Steps to Facilitate Principal-Investigator-Led Earth Science Missions (2004)

This chapter outlines the scientific and programmatic objectives of PI-led missions in ESE and summarizes NASA’s experience with PI-led missions in both Earth and space sciences over the last several decades.

Through the 1980s and 1990s, NASA’s Earth science program built the foundation of its observing system on facility-class missions, including core Earth Observing System projects that have evolved into the current Terra, Aqua, and Aura missions. Facility-class missions have been generally successful in providing an unprecedented quantity of unique, high-quality data for Earth science investigations. However, since each facility-class mission is required to serve many different user communities, individual science objectives and instrument capabilities have sometimes been compromised because of limitations in overall mission resources and spacecraft accommodation.

In addition, facility-class mission costs have been high and development times long, in part because steps must be taken to mitigate significant scientific and programmatic risks associated with the launch or on-orbit failure of large satellites. But the pace of advances in scientific understanding and measurement technology have often resulted in considerable changes in scientific priorities or instrument capabilities during the decade or more that typically has passed between selection of the instrument complement for a facility-class mission and its eventual launch.

ESE established the Earth Explorers Program to address these issues by providing frequent, flexible opportunities for rapid-development flight missions focused on specific Earth science investigations. Earth Explorer missions thus fill a well-defined and focused role in ESE, complementing facility-class missions to achieve ESE’s overall scientific objectives.

The Earth Explorers Program is explicitly designed to support Earth remote sensing missions that contribute directly to:

• Acquiring key additional measurements in response to new scientific understanding, including exploiting scientific discoveries from facility-class missions;

• Proving the concept and scientific utility of new data sets and measurement approaches; and/or

• Ensuring the continuity of critical measurement time series (i.e., “gap filler” missions for critical data sets).

Although the following are not always stated as explicit goals, the program appears also to have been designed to:

• Provide frequent, predictable opportunities for training new investigators and ensuring the continued broad involvement of the scientific community in the overall development of ESE satellite projects;

• Encourage direct involvement of university faculty and students in all aspects of ESE flight mission planning and implementation, and expand the base of academic institutions that have the capability (through experienced faculty) to manage satellite-related technical projects; and

• Foster development of innovative teaming arrangements that optimize the contributions and minimize the costs of industry, university, and government partners.

The Earth Explorers Program also has a wide-ranging set of explicit constraints that greatly influence all aspects of its sponsored missions.¹ These constraints fall generally into the areas of (1) mission scope and capabilities, and (2) mission selection and management.

• Mission Scope and Capability Constraints

  • Acquisition of measurements from space must be central to each mission, with each mission acquiring all unique measurements and producing all fundamental new products needed to solve the defined scientific problem.

  • Prelaunch development time must be less than 36 months.

  • All data required to make substantial progress on the defined scientific problem must be acquired within 3 to 4 years after launch.

  • Mission success must not require a sustained, long-term commitment on the part of the responsible institution or PI beyond the expected on-orbit mission lifetime.

  • NASA costs are capped, although collaborations with domestic and international partners are allowed in order to increase scientific return without requiring additional direct NASA funding.

  • NASA funding for launch vehicles is separately limited and can only be applied to a specific list of launch vehicle options.

• Mission Selection and Management Constraints

  • All missions are selected competitively in response to open AO solicitations.

  • Missions are led by a principal investigator.

Until FY2002, the Earth Explorers Program included two primary PI-led mission components with related but distinct goals and scope: the Earth System Science Pathfinder (ESSP) and the University Earth System Science (UnESS) components.² Although UnESS was not funded by Congress in FY2002 and is not part of the approved FY2003 budget, the committee believes that its objectives were unique, important, and contributed directly to the development of end-to-end remote sensing expertise in the academic community. Thus, while this report focuses

on ESSP as the primary remaining PI-led mission component of the Earth Explorers Program, descriptions and analyses of UnESS are included where appropriate.

ESSP AOs solicit proposals for focused, unique PI-led satellite missions that support any of the ESE Earth science objectives.³ ESSP missions are selected to provide new space-based observations that are complementary to, and not duplicative of, existing and planned data sets. Although ESSP missions may demonstrate the utility of new measurement types and approaches, the conduct and substantial solution of a specific, well-defined geophysical investigation using 3 to 4 years of on-orbit data is a paramount programmatic requirement.

The most recent ESSP AO explicitly requires that the proposed science be considered high priority and clearly beyond the scope of that possible from existing or approved missions in order for the proposed mission to be accepted.⁴ Successful ESSP proposals are supported over the total mission life cycle from concept formulation and refinement through integration, test, and launch, followed by production and dissemination of validated geophysical data sets. Proposed missions must include payload, spacecraft, and launch components in addition to data production and validation activities—they cannot be restricted to provision of a limited suite of instruments to fly as a partial, or secondary, payload in association with another mission.

The NASA cost for each ESSP mission is capped at a level defined in the AO. To maximize the scientific value and minimize NASA’s cost, PIs are encouraged to collaborate in a cost-sharing mode with non-NASA organizations, including foreign organizations and agencies.⁵

Table 2.1 summarizes the ESSP selections to date. The first ESSP AO was issued in 1996 and resulted in the selection of the Vegetation Canopy Lidar (VCL) and GRACE missions. CloudSat and what is now called CALIPSO were chosen from the second ESSP AO in 1998; Aquarius and the Orbiting Carbon Observatory (OCO) were selected during the third AO in 2001.

The life cycle of a PI-led ESSP mission begins with a science-oriented Step 1 proposal effort. Selection at the end of Step 1 leads to a recommendation for submission and evaluation of a competitive Step 2 proposal, which must include the technical aspects of mission development as well as the science. Selection after a Step 2 proposal results in a funded mission project, which leads to a mission confirmation review. After passing mission confirmation, the project moves on to implementation and launch. Following launch and after up to 3 to 4 years of on-orbit data acquisition, the baseline PI mission is regarded as complete. Figure 2.1 provides an overview of the selection process and other mission stages.

UnESS sought to develop PI-led spaceflight missions in which all aspects of the scientific formulation, mission design, implementation, and data analysis would be performed substantially by students. Like ESSP, UnESS missions were to involve the use of space-based data to address geophysical problems related to the ESE science strategy and complementary to other ESE investigations and missions. However, in contrast with other ESE flight programs, UnESS objectives and mission selection criteria gave equal weight to science and to the training and education of researchers, engineers, managers, educators, and entrepreneurs through extensive hands-on involvement throughout the mission.

UnESS missions were designed to be smaller than ESSP missions and to be developed at lower cost and on a faster schedule. Mission scope was not expected to be as encompassing as for an ESSP mission. NASA costs were capped at $15 million; typically, missions were intended to have a 9-month concept study followed by 24 months for definition, approval, design, and development. Unlike ESSP missions, UnESS investigations also did not have

to be complete flight missions in which the proposed instrument (or instrument suite) was the primary payload on the spacecraft and launch vehicle. Instead, UnESS proposals could address partial missions, in which the instrument was a secondary payload on the spacecraft.

As originally designed, UnESS was to release biennial solicitations, from each of which two missions would eventually be selected for flight, leading to a flight rate of about one UnESS mission per year. The first (and only) UnESS AO was released in late 1999; of the 24 proposals submitted, 4 were selected for further concept definition (a fifth effort, resulting from combining two separate proposals, was also supported in the concept definition phase). But UnESS was not funded by Congress in the FY2002 budget, precluding advancement of any of the missions to development or flight.

Distinctions in scientific focus and mission management differentiate Earth Explorer missions from the larger and longer-duration facility-class missions.

Facility-class missions are justified based on the breadth of the science enabled by the data they collect and the contributions of the measurements they make to the generation of multidecadal time series of key quantities. To ensure widespread exploitation of the data, ESE selects interdisciplinary science teams for the facility-class

FIGURE 2.1 Steps in the development and execution of an ESE PI-led mission.

missions through competitive research announcements. During the prelaunch phases of the missions, competing and evolving requirements of different scientific users can influence the design of the instruments, data products, and calibration/validation activities in the quest for broad scientific utility. Facility-class missions thus fulfill an ESE commitment to acquire data for many segments of the Earth science community; however, this commitment can result in the degradation of the measurements’ utility for specific investigations in exchange for their contributions to a wide range of multidisciplinary studies.

In contrast, PI-led Earth Explorer missions are distinguished by their focus on well-defined, important geophysical problems that are amenable to substantial progress using 3 to 4 years of on-orbit measurements; indeed,

the ranking of the proposed investigation’s contemporary scientific importance is a primary selection criterion for Earth Explorer (and especially ESSP) missions. The focus on addressing a small set of specific scientific problems ensures that trade-offs made during the development stages (for example, to accommodate mission constraints) directly support the particular scientific goals that were the basis for the mission’s initial acceptance; at least in principle, design and implementation trade-offs do not require balancing the needs and desires of diverse research communities addressing a wide range of geophysical investigations. All of the Earth Explorer AOs to date have limited the funding of the science team to the design, production, and validation of new geophysical data products, rather than the solution of an identified problem or the conduct of the specific investigation(s) that formed the basis for the mission’s selection and justification. Funding for the latter is awarded competitively as the mission approaches operational status through a separate Science Data Analysis Program (SDAP) NASA Research Announcement (NRA).

The NASA-mandated PI-led mission management approach is fundamental to the Earth Explorers Program and is a significant differentiator between Earth Explorer missions and facility-class missions. There is, however, no unique definition of a “PI-led management approach.” As articulated in presentations to the committee by NASA program and NASA center officials, as well as in the most recent ESSP AO,⁶ the intention of the PI-led management approach is to vest end-to-end mission responsibility (from original concept, through implementation, to generation and distribution of validated data sets and products derived from remotely sensed measurements) in a single, identified PI, working with a team of his or her choice.⁷ The PI is accountable to NASA for overall mission success, including maintenance of the scientific/applications integrity and success of the mission.⁸ To achieve these ends, the PI is formally empowered to manage cost and schedule milestones at every stage of the mission;⁹ in particular, the PI is responsible for making key science trade-offs, including those required by resource limitations (e.g., funding, mass, power, accommodation).¹⁰

Over the last decade, PI-led missions have been used increasingly as a fundamental building block of NASA’s ESE and Space Science Enterprise (SSE). PI-led missions represent a trade-off between two important but potentially conflicting objectives: innovation derived from direct ownership of mission success by PIs in the science community, and mission success based on the long heritage of successful projects performed in the NASA centers. Both ESE and SSE have begun to recognize that a balanced scientific program includes elements that accept some increased risk to mission success in return for enhanced scientific and programmatic innovation.

The PI-led mission paradigm has indeed led to innovative science that has been successfully executed. Within SSE, these projects include the NEAR mission, which achieved the first rendezvous with an asteroid, and the MAP mission, which is redefining our understanding of the cosmic background radiation and the early universe. Within ESE, projects include the GRACE mission, which is providing profound insights into Earth’s structure, and the recently launched SORCE mission, which is investigating the role of solar variability in climate change.

The committee believes that the PI-led approach to implementing missions is fundamentally valuable, subject to acceptance of the trade-off between innovation and risk. As one element of an overall scientific program, PI-led missions provide an important complement to facility-class missions and other means for obtaining scientific data.

A mixed portfolio of mission sizes was endorsed in a previous National Research Council report,¹¹ which also provides insight into many issues associated with PI-led missions, as summarized in Box 2.1.

Finding: The PI-led mission paradigm represents a valuable approach to soliciting and executing missions involving focused science objectives, with demonstrated success in both the Earth and space sciences. PI-led missions provide an important element of the overall ESE observing strategy, complementing other elements such as facility-class missions and data buys.

Recommendation: NASA’s Earth Science Enterprise should continue to employ PI-led missions as one element of the ESE observation system. It should ensure regular review and improvement of the programs that employ or are associated with PI-led missions to increase their effectiveness and value to ESE and the science community.