7
Management Models for Future Seismological and Geodetic Facilities
The second day of the workshop focused on future seismological and geodetic facilities and the scientific advantages of different management structures that might be applied at them. The workshop planning committee once again preassigned workshop participants into one of four moderated breakout discussion groups (see Session 5 in the workshop agenda, provided in Appendix B). Two of the breakout groups discussed management structures that would be appropriate for distributed or partially distributed seismological and geodetic facility capabilities, and two groups discussed more centralized management structures for seismological and geodetic capabilities for which facilities might be combined. A moderator and a rapporteur were assigned for each group. Each group received prompting questions appropriate to their topic. Workshop participants were asked to focus on the scientific advantages and disadvantages of independent or centralized seismological and geodetic capabilities while also exploring management and decision-making structures that could be flexible in the face of new science and technology developments. Participants then reconvened in plenary, and rapporteurs provided their summaries.
This chapter is divided into three sections. The first section describes the summaries from the breakout groups that discussed management structures for independent seismological and geodetic capabilities. The second section describes the summaries from the groups that considered centralized seismological and geodetic capabilities. The third section presents the themes discussed during the plenary session.
MANAGEMENT STRUCTURES FOR INDEPENDENT SEISMOLOGICAL AND GEODETIC CAPABILITIES
Two of the breakout groups were assigned to consider management structures suitable for supporting the capabilities of independent seismological and geodetic facilities. Box 7.1 lists the prompting questions provided to the groups. Summaries of the key themes identified by group rapporteurs are provided in the sections that follow.
Summary of Breakout Group 1 Discussion Themes
Donna Whitney, University of Minnesota
Donna Whitney’s report of her breakout group’s dialogue is divided into discussions of the advantages and disadvantages of distributing capabilities among multiple facilities, and a discussion of how distributed management may or may not allow flexibility to respond to scientific and technology developments. Whitney stressed repeatedly that her group’s discussions were about hypothetical rather than about existing seismological and geodetic facilities.
Advantages to Distributing Capabilities Among Multiple Facilities
Whitney reported her breakout group’s discussion on the advantages and disadvantages of distributing capabilities among separate seismological and geodetic facilities. In terms of instrumentation, she said that distributed facilities allowed expertise to be spread out and more people trained in various instrumentation and techniques. It might also allow for better integration of skills associated with instrumentation and data management if the same personnel were responsible for both parts of the operation. Whitney reported that this would advance science, and might allow for greater flexibility and adaptability for instrumentation. In terms of user support services, Whitney reported that the distributed model may be the best way to support users because domain experts from the facilities could work with specialized user groups. This might lead to more innovation. Likewise, the group discussed that it may be advantageous to distribute capabilities among multiple facilities if the data are particular to different user groups, or if there is a need to focus on a particular type of data.
According to Whitney’s group, education and outreach and workforce development are enhanced through distributing capabilities among multiple facilities in part because there would be more facilities, more people, and more scientists connected to these activities. The number and types of activities could be increased, as could the opportunity to reach different audiences. Workforce development might be enhanced because there is a greater spread of specialized skills and highly technical knowledge. Whitney described that there could be more entry points for employment, and opportunities for both stability and portability for the workforce. Capabilities distributed among multiple organizations might offer multiple opportunities to broaden participation in science among, for example, under-represented groups.
Disadvantages to Distributing Capabilities Among Multiple Facilities
There is increasing overlap of interests and research within seismology and geodesy, Whitney reported. She described how distributed capabilities might result in a silo mentality that inhibits the potential development or use of instrumentation that could be applied to research in both fields. Distributed capabilities might also result in a lack of cross-trained specialists. Distributing instrumentation capabilities may create obstacles to integration and collaboration, although Whitney noted that there are opportunities for integration and collaboration among facilities. Increased competition between facilities for funding might be an advantage or disadvantage, she reported. Hypothetically speaking, the biggest disadvantage to distributing user support capabilities among facilities is likely a duplication of efforts, which could inhibit sharing and may result in a lack of standardization.
Whitney noted that distributed systems may not recognize synergistic opportunities for enhanced education and outreach activities, but she said the group also discussed how this might be true for any facility management model.
Flexibility of Distributed Capabilities
Whitney reported that the current management structures for seismological and geodetic facilities are quite flexible. The community-based approaches allow for scientific agility and opportunity to focus on multiple science priorities. Distributed capabilities among facilities may also allow for greater flexibility to compete or re-compete for funding. However, her group noted that funding agencies have multiple facilities to administer. Her group wondered if separate organizations take full advantage of opportunities to collaborate on their increasingly overlapping portfolios, but she observed that separate facilities do not preclude cooperation among facilities. She cited the current cooperation between Seismological Facility for the Advancement of Geoscience (SAGE) and Geodetic Facility for the Advancement of Geoscience (GAGE) in data management as evidence.
Whitney concluded with her group’s final thoughts: there are two well-functioning organizations that could evolve as needs arise and as scientific and technological capabilities emerge. Change could start with governance. A simplified, unified executive governance structure could enable and enhance the positive attributes of a distributed system.
Summary of Breakout Group 2 Discussion Themes
Bill Dietrich, University of California, Berkeley
Bill Dietrich’s group considered the scientific advantages and disadvantages of distributed capabilities in light of the differences between seismology and geodesy. His group did not discuss SAGE and GAGE specifically, but did use those facilities as examples.
Advantages to Distributing Capabilities Among Multiple Facilities
In consideration of instrumentation, Dietrich reported his group’s observation that seismological and geodetic instruments have different bandwidth, their site locations are different, there are differences in resolution at depth, and different expertise are involved. He noted areas of convergence—geodesy is aspiring to have dense deployments, but geodesy has different timescales of observation than seismology, different error and noise sources, different field practices, and different lengths of deployment. Although geodesy and seismology have overlapping elements in the questions to which they respond, he said, the science questions are very different.
According to Dietrich’s group, distributed facilities offer advantages for user support services because there is a healthy competition among centers that can lead to better support. For example, response to user requests is timelier because the facilities are smaller and more streamlined. Dietrich reported that Robert Detrick’s “dirt-to-desktop” concept (see Chapter 3) was compelling. It was desirable to have the same group of people managing the data from the time of its collection, to its storage, to making it available to the user. Logistically, distributed systems are advantageous because the data format requirements are different for different groups, and the expertise necessary for archiving those data may differ, as may the information technology security issues. Data products may be tailored for specific communities, and tools for distribution may be specialized, he said.
Dietrich reported on the value of having distributed data centers, and having them compete for funding in a manner similar to the Earth Observing System Data and Information System (EOSDIS) discipline-oriented distributed active archive centers (DAACs) model (see Chapter 4). He said that distributed data centers may offer proximity to various research institutions. Centers could have more access to specialized information by working more closely with scientists. The distributed data centers might offer advantages for education and outreach by increasing the number of organizations that reach the general local public, and by increasing opportunities for graduate students to engage directly. A greater number of domain scientists might have greater access to education and outreach professionals at the centers and achieve broader impacts.
The advantages of a distributed system in terms of workforce development include greater ability to train domain-specific specialists, Dietrich reported. The group noted that distributed facilities also make it easier to develop relationships with potential employers.
Disadvantages to Distributing Capabilities Among Multiple Facilities
According to Dietrich’s report, disadvantages of distributed capabilities may include a lack of standardization of instrumentation and missed opportunities for cross-fertilization
among scientific communities. There is potential for duplication of data management hardware (and cost) and duplication of information technology security resources. Redundant paperwork and increased bureaucracy were also described as disadvantages to distributed facilities.
Scientists working at the intersection of two facilities must work with two organizations, Dietrich relayed. Scientists will have multiple points of contact when trying to develop their plans, and procedures among facilities may conflict. Distributed capabilities also affect rapid response to events (e.g., earthquakes)—Dietrich described increased bureaucracy and a need to pull resources from multiple facilities. Distributing data capabilities also means that data sets are dispersed and may need to be brought together for certain investigations. If the distributed data facilities are not integrated, it is more difficult to integrate data sets, and more difficult to generate integrated data analysis tools, he said. Furthermore, Dietrich reported, other scientific communities (e.g., the meteorological community) may not know how to find and incorporate data stored in distributed facilities into their own work.
There are disadvantages to distributed capabilities in terms of education and outreach. Centralized education and outreach can produce broader materials with greater efficiency, according to Dietrich’s report. In terms of workforce development, geographically distributed facilities might result in increased competition for competent staff.
Management Practices to Address Future Challenges and Allow Flexibility
When thinking about flexibility of the distributed capabilities, Dietrich’s group focused on lessons learned from the EOSDIS DAAC. The geographic distribution of the data centers, and of expertise with programmatic oversight, allows the program to take advantage of scientific expertise across the country. Geographic distribution offers the possibility of going to where a capability is emerging, he said, and there is flexibility to re-compete for a single center or capability. Additionally, geographic distribution also has opportunities for broad educational outreach. Dietrich noted that the Incorporated Research Institutions for Seismology (IRIS) does distribute efforts, for example, the Portable Array Seismic Studies of the Continental Lithosphere (located at the New Mexico Institute of Mining and Technology), and work done elsewhere.
Dietrich’s group considered how to respond to emerging efforts, raising the example of increased interest in distributed acoustic Doppler sensors. He said the breakout group wondered if the community response could be to find an independent scientific group that is working on the emerging topic and encourage them to develop it. Or is it more advantageous to bring that effort into a central facility? The DAAC model might be one that is applied here. Another example given by the group was that of the Department of Energy’s large-scale active source testing. Active source experiments with National Laboratory partners is another option for development of emerging technology, he reported.
Flexibility in response to emerging science and technology might be achieved with tiered but geographically or topically distributed capabilities. Such a model could also allow for cross-fertilization. The group noted that an essential component of future seismological and geodetic facilities is a facility governed by a community board of directors or council.
MANAGEMENT STRUCTURES FOR CENTRALIZED SEISMOLOGICAL AND GEODETIC CAPABILITIES
The remaining two breakout groups were assigned to consider management structures suitable for supporting centralized seismological and geodetic capabilities. Box 7.2 lists the prompting questions provided to those groups. Summaries of the key themes identified by group rapporteurs are provided in the sections that follow.
Summary of Breakout Group 3 Discussion Themes
Carolina Lithgow-Bertelloni, University of California, Los Angeles
Carolina Lithgow-Bertelloni opened the summary of her group’s discussion on centralizing some or all seismological and geodetic capabilities with the following statement: “Any contemplated merger should be one that [the] community/users are excited about.”
Scientific Advantages of Centralized Seismological and Geodetic Capabilities
Lithgow-Bertelloni described how multi-disciplinary projects, particularly those in remote places (e.g., polar environments or the deep ocean), would benefit from centralized facilities providing access to multi-instrument platforms and field engineers with diverse areas of expertise. Centralization—perhaps organized using a matrix management structure—could result in having enough people and expertise to serve multiple projects and
create a “one-stop shop” for equipment and logistical needs. Engineers in different domains might have an opportunity to interact more, investigators could prepare a single proposal package to request access to a single facility, and scheduling experiments might be easier, according to the group. And, whereas geodesy and seismology do not overlap perfectly, her group predicted that the intersection of geodesy and seismology instrumentation will increase with new technologies. In that case, she said, a centralized facility would be useful. Lithgow-Bertelloni also reported her group’s observation that a centralized facility would provide economies of scale when purchasing new equipment.
In terms of logistics and user support services, Lithgow-Bertelloni reported that a centralized facility with a single pool of engineers would facilitate best practices for all instrumentation, for example, in terms of power supply, communications, and instrument enclosures. Likewise, centralization might simplify some logistics associated with international projects (e.g., import–export issues, customs, power requirements, and communications requirements). Centralized capabilities would allow for close coordination for large projects within the facility. Lithgow-Bertelloni said her group considered if centralization might improve communication with other institutions, such as the U.S. Geological Survey (USGS). IRIS already has a good relationship with the USGS (e.g., the Global Seismic Network), and UNAVCO is collaborating with the USGS on early warning applications. A bigger centralized facility might facilitate more of these collaborative efforts.
Lithgow-Bertelloni summarized her group’s discussion about data management by saying that data distribution is not just about bytes of data. Her group agreed with the observations of others that maintaining a thread from data collection through archiving was important. Domain expertise remains important, and information needs and curation do not go away with a centralized facility. Her group observed that any efficiencies of centralization would not be gained as a result of reduced staff related to data management because individuals’ expertise remains vital. Furthermore, she said, any contemplated merger of facilities would need to preserve the connections between the user community and the data. Her group noted that there could be new synergies as a result of centralized data management, for example, in the form of geospatially referenced data. Other benefits of centralized data management staffing might include the development of creative ways to curate and therefore examine data given the greater number of domain experts, or the development of new or specialized output for educators.
A centralized data management facility, or potentially pooled separate systems, might be advantageous for management of the greatly increased amounts of data that are anticipated, according to Lithgow-Bertelloni’s report. It will be important to archive and manage those data, and also to work with those data where they reside. From a logistical point of view, a centralized facility might be able to negotiate better terms from vendors (e.g., commercial Cloud service providers).
In terms of education and outreach, Lithgow-Bertelloni reported a diversity of opinions. Centralization could be useful, but she described that there is already strong communication and collaboration between existing facilities and that centralized education and outreach may not necessarily represent efficiencies in terms of staffing needs. A centralized effort might, however, result in greater efficiencies for the public because it would not be necessary to go to multiple websites to download different products. Centralized education and outreach might also be more easily and better integrated with the outreach efforts of other institutions.
In terms of workforce development capabilities, Lithgow-Bertelloni’s group observed that a centralized facility might better connect to undergraduate institutions, but might have more difficulty reaching those schools or institutions with large populations of under-represented individuals, regardless of the effectiveness of the facility’s internship program. With respect to staff development, Lithgow-Bertelloni’s group thought centralization could provide opportunities for all staff scientists to expand expertise. The larger critical mass would offer greater flexibility for staff so they could move around in their careers.
Management and Governance Structures to Accommodate Unified Capabilities
The group noted that governance will change if capabilities are unified. Workshop discussion from the previous day informed the group that there is widespread demand for effective community governance regardless of the facility configuration. With multiple awards, there are different management responsibilities, and multiple boards of directors that are each accountable to the National Science Foundation (NSF). If unified, there is one joint board and governance management team overseeing everything. However, because it is desirable to maintain domain expertise, a joint board will be large and more logistically cumbersome to operate, she said. The National Ecological Observatory Network (NEON) was mentioned as an example of how technical working groups could be formed for each discipline, but Lithgow-Bertelloni reported her group’s concern about the layers of management that would likely be added to the structure.
Seismological and Geodetic Capabilities That Might Not Be Centralized
Lithgow-Bertelloni reported there could be some disadvantage to centralizing any seismological and geodetic facility capabilities because of the different domain expertise required and the different data needs. Her group’s session ended with broad thoughts about caution being exercised with any centralization, that there be an equal commitment among partners, that the enthusiasm of the scientific communities is important, and that there is a commitment to maintaining domain expertise.
Summary of Breakout Group 4 Discussion Themes
Michael Foote, The University of Chicago
Michael Foote summarized his group’s discussion about centralized seismological and geodetic capabilities. His group chose not to use the prompting questions (see Box 7.2) to guide their discussion. A common theme of the group’s discussion, Foote noted, was concern about excessive focus on areas of overlap in seismology and geodesy at the expensive of the fundamental elements of each.
The group noted that centralization does not necessarily mean geographic consolidation of all facility functions, but rather an administrative consolidation. He said that centralization of some instrumentation could be advantageous. Centralization also does not necessarily mean downsizing—the amount of work done by a joint operation will not
necessarily decrease. He reported that the complementary work of the combined staffs could increase the quality of service, and that centralizing existing facilities might benefit from pulling smaller groups into the enterprise. Disadvantages to centralization were related to efficiency and agility in some areas. Foote reported that there may be increased administrative lag when trying to reallocate funds, for example, but efficiencies could be gained in terms of maintaining and distributing data. He stressed the importance of continuity of domain expertise throughout the data life cycle.
Education and outreach endeavors might benefit from centralized management by avoiding the need to “reinvent the wheel,” Foote said. He stated that different science organizations may be independently trying to determine the best mechanisms for education and outreach. Foote added from his own experience that some of that reinvention may not be well informed by pedagogy.
With respect to workforce development, Foote reported that people joining an institution do not necessarily expect to stay there for their entire careers. With a merged or shared facility, the workforce may be exposed to a greater variety of experiences that could contribute to long-term career development. A question was raised during the group discussion about the ability to retain skilled staff of both organizations if seismological and geodetic functions are centralized—assuming a reduction in staffing is a requirement of the centralization. Staff also may choose to leave because of changes in the workplace environment.
Foote reported mixed opinions regarding whether the ability to respond to new opportunities would be enhanced or diminished by centralized capabilities. If centralization results in a larger administrative structure, it could be more difficult to get approvals needed to respond to new opportunities. On the other hand, money merged into a single fungible pool might make it easier to be flexible and respond more quickly. Foote reported agreement among the group, however, on governance: a community governance model was preferred with a board of directors selected by and accountable to its constituents.
PLENARY DISCUSSION
Following the reports from the breakout groups, the floor was opened to comments and group discussion. The following sections summarize some themes raised.
Maintaining Enthusiastic Staff Support
A workshop participant expanded on Lithgow-Bertelloni’s remarks regarding the need for enthusiastic support from the user community for any merging of facilities. The participant added that staff of the facilities also need to be enthusiastic about changes. Another workshop participant added an example from personal experience: for strategic reasons, a company for which he worked decided to move all researchers from one part of the country to another. Of the 120 senior scientists, only one elected to move. According to the participant, it was a brilliant business plan, but because of poor execution the company lost a lot of talent.
Scientific Community Convergence at Workshops
One workshop participant stated that workshops held by both IRIS and UNAVCO allow for convergence across research communities, community feedback to the facilities, and the opportunity to build cultural and scientific connections over time. These can lead to greater efficiencies and commonalities. Meghan Miller of UNAVCO countered, however, that she and Bob Detrick of IRIS have independently tested levels of community overlap at their workshops and observed that other than staff and sponsors, there was only narrow overlap between the seismology and geodesy communities at those workshops, on the order of a dozen or less out of a workshop of 100 people or more.
Knowing How to Access Data
Behnke took issue with a speaker comment that people know where to get seismology or geodesy data. Behnke stated that someone unfamiliar with IRIS may insert “seismology” in an Internet search engine and receive numerous results of varying relevance. She suggested caution about assuming familiarity, citing this as a problem at EOSDIS. She also described that potential users may have difficulty locating data because EOSDIS terminology has been duplicated across agencies. Data related terminology may no longer be unique to EOSDIS data, or it may not be used consistently across agencies.
International Participation
A workshop participant brought up the importance of international participation on facility boards. Lucy Flesch reported that the UNAVCO Board of Directors includes member representatives from the Netherlands and New Zealand. The number of international members is driven by needs, as identified by the nominating committee. Lithgow-Bertelloni said that, based on her own experience with large international organizations and small organizations, unless international participation on boards is explicitly designated, there is a tendency to vote U.S. members onto boards.
Community Governance Models
While one participant noted that a common theme during the workshop was that community governance was a vital aspect of seismological and geodetic facilities, Bradford Clement of the International Ocean Discovery Program observed that community governance does not guarantee a well-run organization. At present, IRIS and UNAVCO seem to have excellent leadership and boards, he said, but that might not always be true for the future. Another workshop participant spoke against community governance by stating that although community input is vital, leadership is necessary when “democracy goes wrong.” He said that leadership and oversight, and possibly redirection, needs to come from NSF. He did say he felt it appropriate to endorse current seismological and geodetic facility leadership for what they have done in the past few decades. Another workshop par-
ticipant suggested that because the seismology and geodesy communities rely heavily on global instrumentation networks, scientists in those fields are forced to think beyond their focused research. For this reason, they may think about community research more broadly, and thus about community governance. He suggested this might be why community governance works for these research communities.
This page intentionally left blank.