Project Leadership and Management Philosophy

Proper organization, management and coordination of all aspects of the project is necessary for success. The leadership of the project at all levels will operate with transparency and openness, while simultaneously accepting responsibility for achieving the project’s vision and goals. Formally, the E3SM is a Scientific Focus Area (SFA) project funded by the Office of Biological and Environmental Research (BER) and overseen by a BER Program Manager (PM). To achieve the goals of Phase II, we envision two classes of project functions – “core activities” and “next generation development (NGD) sub-projects”. The core activities primarily will target the project’s infrastructure, the execution and analysis of the v1 simulation campaign, the completion of the v2 model, and the execution and analysis of the v2 simulation campaign. The NGD sub-projects will target strategic, longer-term development for the v3 and v4 E3SM versions.

The organizational and functional structures of Phase II are briefly described below. The new structures adopt different approaches in managing the core and NGD activities, while promoting interactions between staff performing those functions.

Project Structure

The Executive Committee (XC) is comprised the Council Chair (CC, and project Principal Investigator), the Chief Scientist (CS), the Chief Computational Scientist (CCS) and the Project Engineer (PE), who were selected by the PM in consultation with the DOE Earth and Environmental Sciences Division Director. The XC has overall responsibility for achieving project goals and delivering products. Accordingly, in consultation with the PM, it will determine the project team structure, organization, staffing and allocation of resources. The Council Chair as SFA Manager, is ultimately responsible to ensure delivery of Project products and report progress to DOE. The CS will lead Council (described below) and leadership team planning and execution concerning the scientific goals and deliverables for Earth system science. The CCS will lead Council and leadership team planning and execution concerning computational goals and deliverables. The PE functions as the project’s chief operating officer, overseeing the coordination of project elements and the tracking of project performance and deliverables, as well as providing scientific and technical expertise to the project.

Core Activities are organized and distributed across five groups, each with a single leader. Some groups have deputies to assist the leader.

Water Cycle - Christophe Golaz, leader; Luke Van Roekel, deputy

Cryosphere – Steve Price, leader, Wuyin Lin, deputy, Mark Petersen, deputy

Biogeochemical Cycles – Katherine Calvin, leader; Susannah Burrows, deputy

infrastructure and data management – Robert Jacob, leader; Chengzhu (Jill) Zhang, deputy

Performance – Philip Jones, leader, Sarat Sreepathi, deputy

The science question groups will complete the v1 experimental campaign, which may include additional necessary changes to the model system. They will also complete the integration and testing of new developments for the v2 modeling system, including the completion of the code review process for each configuration, prior to the start of the v2 experimental campaign. The infrastructure has the responsibilities of maintaining and extending the project’s software system, associated scripts, analysis software and management, curation and distribution of production simulation output. The performance group will test and apply optimal model layouts for the simulation campaign configurations and solve problems associated with changes in the software stacks (compilers, libraries, I/O systems, etc.) at the Office of Science computing centers.

Most of the project’s work will be performed by small teams of staff working on well-defined “Epics,” a term borrowed from our project management system. Epics are defined in the project road maps and have a well-defined deliverable and due date, allowing the team to operate day-to-day without direct involvement by the leadership. Epic teams typically are comprised of between three and eight people, with a single leader, who is accountable to the group leader. A single project staff member may work on two or more Epics at any given time.

Coordination of the teams and integration of the Epics is the function of the leadership. Overall planning for the core activities will be accomplished through 3, 6 and 12-month road maps.

Next Generation Development Sub-Projects pursue higher risk developments for versions v3 and v4, but with a path for integration into E3SM within five years. Subprojects will have less intensive oversight than core activities, but are still accountable for plans, progress and deliverables. The sub-projects will operate as semi-autonomous, highly focused teams pursuing aspects of v3 and v4 development. NGD sup-projects generally will fill needs and requirements for the longer term strategic plan priorities. With proper coordination and review, we fully expect that most, if not all, NGD projects will be able to adjust to changing circumstances without major disruption. Nevertheless, it is possible that an NGD project may be redirected, reduced or eliminated before the end of the three years depending upon progress.

Sub-project foci will be in the following areas, each with a leader to coordinate the work:

Software and Algorithms – Andrew Salinger, acting leader

Nonhydrostatic Atmosphere – Peter Caldwell, leader

Atmospheric Physics – Shaocheng Xie, leader

Land and Energy - Benjamin Bond-Lamberty, leader

Two smaller NGD “mini-projects” in coastal modeling and implementation of the BISICLES dynamic land ice model will be overseen directly by the XC. As with core activities, NGD sub-project research will be performed by small teams of staff working on well-defined Epics, each with a leader and deliverable. Because NGD sub-projects are higher-risk, they will have greater independence and utilize six and twelve-month roadmaps to allow more flexibility.

Governance

The Council will be composed of the XC, the five core activity group leaders and the leaders of the four NGD sub-projects. All major project decisions will be discussed and vetted through the Council. In practice and based on experience, most major decisions are achieved through consensus of the Council. In cases for which that is not possible, the Executive Committee will make all final decisions. The XC and DOE National Laboratories’ managements will identify project staff members to serve as points-of-contact at each of the participating laboratories (Lab POCs). POCs will report to the XC on laboratory participation and costs, represent the project to their lab managements, and work with lab, project, and DOE management as needed to resolve any problems. Figure  shows the organizational structure of the E3SM Phase II project.

Project execution

The management paradigm delegates authority down to the Epic leaders within the five groups and four NGD subprojects. This is shown in Figure 23. The leadership's role is to provide assistance, when needed, as well as overseeing the coordination and integration between and among the project elements. The roadmaps, which will be derived from the strategic plan and proposal, will be sufficiently defined and detailed so that they provide scope and guidance to each Epic team and its leader.

Communication and collaboration - Good communication is essential within the project. We will continue to use the Confluence, Jira, github and other tools for communication, collaboration and project management. Regular meetings and telephone conferences are also essential. For example, a possible meeting schedule would be:

• The full project will have meetings semi-annually;
• The Council will meet by phone monthly, and in person semi-annually;
• The EC will meet by phone bi-weekly with each of the five groups’ leaders;
• The EC will meet by phone monthly with each of NGD sub-project leaders;
• Within the core activity groups and NGD subprojects, the leaders will determine the type and frequency of meetings with their Epic leaders;
• Within each Epic, we anticipate weekly or more frequent communication. This is where the work gets done.

Project Team

The Project Team includes all DOE national laboratory scientists and close external collaborators executing technical work in the project. The XC recruited staff members from across the laboratories who understand and embrace the overall priorities and objectives of E3SM. DOE laboratory staff participation in E3SM requires a minimum 50% commitment. Based on the experience of Phase 1, and learning from other modeling centers, Phase II requires more dedicated effort for much of the work in the project than was available in Phase 1. Consequently, DOE lab staff were recruited for both “Dedicated” and “Regular” positions. The roughly 40 dedicated staff committed to E3SM at the 75%-100% effort level for a period of two years and agreed to make E3SM their top priority, with at least 50% effort devoted to core activities. An additional ~45 DOE lab participants are considered Regular staff and will work on either core activities, NGD subprojects or both. Within the requirements for time commitments, the project considers participation in “Ecosystem” projects. Project participants from outside the DOE lab system, especially the academic sector, are included the project may solicit specific expertise from non-laboratory institutions, especially in the academic sector will be supported through subcontracts.

External Advisory Committee

The Council will establish an External Advisory Committee of five to seven qualified technical experts in the various aspects of the project to provide frank and candid advice to the Council on the plans, progress, and direction of the project. The first meeting of the Advisory Committee will be held approximately three months after the start of the project to review the initial plans, and will meet approximately annually thereafter, typically at one of the quarterly Council face-to-face meetings. Advisory Committee members will be invited to all semiannual Project Meetings.