastructures in Europe ..................................................... 7 Cross disciplinary challenges .............................................................................................. 10 Information and knowledge gaps ....................................................................................... 12 Summary of the state of European Environmental Research Infrastructures .... 14 Vision for the Environmental Research Infrastructures ................................... 15 Envisioned benefits of the Holistic understanding of Earth System ...................... 16 Strategy for integrated holistic Earth System understanding ......................... 18 Technological Capital development .................................................................................. 19 Instruments and measurement station development .......................................................... 19 e-­‐infrastructure components ........................................................................................................... 20 Cultural Capital development .............................................................................................. 23 Disseminating the idea of a common goal in Earth System sciences ............................. 23 Building the culture of open research .......................................................................................... 24 Standardize the language and terminology ............................................................................... 24 Enlarging the view ................................................................................................................................ 26 Organizational framework ................................................................................................................ 26 Human Capital Development ............................................................................................... 28 Curriculum development for data scientists, connected to disciplinary science ...... 28 Need for a wide enough perspective outside the own discipline. .................................... 29 Geographical and interdisciplinary mobility programs ....................................................... 29 Serving the citizen scientists ............................................................................................................ 30 Final words ........................................................................................................................ 31 Connection to other sciences, regions and infrastructures ....................................... 31 Summary and view to the future ........................................................................................ 32 Abstract Environmental Research infrastructures are facilities, resources, systems and related services that are used by research communities to conduct top-­‐level research. They are designed as long-­‐term entities in order to meet the requirements of continuous environmental observation. This longevity makes the environmental research infrastructures ideal structures to support long-­‐
term development in environmental sciences. The vision for environmental research infrastructures for 2030 is based on holistic understanding of our planet and it’s behavior, processes, feedbacks, and fluxes; developing an environmental system model, a framework of all interactions within the Earth System, from solid earth to near space. Scientists that within their own science contribute with data, models, algorithms and discoveries should feel that this serves a greater good, namely a contribution to this understanding. To grasp the Earth System as an interlinked system, we aim for a systems approach, mainly because modern science, engineering and society are increasingly faced with complex problems that can only be understood in the context of the full overall system they belong to. The strategy is based on developing three key resources of the Environmental Science: technological, cultural and human capital. The technological capital development concentrates on improving the capacities to measure, observe, preserve, compute, and predict. This requires staff, technologies, sensors, satellites, floats, software to integrate and to do analysis and modeling, including data storage, computing platforms and networks. The cultural capital development addresses issues such as environmental literacy, open access to data, rules, licenses, citation agreements, IPR agreements, and technologies for machine-­‐machine interaction, workflows, metadata, and RI community on the policy level. Human capital actions are based on anticipated need of specialists, including data scientists and ‘generalists’ that oversee more than just their own discipline. To achieve the overall goal, the strategy has a list of action items that contains intermediate aims, bigger and smaller steps to work towards the development of the chosen approach. Scope and purpose of this document This document provides a Strategy for the European (in-­‐situ) Environmental Research Infrastructures (RIs). Although the strategy is build from RI point-­‐of-­‐
view, it is meant to be used by the wider Earth System science community. The role of RIs as long-­‐term actors in the science community makes them natural platforms to initiate and develop strategic development plans, bringing the needs and ideas of their respective sub-­‐domains to more general overall picture. This document is meant in order of importance to the following groups • To the Environmental Research Infrastructures (in Europe and beyond) to describe the
•
•
•
•
•
commitment of progress, give overall direction of development and to facilitate cooperation; To the emerging Environmental Research Infrastructures (currently in operating as RI
networks or projects) networks for defining the RI development and co-operation;
To the national, European and international policy stakeholders (funding performing
organizations, ministries, policy and strategy bodies such as ESFRI, that are setting the
scene for policies by directing the bottom up needs and potentials of the Earth System
science community, and the knowledge related to changes regarding the Earth System
science;
To the Environmental and IT science community, to express the needs of Environmental
RIs in the future To the education sector, explaining the need for new specialists in interdiciplinary research To the private sector, highlighting the need of their participation in the Environmental RI
development, via various stakeholder groups or direct involvement (e.g. by instrument
development) This document was made as a part of ENVRI FP7 project, in collaboration with the participating RI representatives in the ENVRI Stakeholders Advisory Board. Individual actions and tasks presented in this document are presented as examples, not as detailed requirements. This strategy does not necessarily cover all requirements and development needs of individual participating Research Infrastructures. State of Environmental research infrastructures Environmental challenges now and in future Many of the most urgent challenges human societies are facing, such as climate change, energy use, water availability, food security, land degradation, hazards and risks, life in mega cities and human health are closely related to complex interactions with the environment (Figure 1). Tackling environmental challenges is crucial for mankind and for life on Earth. Given the scale and complexity of the challenges, the focus area of the environmental research is our whole planet. Global-­‐scale environmental research and its supporting data delivery with analyses from observations, experiments and modeling are, therefore, key factors in improving our knowledge of the complex interactions between human activities and our ecological life support systems. The complexities of the Earth system as well as the critical challenges defined above require appropriate scientific approaches including inter-­‐disciplinarity and the systems approach. Figure 1. Landscape of environmental challenges Table 1. Simplified typical features of different public research components University research Research Institution Research infrastructure Typical funding model Projects Projects and institutional funding Governmental strategic funding Typical time period of activity Short (1-­‐4 yrs.) Medium to long term Long term Main contribution to scientific process Education of researchers, cutting edge research Medium to long term continuity in research, operational research, applied research Maintaining the research capabilities and prerequisites (instruments, standards, policies) For the Earth system, as for any system, our capacity to understand is predicated on our capability to describe various elements and their interactions. The intelligent design of human activities to minimize their negative impacts on environment will only be successful if it is based on holistic knowledge and adequate predictive capabilities. By carefully exploring and studying the basic environmental processes and synthesizing our observations into an overall understanding, new scientific breakthroughs can be achieved and environmental challenges can be tackled. Environmental Research Infrastructures in Europe Research infrastructures are defined in this document as long-­‐term facilities, resources, systems and related services that are used by research communities to conduct top-­‐level research in their respective fields. The focus is in in-­‐situ or short-­‐range remote sensing infrastructures. By this definition (Table 1): ● RIs are designed as long-­‐term entities in order to meet the requirements of continuous environmental observation. ● RIs comprise major scientific equipment or sets of instruments, as well as knowledge-­‐containing resources such as collections, archives and thematic data infrastructures and the associated human resources. ● RIs support access and services within their facilities. This access to infrastructures is a crucial part of scientific progress, and enables scientists e.g. from developing countries benefit of state-­‐of-­‐art facilities. ● RIs may be “single-­‐sited”, “distributed”, or “virtual” (the service being provided electronically). RIs mentioned in this document all contribute to the development of a global network of environmental research infrastructures need for a comprehensive integrated research of the Earth system, already envisaged by the Group of Earth Observations (GEO). Developing a EU policy for RIs has been a constant subject of high-­‐level discussions at the national and European level and this stays a priority also for future. At European level, the European Strategy Forum on Research Infrastructures (ESFRI) has been instrumental for supporting decision making on RIs at the pan-­‐European level, and promoting the integration of European scientific research as well as strengthening its outreach capabilities. Most ESFRI infrastructures serve (or intends to serve) a broad community of researchers working in a particular field of research by providing physical, virtual, and/or data access and computational services for the users which has not been previously available to that community or was operating on a fragile project basis. Each research infrastructure has its own particular set of science questions and foci that it must solve to achieve its objectives; however every research infrastructure is also providing its data and services to the wider user communities and thus contributing to the wider, trans-­‐ and interdisciplinary science questions and grand environmental challenges regardless of its particular field of interest. There are many issues that most of the RIs share, for example data collection, preservation, quality control, integration and availability, as well as providing the computational capability to perform the analyses of interest to researchers (or vice versa). Moreover, whilst each RI is separately concerned with the integration of data within its domain of interest, it is also imperative to find robust yet lightweight means to integrate data and computation across RIs to serve an increasingly multidisciplinary scientific community. As an example, the existing infrastructures for atmospheric observation in Europe are highly developed in the different atmospheric domains and topics, while an over-­‐arching structure for higher-­‐level integration is missing. However, a number of issues regarding atmospheric research infrastructures require decisions more efficiently taken at a higher level of integration than at the single infrastructure. While the independent infrastructures focus on their specific research objectives, the implementation of a single interface and approach to the atmospheric domain will offer unique opportunities for joint research activities, mirroring the full complexity and 4-­‐dimensional structure of atmospheric processes. In addition, transnational access and services to the scientific community will be provided efficiently through a single interface to the entire atmospheric domain. The same characteristics and developments applies to the marine, ecosystem/biodiversity and solid Earth domains The current Environmental Research Infrastructures (RIs) in the ESFRI roadmap are represented in Figure 2 with the ones participating in the ENVRI project marked with yellow. The environmental RIs and Integrated Infrastructure Initiatives (I3s) in the environmental sector that have contributed in the development of this strategy paper are: • ACTRIS (Aerosols, Clouds, and Trace Gases Research Infrastructure Network), a ground-­‐based integrated atmospheric observing system for aerosols, clouds, and trace gases. • EISCAT_3D, a project led by EISCAT (European Incoherent Scatter) which seeks to construct a three-­‐dimensional imaging radar to make continuous measurements of the geo-­‐space environment and its coupling to the Earth's atmosphere. • EMSO (European Multidisciplinary Seafloor Observatory), a European network of fixed-­‐point, deep-­‐seafloor and water column observatories. • EPOS (European Plate Observing System), which integrates existing European facilities for solid Earth science into one distributed and coherent multidisciplinary infrastructure. Part of this infrastructure addressing surface and lithosphere processes are well suited for integration with other environmental observation infrastructures. • Euro-­‐Argo, the European contingent of the Argo project, a global ocean observing system comprised of a large network of robotic floats distributed across the world’s oceans. • IAGOS (In-­‐service Aircraft for a Global Observing System), dedicated to global in-­‐situ monitoring of greenhouse gases, reactive gases, aerosol particles and clouds by means of instrumented passenger aircraft. ICOS (Integrated Carbon Observation System), dedicated to the monitoring of greenhouse gases (GHG) through its atmospheric, ecosystem and ocean networks. • InGOS (Integrated Non-­‐CO2 Greenhouse Gas Observing System), dedicated to the monitoring of non-­‐CO2 GHG such as SF6, CH4 or N2O. • LifeWatch, a research infrastructure for studying biodiversity and the Earth’s ecosystems. Dedicated coordination RI projects (e.g. ENVRI, COOPEUS, GEOCARBON, CReATIVE-­‐B), serving as forums for cooperation of the existing environmental RIs. However these are only short-­‐term activities and currently, no overall joint long-­‐term framework exists which looks at priority setting, identification of possible new areas of cooperation, mutual use of global and large scale research infrastructures, or strategies towards the ‘globalization’ of national and regional infrastructures. Figure 2. Current (2013) European Environmental Research Infrastructures in the ESFRI roadmap. Infrastructures marked with yellow text show the beneficiary infrastructures in the ENVRI FP7 Project. The range of different RIs does not yet cover the whole Earth System, and each RI has their own dedicated user groups and stakeholders. Cross disciplinary challenges The Earth System is far too complex and interdependent to be studied from one perspective. The environmental scientific community has used several different ways to slice this complex field to more manageable regions of study. The Earth System is commonly separated into “spheres” or “domains”, mainly dividing the living organisms (biosphere) from the rest, and then using physical phases as the main argument for the separation to atmosphere, lithosphere and hydrosphere (including cryosphere). This traditional view of the Earth system has also been the background for development of many of the Research Infrastructures in the field (Figure 3), although some of the current RIs are more concentrated on the geographical areas (e.g. SIOS, or more generally all European RIs). Another possible way of categorizing is the use of traditional disciplines (physics, chemistry, forestry, biology, etc.), which can operate on several spheres and geographical areas. Similarly, the categorizing can be done by measurement platforms, such as between IAGOS using instrumented in-­‐
service aircraft for atmospheric composition monitoring, and ACTRIS or ICOS using ground based remote sensing or in-­‐situ measurement for similar purposes. These separations are necessary, but they implicitly build their own barriers for scientific information exchange and fracture the understanding of the whole Earth System. Different RIs from different disciplines have different initial user bases, commonly concentrated in the research field they are based on (thick arrows in Figure 3), with less useful information provided for the scientists outside of the field in question or to the society in large (thin arrows). Figure 3. Example of the cross-­‐disciplinary challenges in current (2014) environmental RIs from the Earth System viewpoint1. The RIs only cover a minor part of their respective parts of the Earth System, and provide products mainly for their own communities (thick arrows). Science products and services are not necessarily well suited for users from other communities or from the larger society (thin arrows). 1 in this and following figure “Lithosphere” is to be considered containing all solid Earth disciplines. Figure 4 Different viewpoints of the Earth System sciences. The top-­‐leftmost figure is a schematic viewpoint of scientists, based on the physical divisions. The top-­‐middle figure shows the viewpoint from the common Environmental challenges and the top right an example of the viewpoint of the societal actors. A key idea is to understand that all of these viewpoints are just different ways to look at the Earth System, as all of the possible divisions are strongly interdependent. Figure 4 shows (one of many possible) viewpoints to the Earth System. Where the scientists commonly view the System from the physical and ecological viewpoints (cf. Figure 3), the environmental challenges are typically strongly cross-­‐cutting the traditional domains. Similarly, the societal viewpoint shows examples of needs, which are not answerable from other viewpoints: They are intangibly connected to the whole Earth System and to several environmental challenges. As long as the domains stay fractured, the Earth System “cube” is not complete, and answering the societal and environmental problems remains difficult. Information and knowledge gaps There are also computational and informational barriers to cross. Environmental sciences are already facing a tsunami of data. The amount of environmental science data flow is envisioned to increase tremendously in the near future. As the amount, variety and complexity of environmental data increases, it becomes more and more apparent that the construction of well-­‐
articulated information technology is needed to combine and assimilate environmental information from disparate sources (e.g. field observations, monitoring programmes, experiments, models and simulations). The data tsunami comes with the interoperability challenge -­‐ the processing, transport, storage and combining of variable, and often dynamic, datasets that represent different aspects of our environment at different scales and complexities. As is often the case, it is the comparison and integration of these variable datasets that gives scientists their most powerful tools for innovation; therefore significant gains should be expected when existing data and computation bottlenecks can be overcome. For some disciplines, the data tsunami is still lacking due to label intensive data acquisition or traditional “old-­‐school” individual data collection and archiving. The unexpected discoveries will still stimulate the environmental sciences. They will be more easily obtained thanks to data quality control validated on large scale databases, synthetizing models, and integrated in the global knowledge due to a true cross disciplinary approach. In this respect, environmental research infrastructures are very important nodes of innovations, since they serve as joint working environments for academia, industry, SMEs and NGOs. For all these purposes, collaborative work towards innovation products, downstream services development and on interoperable physical and e-­‐
infrastructure operations has been started among ESFRI environmental RIs, but require many more investments and efforts to be practical. Information and knowledge gaps are not always related to technical issues but there are many other perspectives that may decrease in the applicability of data products and usage of the RI services among the disciplines. The Figure 4 illustrates four categories of the different barriers of information exchange in Earth System sciences: Discovery: Disciplinary, regional and study area boundaries make it hard to find
correct and useful datasets. This can be as simple as not being aware of RIs
operating in specific needed region, or more ingrained, due to cultural research
differences, making it difficult to understand necessary feedbacks within the Earth
System. The differences in terminologies, time and spatial scales can make data
usage very difficult for a scientist from a different discipline;
Access: Open or relatively easy access is strongly required for data usage. The
access requirements also include access to non-scientists. Complicated or
discipline-specific user interfaces and licenses can effectively make the data nonaccessible. Lack of standardized formats, dictionaries, data identifiers and
metadata can create practical problems on data access and thus use;
Understanding: The data user must understand the formats, versioning, language,
metadata, and discipline specific jargon of the produced datasets. The RIs can also
make several assumptions in the data production regarding assumed data use, such
as required temporal or spatial resolution, derived data products or even in the data
screening procedures. This lack of understanding is then not only due to lack of
user understanding but also lack of RI knowledge of all potential stakeholder
groups, and a lack of uniform documentation (e.g. reference model) of all issues
regarding data among RIs
Trust: The data user must trust the data producers and the documentation. The
farther the data user is from the data production community, the more complete
documentation and explanations are needed before the user can be confident that
the produced datasets are usable and most updated. Similar lack of trust of users
from other disciplines can also hinder the activities of the RIs, as erroneous
conclusions can be derived from poorly understood datasets. As long as the data
citation issues are not standardized, also the credit and documentation of the data
use might suffer in cross-disciplinary work.
Figure 4 Processes decreasing RI information usability for a scientist not from the same discipline. Lacks of discovery, access, understanding and trust have several symptoms and root causes. Summary of the state of European Environmental Research Infrastructures •
•
•
•
The European Environmental Research Infrastructure field has been strongly growing via construction of several RIs and by institutional developments within the community. The RIs are envisioned as the long-­‐
term actors in the Earth System sciences, thus also responsible to provide long-­‐term strategic approach. All of the current RIs are based on originally bottom-­‐up processes in individual disciplines and Earth System domains. Cross-­‐disciplinary and cross-­‐RI coordination is seen as important by the RIs, but is not yet followed up in the current cluster projects with adequate resources. Computational challenges are common to all RIs. Data usage barriers are significant, making interdisciplinary science difficult and increasing the cost of answering to Earth System challenges. Problems related to data discovery, access, understanding through reference modeling, and trust remain to be solved. Diversity of the RIs brings great potential for observation of the Earth System, but often also makes it difficult to understand underlying processes and feedbacks. Many of the environmental and societal challenges are strongly connected to several Earth System domains and require integrated multidisciplinary understanding. Vision for the Environmental Research Infrastructures The vision for environmental research infrastructures for 2030 is aiming towards universal understanding of our planet and it’s behaviour. This should result in the evolution of a seamless holistic understanding of the Earth System, an environmental system meta-­‐model2, a framework of all interactions processes within the Earth System, from solid earth to near space. Scientists that within their own science contribute with data, models, instruments, algorithms and discoveries should feel that this serves a greater good, namely a contribution to this understanding. Figure 5. Overview of the Vision for the European Environmental in-­‐situ Research Infrastructures. The current situation of separated domains, RIs and disciplines lead to fractured research and fractured research products. The products and services are difficult to build to answer the Environmental Challenges and Societal needs. By investing to different aspects of the Earth System science community integration, we will gain more holistic understanding of the operation of the Planet. This understanding makes possible to use system science approach, generating flexible multidisciplinary services, answering to environmental challenges and providing key products for societal needs. 2 A term model here is not to mean a computational model, even though that can be one realization. Here the term is meant to describe a self-­‐contained and consistent contextual model, which describes our understanding of the system, its linkages and feedbacks. Envisioned benefits of the Holistic understanding of Earth System The Vision is based on the observed deficiencies in current Earth System sciences as described in the previous chapter. The fractured nature of the Earth System sciences is a natural product due the multitude of disciplines and motivations of the research, and the large scope of the research questions. However, this heterogeneity makes it difficult to approach cross-­‐cutting problems from both environmental and social viewpoints (Figure 5). Finding a holistic common view of the Earth System makes the current disciplinary borders of the observations less strict. The need of a standardized holistic understanding of the Earth System is all the more imminent in the light of present societal questions that arise from Global Policy platforms. IPCC is established to answer to the issues around climate change like rising sea levels, desertification and drinkwater supplies; IPBES follows this development in regard to the threads to biological diversity and the attached risks like food security, medical discoveries and pandemics. GEO/GEOSS touches upon similar and other topics like natural disasters caused by geological phenomena. This holistic understanding will make it possible to approach the entire Earth System from different perspectives, and chose the portions of the whole conceptual understanding which are relevant to the problem to be solved. This approach makes it possible to do new and flexible services, answer environmental and societal challenges. Most importantly this approach is also of aiming to be complete: Any emerging issues can be tackled on the framework of this understanding, enabling tuning and improvement of the understanding and building the connections to other scientific fields, such as social sciences. To understand the Earth System as an interlinked system, we aim for a systems approach. Environmental Sciences are rapidly moving to become one system-­‐
level science, mainly because modern science, engineering and society are increasingly faced with complex problems that can only be understood in the context of the full overall system they belong to. There are several reasons and enablers for this shift: • Technology push. Technology innovations on, for instance, detectors and sensors with ever increasing resolution, allow deep observations of scientific phenomena important for the better understanding of a whole system. Connection between these new observations to the whole Earth System requires ways to integrate between the domains. In addition, information technology innovations, such as digitalization of collections, also unlock resources at a systems-­‐level. • Demand pull. The questions scientists are faced with nowadays (not only arizing from curiosity, but also from policy like IPCC, IPBES and GEO/GEOSS and societal needs) can simply not be solved using the traditional sources of information. Without access to information from adjecent disciplines, the answers scienstists can give will increasingly be partial and incomplete, an therefore less ground breaking or even useful. •
Globalization. Like in economics and society, science is experiencing an up-­‐scaling due to globalization. Establishing and managing big data and information repositories often demand an international effort. This can also be observed from the ever-­‐increasing aggregation of research funding, such as ESFRI. •
Resource integration. Never before did researchers of so many domains have such a wealth of resources at their disposal. The integration of these worldwide available resources has further fueled system-­‐level research. An important contribution of e-­‐Science as a system-­‐ level science is its potential for integration of information. •
Science Integration. These developments in science in general offer an excellent opportunity to approach the Earth as an integrated system, with an eye out to related sciences, such as social sciences and life sciences. Strategy for integrated holistic Earth System understanding The vision requires understanding and a conceptual model that is capable of providing a definitive answer, which is reliable and credible. However, building such integrated view is not straightforward. Development of this view requires resources, and we identified 3 interdependent resource capitals which need to be improved to gain this vision: Technological Capital: Capacity to measure, observe, compute, and store. This requires materiel, technologies, sensors, satellites, floats; software to integrate and to do analysis and modeling and processing; building observational, computational and storage platforms and networks. Cultural Capital: Open Access to data, services etc from other RIs. This requires rules, licences, citation agreements, IPR agreements, technologies for machine-­‐machine interaction, workflows, metadata, data annotations etc. Goal should be to contribute always to the standard understanding, the systems approach; RIs work together as a community on the policy level. Human Capital: specialists to make it all work. This requires data scientists rather than discipline scientists. We also needs ‘generalists’ that oversee more than just their own discipline; Describe requirements for the future environmental scientist; How to include and train citizen scientists. It should be noted that all of the actions proposed below are examples, and might not be suitable for all RIs and disciplines. Technological Capital development The Environmental science research community has much to gain from the rapid advance of technology in many areas. The research infrastructures on the field are mostly focused on collecting and acquiring observational data, which can be used to gain comprehensive understanding of the Earth System. The technology used for acquiring, storing and processing the environmental research data includes instrumentation such as sensors, floats, radars and integrated measurement stations; as well as e-­‐Infrastructures (in their wide meaning), including networks, computing servers and storage, together with software components enabling the processing of this data. The necessary developments on these areas of technology are described in the following subsections. Instruments and measurement station development Tremendous amounts of measurement data on environment is collected globally each day and night, in many cases operated by the current RIs. Many sensors work automatically 24 hours a day and 7 days a week; others provide large amounts of data during campaigns, either by sensors or by human observation. However, there is still need and potential to increase the data streams significantly. Sensors with more resolution and accuracy will become available. The measurement station network will need to be extended to cover areas of the globe that have currently few such capacities, mainly areas outside Europe, North America, Japan and Australia. Additionally, novel types of sensors and platforms (e.g. distributed or independent platforms) are going to become available as technology advances, resulting in more variables measured from atmosphere, oceans, geosphere, and other parts of the Earth System. In some cases, even the need of specific kinds of sensors is not yet known. Another development foreseen is the developing multidisciplinary and comprehensive stations to measure multitude energy and material flows between and within the Earth System domains. The advantage of an integrated station is that it can produce data on multiple phenomena at the same time and location, thus allowing to study correlations and interdependencies between variables in great detail. Altogether these developments will produce research data that is more accurate and more widely describing the environment than before, and it will be collected from multitude of locations around the globe. The system-­‐level science towards the Vision will greatly benefit from such resources provided that matching capacity of data management and processing technologies are made available. Example actions related to this process Enlarging and developing the current RI coverage of the planet Actor: Current and future RIs. Outside-of-Europe partners Description: Enlargement of current RIs in the fields where this is necessary. Contains
issues, such as improved network densities, inclusion of new instrumentation,
independent observing systems (e.g. automated independent systems), sensor
miniaturization, and cross-national support actions Needs: Gap analysis for prioritization, improving global cooperation, global demand for
larger or integrated Environmental RIs, enhanced automatic, remote data transfer
and data receiving, enhanced development of hardware and software of automatic
measurements Benefits: Far advanced data collection of Earth system, also outside the developed
countries Supporting cross-use of experimental research platforms and vessels towards understanding
the gaps of our knowledge
Actor: Current and future RIs, research organizations, universities Description: providing multidisciplinary, cross-RI community access to campaigns towards
finding out the key gaps in our knowledge, and generating recommendations for RI
development. Needs: Coordination of cross-RI campaign programmes Benefits: Cost-efficient provision of the comprehensive datasets and RI access to the
scientists
Supporting scientific projects towards understanding the gaps of our knowledge
Actor: Current and future RIs, research organizations, universities Description: Supporting scientific projects by direct participation or acting as a stakeholder.
These projects are aimed towards finding out the key gaps in our knowledge, and
generating recommendations for RI development. Needs: Development of funding models and programmes Benefits: Concrete recommendations towards filling the key gaps in knowledge
Feasibility analysis of measurement station integration between RIs
Actor: Current and future RIs Description: Analysis of which measurement stations could be used by several RIs Needs: Gap analysis for prioritization, international demand for larger Environmental RIs
(Dissemination of Standard Model idea) Benefits: Synergy benefits, and possible interdisciplinary co-location benefits
Development of new types of sensors and observation technologies
Actor: Current and future RIs, research organizations, universities, technology companies Description: Development of new measurement technologies based on research needs and
technology advances Needs: Collaboration between academia and industry Benefits: Novel measurements revealing information on the Earth System
e-­‐infrastructure components Complex system-­‐level science can not be realized with the available concepts, methods, tools, and infrastructure that were designed and developed to meet the needs of traditional pursuit of science. System-­‐level science demands new innovative versions of all these elements. e-­‐Science is the science paradigm that enables modern system-­‐level studies. It uses computer techniques to handle and harness the exponentially increasing amount of data, as well as the complexity arising from these system-­‐level studies. It also addresses the important aspects of multidisciplinary approach and allows remote collaboration. e-­‐Science develops the required concepts and methodologies, implemented via software tools which can be applied as part of an ICT based research infrastructure for system-­‐level science in virtually all science domains. The strength of the e-­‐
Science approach is that it supports real world and simulated experimentation plus the powerful combination of both. At the same time, however, e-­‐Science enabled system-­‐level research and the corresponding ICT based research infrastructures have become so complex that they can only be realized by multidisciplinary collaborations, often dispersed all over the globe. Hence, developing the necessary integrated infrastructure is a challenging research problem in itself that can only be tackled by multidisciplinary research teams. e-­‐infrastructure in this context can be defined to mean ICT capacities that enable eScience approach to system-­‐level science. It can roughly be divided into hardware and software components. Advances in instrumentation require significant increase and development in e-­‐infrastructure capacities in order to utilize the scientific potential of the collected data. A starting point is that fast network connections are needed at the larger measurement stations in order to transfer the data to a storage facility. At some integrated measurement stations or large instrument sites a hierarchical approach to data transfer is needed so that real-­‐time automatic processing is done on the data before it is feasible to transfer it to a central facility. Also large storage capacities with large writing bandwidth are needed to store the measurement data. The data should be stored as much as possible to open repositories, and supplemented with descriptive and machine-­‐readable metadata. The research data is made discoverable by various metadata catalogues that are globally interlinked and connected. The data should be easily available for high-­‐performance computing (HPC) systems for processing and analysis. The vast amounts of data need next generation computing power for processing, in some applications parallel supercomputers with fast interconnects are needed, and some cases apply massive amounts of independent computations. Besides hardware more advances in software and ICT services are needed to reach the vision. Modeling and data analysis software need to be able to analyze much larger datasets and utilize much larger computing systems. Also intuitive and easy-­‐to-­‐use interfaces for ICT capacities are needed, as the complexity of the systems increase, for the full user community to be able to benefit from increased capacities. For the same reason portals, workflow tools, Virtual Labs and Virtual Research Environments will become more important. The access to the e-­‐infrastructure resources would be best provided through flexible and scalable services that may be tailored towards the needs of specific user communities. The service based approach puts the focus on the users – the researchers – and not on the technology. The backend technology should be run on modern, flexible, cost-­‐effective and electricity-­‐effective computing systems over a high-­‐speed network. The technology behind the services is provided as open source to enable RIs as well as other parties to investigate and modify it as needed. The development and provision of e-­‐infrastructure and eScience components should rely on a heterogeneous landscape of service providers: sustained international e-­‐infrastructures, computing centers, and also SMEs. Finally, the infrastructures both on instrumentation and ICT levels should become globally integrated. Interoperability between regions and between disciplines is crucial for system-­‐level science. For example connecting repositories with open interfaces to data catalogues will give users the possibility to discover data from various locations and infrastructures from one entry point. To accomplish interoperability steps towards common standards are needed, such as agreeing on a joint reference model to describe infrastructure components. The technological capacities need to be open for new research infrastructures to start providing and utilizing common data. Example actions related to this process Establishing fast network connections and methods of on-site data reduction for
environmental RIs
Actor: Current and future RIs, IT experts and service producers, National Research and
Education Network (NREN) providers, Description: New fast network connections for selected sites, building of on-site data
reduction methodologies and infrastructure Needs: Measurement systems providing large datasets Benefits: Fast and economical system for Earth System observation data distribution
Building of large storage capacities for RI use
Actor: IT service producers Description: Establishing the IT infrastructure needed for the data storage and retrieval Needs: Standardization of IT requirements, metadata and data identifiers, a common
reference model Benefits: Necessary storage requirements for data providers and quick access possibilities
for data users
Development of software solutions for Environmental Big Data applications
Actor: IT service providers, universities Description: Development of software solutions for effective data management, discovery
and retrieval Needs: Standardization of IT requirements, metadata and data identifiers Benefits: Effective and affordable solutions for big data applications, reduction of need for
Easy-to-use interfaces for e-science use in Earth System RIs
Actor: IT service providers, GEOSS Description: Generation of web-based analysis tools for holistic Earth System data analysis,
in co-operation with international and global operations (such as GEOSS) Needs: Standardization of IT requirements, metadata, data formats and data identifiers Benefits: Enabling data users for all disciplines to effectively use the datasets. A crucial
requirement for system approach.
Generation of standard APIs and services for Earth System Sciences
Actor: IT service providers, RIs, Scientific community, RDA Description: Standardization of IT requirements, APIs, metadata, data formats and data
identifiers, a common reference model Needs: Requirements from the RIs and data users Benefits: Enabling easy further use of the datasets, requirement for many advances
Building competence centres of Earth System data analysis
Actor: IT centres Description: Establishment of competence centres, which operate to development of
methods for Earth System data analysis and train users Benefits: Long term centralized development of environmental data analysis, training of
new specialists
Cultural Capital development The cultural capital development describes the advances in non-­‐material aspects of the research infrastructures and research landscape. The research infrastructures do not only consist of physical instruments, produced data or people handling them. A significant -­‐for some cases the main-­‐ part of the research infrastructure is the development of methods, policies and ways of thinking needed for successful research endeavors. Without such developments, the goal of having integrated and complete understanding of the Earth System cannot be realistically achieved. The success of e-­‐Science also heavily depends on the ability and willingness of researchers to accept a cultural shift in the way they undertake science. It is imperative to structure the e-­‐Science process along the full technology chain in order to clarify the role of the various activities and stakeholders involved (physical instrument engineers, computer and computational scientist and engineers, application domain end-­‐users). The vision is also directly connected to the key barriers within the current methodologies of Earth System science described in the previous chapter. The acceptance of the need of system understanding will implicitly require common discovery tools, including standardization of language, and understanding of different scales of processes. The access to the knowledge and datasets must be open to all scientists in the system analysis, including necessary information to use this knowledge efficiently. The trust between different disciplines is built on this common goal towards common understanding, helped by standardization efforts and increased communication. Disseminating the idea of a common goal in Earth System sciences The key cultural development is the acceptance and dissemination of the overall vision of this strategy. Building such environmental literacy is absolutely necessary not only to RI personnel, but also to the whole Earth System science community, funding agencies, citizen scientists, science-­‐publishing sector, industries, and to the society in general. This process is crucial to the success of the strategy. Example actions related to this process Publication of the strategy and goals of the European Environmental Research Infrastructures
Actor: RIs, Earth System scientists Description: Publication and dissemination of the strategy: White papers, opinion papers
and presentations in national and international meetings. Consultation and direct
communication of the Vision to the relevant funding agencies and to the general
public. Needs: Acceptance of the Vision by the RIs Benefits: Enabling the overall guiding vision for the Earth System sciences.
Building the culture of open research The standardized understanding of the Earth System is as good as the datasets, models and understanding available for it’s creation. Restricted datasets and models do not facilitate scientific work, and can make it harder to get the getting overall picture of the system affecting the Earth system. The culture of open research does not happen by itself. The scientists must have clear and effective benefits for giving their datasets and methods to open access. The progress is similar to the start of the scientific revolution: Opening up the results of research lead to huge increase in scientific progress, but require the scientists to trust that their contributions will be properly attributed to them. Building this trust is a key to the change in the research culture. The benefits of sharing the datasets must be made clear by the funding agencies and should lead to direct professional development of the scientists. This requires first of all a proper and widely acknowledged way to cite datasets for their use, and a way to get credit for these citations, for both to show the importance of scientific work to the funding agencies, and to show the quality of the individuals when deciding on career progress. The Environmental Research Infrastructures must be in the forefront of this progress, and actively implement the citation methods and authorship follow-­‐up in their e-­‐science contributions. They must also implement the data publication record as a crucial part of career advancement, and support similar actions in their respective scientific disciplines. Example actions related to this process Requirement of open data access as the standard clause in any public science funding
Actor: Funding agencies Description: Requirement of data openness for publicly funded studies with exceptions
only in very rare cases Needs: Pressure from the scientists Benefits: Making open data access the norm in Earth System Science
Common and widely used data citation mechanisms and citation indices
Actor: Journals, data citation index managers, data centres Description: Effective data citation for both credit and reproducibility purposes Needs: Data centres with applicable resources Benefits: Necessary minimum requirement for efficient data sharing culture
Taking data publication and citation as a key parameter for merit determination in RIs Actor: RIs, funding agencies Description: Data publication and citation will be considered as one of the key products for
scientists and data scientists Needs: Data citation mechanisms, citation indices Benefits: Acts as a powerful enabler for data publication
Standardize the language and terminology The increased collaboration and cross-­‐cutting science also requires that the scientists from different disciplines speak the same language. Terminologies of the different sciences need to be harmonized, or at least well documented, to enable easy use of datasets across the different user groups. Data discovery and correct use is strongly dependent on the users’ ability to understand the context of the datasets. The development of dictionaries, ontologies and standardized terminology are specific tasks, but the change must come from the scientists themselves, who need to understand the need of providing their data with documentation in the common terminology, and the need to understand the terminology to find correct information. The domain-­‐specific terminology is not expected to disappear quickly, and thus the use of standardized names is a long process. A good example of such advances is the NETCDF CF approach, where the standardized terminology for climate model applications is created as a community effort. Standardization is needed for the descriptions of systems as well. The differences on the terminology of research infrastructures can lead to misunderstandings and barriers on cross-­‐infrastructure co-­‐operation. Example of such standardization is the ENVRI Reference Model, which describes the processes in environmental research e-­‐infrastructures from data collection to discovery and user feedback in an understandable, common format. Developing such approach also for the scientific work connected to the ‘Standard Model of the Planet’ is crucial. Such reference model could then give the terminology for roles, processes and actors in such research, giving a transparent standard understanding of the research activity. This would give a common way to understand the workflows of the scientific activities is important to understand the scientific thinking and possible limitations of different datasets. The data handling and distribution culture must also take into account machine readability. The documentation and licensing system must include directly machine-­‐readable part to make sure that the data discovery and collation tools can correctly state the usability of the data for the research purposes both from the scientific and from the usage rights points-­‐of-­‐views. Example actions related to this process Harmonisation of terminology within the Earth System Sciences Actor: Earth system science community, RIs Description: Standardization of terminology for metadata descriptions and process
descriptions, including explanation of uncertainty Needs: Community effort (e.g. within RDA)
Benefits: Discoverability, improved understanding
Establishing standards for scientific data including contextualization Actor: Data journals, IT centres, scientific community, RIs Description: Standardization of scientific publication, use of accepted data centres for
storage and dissemination, standardization of metadata and other contextualization Needs: Efficient data citation standards, data storage capacities, metadata standards and
requirements, community participation Benefits: Simple ways for the researchers and RIs to publish datasets and gain benefit for
citations: clear rules of data usage and citation
Development and adoption of ENVRI reference model to map RI e-­‐infrastructures Actor: IT service providers, RIs Description: Further Needs: Requirements from the RIs and data users, implementation tracks in all Ris related
to environmental science Benefits: Standardization of terminology for environmental RI data production, ways to
document processes and workflows, development tools for new RIs
Developing and implementation of machine readable documentation and licensing standards for Earth System data Actor: IT service providers, RIs, Scientific community, RDA, license communities (e.g.
Creative Commons) Description: Standard machine readable licensing, and documentation of the datasets Needs: Requirements for license and documentation standards Benefits: enabling combining datasets together with the license requirements and
automatic inclusion and collation of documentation Enlarging the view The Group of Earth Observations (GEO) and associated Global Earth Observation System of Systems (GEOSS) are crucial coordinating efforts in building the Standard Model of the Planet. They facilitate direct access of Earth System data, mostly concentrating on the policy aspects. The GEO/GEOSS activities can form the core or at least a strong part of the data discovery and standardization requirements, especially on global scales, of the holistic understanding of the Earth System, and the collaboration of the Environmental RIs with them is of crucial importance. Understanding the feedbacks and processes in the Earth system can not only work via observations and modelling of the natural systems. Inclusion of more direct anthropogenic influences (anthroposphere) to our understanding is of great importance, as the human influences on all aspects of Earth System are growing. This approach can lead to far better estimates of behaviour of the Earth System as whole. Example actions related to this process Integration efforts between Earth System in-­‐situ observations, remote sensing communities and Earth System models Actor: Earth System science communities, Satellite observation communities, Modeling
Communities Description: Solidifying the interaction between different kinds of observations and
modeling communities – work towards integrated view of the Earth System
interactions. Needs: Connection and trust building. Data infrastructures and data scientists. Benefits: Directly benefiting the vision.
Integration efforts towards including human activities on the Earth System analyses and databases Actor: Earth System science community, Social Sciences community Description: Establishing the connection of social science methodologies, models and RIs
and Earth system observations to holistic view of human-environment interactions. Needs: Connection and trust building towards social sciences community Benefits: Holistic view of human-environment interactions – taking human activities as
internal part of the model of the planet
Organizational framework The European Environmental RIs need to have a sustained method of co-­‐
operation, coordinating the development, methods and policies between the Environmental Research Infrastructures. There are many benefits for such co-­‐
operation. Working together makes it easier to provide prioritization for the developments, common strategies, develop necessary critical mass to influence outside communities, enable communication between the RIs and the scientific disciplines they represent, and to coordinate the actions needed for the Standard Model development. This coordination structure should include members from all RIs developing the ‘Standard Model of the Planet’, and include in some way the representations of science, policy, and societal stakeholders. The coordination structure can also support and give guidance to new research infrastructures, in Europe and beyond. The efforts required for the ‘Standard Model of the Planet’ building need motivation from the scientific community, funding agencies and policy makers. To achieve this kind of common interest, a significant effort need to be made to make the need of a ‘Standard Model of the Planet’ known. This achievement requires a convincing storyline, disseminated via publications, education efforts both for old and new scientists and direct communication to funding agencies. The role of the suggested coordination structure is manifest in such activities: A strong common voice of the need of these improvements is a strong asset and can lead to common and wide acceptance of the overall holistic view of the Earth system, in all of it’s components and processes. Example actions related to this process Establishing coordination structure of Environmental Research Infrastructures Actor: Environmental Research Infrastructures Description: Finding semi-permanent way for communication between the environmental
RIs, e.g. by founding a coordination board Needs: Willingness for collaboration, funding Benefits: Possibility of strategic development, common voice
Dissemination of the Standard Model of the Planet idea to Scientists, Funders, and end user groups Actor: RIs, science community Description: Dissemination of Standard Model of the Planet idea widely to key groups
within the environmental communities. Building the culture of system understanding
to all Earth System sciences Needs: Development of a convincing storyline. Active participation of RIs and scientists on
dissemination, co-operation between RIs Benefits: Finding common cause across the Earth System sciences, far improved
understanding of the planet, establishing true interdiciplinarity.
Human Capital Development Curriculum development for data scientists, connected to disciplinary science A complete new type of data specialists will be needed, concentrating on producing and maintaining the data products of the RIs. Earth Science data specialists have a strong information technology background, but are well versed in the scientific questions and methodologies of the ES sciences. They are capable of handling large datasets, understanding the limitations and benefits of different Earth system observations and experiments and are also responsible on the development of IT infrastructure of the RIs. As they are concentrating on the data production and data assimilation, their career paths include far more data publications than other types of publication, and thus their positions are best developed using the data citation services. The career paths of these specialists require completely new positions recognizing their unique services to the science community. Big Data and Data Intensive Science will require a new type of skills and knowledge that the future specialists: Data Scientists, Data Analysts or Architects should have to effectively operate complex infrastructure and data processing (analytics) applications through all stages of the Big Data lifecycle and deliver expected scientific and business value to science and industry. Researchers, software and infrastructure engineers recognize that Data Intensive Science/technologies to support the idea of the holistic Earth System understanding will constitute a mandatory knowledge and skills for the future science, industry and other human activity domains. The changing methods of (digital) science and research require that researchers, professors and students receive adequate support in computing and networking, as well as in handling, analysing and storing large amounts of data and content. This is the scope of work of the emerging professions of e-­‐infrastructure developers, integrators, operators, research technologists, data scientists and data librarians for which formal education hardly exists today. Professional recognition of this community and the development of appropriate curricula, training and skills are crucial to ensure effective services to institution staff and students. Training opportunities should be available at all levels and for all communities potentially engaged in research related activities. One could think of the following activities to include in curriculum development: • Support the establishment of the profession as a distinct profession from that of a researcher and a traditional computer specialist/engineer. Create a reference model and a Common Body of Knowledge (CBK), which defines their competencies, supported by case studies, and best practices relating to e-­‐Infrastructure skills, human resources management, support tools and related institutional practices. • Define or update university curricula for the e-­‐infrastructure competences mentioned above, and promote their adoption. • Support networking and information sharing among already practicing e-­‐infrastructure experts, research technologists, data scientists and data curators working in research institutes and in higher education. Example actions related to this process Establishing the career paths for Data Scientists in Earth System Sciences Actor: Universities, research institutions, funding agencies, RIs Description: Create career paths, including incentives, rewards, goals and performance
statistics. Founding of positions for different parts of Data scientist careers. Needs: Education of Data Scientists, data citations Benefits: Create professional and motivated Data Scientist profession
Establishing education and training requirements for Data Scientists Actor: Educational institutions Description: Standardized requirements for a Data Scientist curriculum Needs: Career paths for Data Scientists, knowledge of key skills required Benefits: Create professional and motivated Data Scientist profession
Need for a wide enough perspective outside the own discipline. The Environmental sciences produce and need new specialists for their work. Earth System scientists in environmental sciences have a scientific career concentrating on the boundaries and communalities of the current disciplines and Earth System domains. They give the necessary scientific advances on the larger scale issues of the Earth System, or from phenomena and processes affecting several of the traditional fields of science. They are often generalists, operating on large or small-­‐scale interdisciplinary efforts, combining the methodologies and study areas from several fields, including social sciences and information technologies. They depend on common language, ontologies and policies between the Earth System observation and experiment research infrastructures, and will benefit from the far better interoperability between the traditional study areas than today. The training of these specialists can be originally from specific disciplines, but can also be based on multidisciplinary approaches from start up. Their career paths follow the traditional scientific researcher careers, although their cross-­‐disciplinary nature can require new kinds of positions. Example actions related to this process Establish curriculum and programs for cross-­‐disciplinary studies Actor: Education institutions Description: Establishing requirements, curriculum and programs for cross-diciplnary
expert training Needs: Common acceptance of cross-disciplinary approach to Earth System science, data
availability, and information on key skillsets needed
Benefits: New scientist types, which are directly capable for holistic studies of complex
Earth System analysis
Geographical and interdisciplinary mobility programs Part of the training for the new breed of scientists that have the 'Model of the Planet' as their overarching theory of things is the capability for them to move around within the various disciplines that constitute the Earth Sciences. A sophisticated network of RI facilities both geographically and interdisciplinary is necessary to provide the Earth System scientists an environment to move around without limitations. The I3 system as developed by the European Commission could serve as example, but should be implemented on a permanent basis. Example actions related to this process Establish cross-­‐diciplinary mobility programs Actor: Funding agencies, scientific communities, RIs Description: Ways for easy mobility of Earth System scientists between institutions, RIs
and disciplines Needs: Acceptance of the need for cross-disciplinarily, training of specialists capable of
holistic view Benefits: Advancement of cross-disciplinary studies, better understanding of Earth System
in whole, efficient usage of RI products
Serving the citizen scientists Environmental Science relies at least partly on the efforts by citizen scientists. These often operate without any organizational infrastructure behind them. Nevertheless they generate and maintain valuable datasets and constantly increase the amount of data. The power of crowds cannot be underestimated in this sense. To be able to fully profit from these data, the RIs need to supply these volunteers with the possibility to store, curate, maintain and improve their datasets. Quality can only be guaranteed if the Environmental RIs create the right environment. This is not only a technical challenge (which is being tackled in many places around Europe already), but also a cultural capital issue, when it comes to citation, credits and even monetary value of the collections. Establish the methodology and practices for Citizen Scientist interaction with the RIs Actor: IT service providers, RIs, Scientific community Description: Development of basic methodologies for Citizen Scientist participation in
scientific process, two-way interaction with the society Needs: Development of culture of science participation, legal and practical challenges Benefits: Participation and commitment of large parts of society to building of Model of
Earth system
Final words Connection to other sciences, regions and infrastructures Although the RIs preparing this strategy and vision are representing a large fraction of in-­‐situ Earth System observations within the European region, the view of the field is of course not complete. There are several Earth Science and other science activities not directly involved in the preparation of this Strategy, and their actions should be integrated within the actions described in this document. • IT sector, and relevant developments in the other science fields are strongly involved in the actions required for the capital building discussed in this document. Actions by the European IT e-­‐infrastructures, projects and initiatives such as EUDAT, Helix Nebula and EGI.eu are crucial in these developments, as they can provide common and affordable solutions to the challenges presented in this document. Similarly, on the global scale, the data discovery and cultural issues are approached in initiatives such as CODATA and Research Data Alliance, where the Earth System sector activity is strongly needed. • Satellite Remote Sensing observations of the Earth System are currently leading many of the data integration and Earth System observation collaboration actions. ESA is developing a set of Thematic Exploitation Platforms (TEP), which can provide new inventive ways to handle large data access and handling problems. The idea underpinning TEPs is to facilitate data access and exploitation by moving the scientist “desktop” (and associated software) to the “data”, instead of other way around. This kind of virtualization and access could be extremely useful for the overall development of this vision. • The Earth System modeling community in Europe has published their own strategy for the future, where they concentrate on improving the spatial resolution of the modeling systems. This would also mean that more of the currently less-­‐usable in-­‐situ Earth Observations could be integrated with the modeling efforts – strengthening the view presented in this document. Similarly, the connection between models and observations is powerfully presented in the forthcoming Copernicus initiative, which is directly useful for the overall vision presented here. • Regions outside of Europe have their own actions and strategies for Earth Observation and understanding. However we feel that the overall scientific ambition and vision presented in this document is general and globally needed. All of the actions suggested can and should be generalized to other regions all over the planet. Current activities, such as COOPEUS, CReATIVE-­‐B or in larger scale, Belmont Forum, could be the starting points for this process. • The inclusion of social and human sciences is clearly one of the avenues to actually answer the societal challenges. Building the connection to these communities needs concrete and consistent actions and support from the RIs, funding agencies and individual scientists. Summary and view to the future We present in this document the vision towards a common and well understood Earth System. A framework which is flexible, and makes possible to tackle the challenges of the future from their point of view, regardless of the purpose or original problem why certain infrastructure or facility was originally found. We also present wide range of suggested actions towards this goal. The timeline of these actions is dependent on the prioritization of the needs. The progress must start from the overall acceptance of the vision, and then work in parallel in all categories of development: technological, cultural and human. Some of the actions are clearly interdependent, creating critical paths towards a common goal. Overall the higher-­‐level priorities should then be based on following overall categories 1) Acceptance of the vision, the aim towards a system understanding of the whole Earth System; 2) Establishing the cooperation and communication methods and infrastructures for RIs and other active participants needed for the vision; Dissemination of the vision to stakeholder groups; 3) Work towards the key requirements for system understanding, in cultural, technological and human capital building; 4) Evaluation of the progress within the Earth System understanding, key results and societal benefits; The actions presented in this document are examples of the kinds of actions needed, and should not be considered a complete or needed for all Earth System disciplines. They do however suggest commitment of new resources to the Earth System observations, but would also make the current investments far more useful for the society at large, generate top-­‐level scientific results and facilitate answering to key global challenges. The selected vision should also be considered as a tool to determine the usefulness of current and future investments in the Earth System observation, modeling and research.