In This Issue
Fall Issue of The Bridge on Space Exploration
September 1, 2021 Volume 51 Issue 3
Close collaboration between engineering and science has enabled marvels of space exploration over decades. Eight exemplary missions are described in this issue, conveying the excitement, challenges, and breakthroughs involved in efforts to better understand the wonders and mysteries of this solar system.

MAIA: Opportunities and Challenges Associated with a Small, Cost-Capped Satellite Mission

Monday, September 13, 2021

Author: Stacey W. Boland and David J. Diner

MAIA breaks new ground with its emphasis on human health applications and efforts to use commercial access to low Earth orbit.

In 2007 an Earth science decadal survey outlined a vision for “a decadal program of Earth science research and applications in support of society” and called for “a renewal of the national commitment to a program of Earth observations from space in which attention to securing practical benefits for humankind plays an equal role with the quest to acquire new knowledge about Earth” (NRC 2007, p. 1). The report considered that “the societal benefits accruing through improved human health should be ­fundamental to defining the research and applications goals of the Earth science ­agenda, including the need for an intellectual framework that directs bridging research between the Earth system framework and the public-health response and decision-maker community” (NRC 2007, p. 155).

An important outcome of the 2007 report was the establishment of NASA’s Earth Venture program, a class of PI-led, low-to-moderate-cost missions managed by the Earth System Science Pathfinder Program Office (ESSPPO). With frequent calls for proposals, the Venture program provides flexibility to enable new research avenues, accommodate scientific ­advances, demonstrate key application-oriented measurements, and explore novel implementation approaches (NRC 2007; Peri and Volz 2013).

The Multi-Angle Imager for Aerosols (MAIA) investigation was selected in 2016 in response to NASA’s third Earth Venture Instrument (EVI) solicitation. MAIA’s primary objective is to explore the human health effects of exposure to different aerosol types[1] (Diner et al. 2018). As noted by the National Research Council (NRC 2004, p. 89), “a better understanding of [the] characteristics of particles that modulate toxicity could result in targeted control strategies that would specifically address [the] sources having the most significant effects on public health.” With a multidisciplinary science team that includes epidemiologists, MAIA is the first competitively selected NASA investigation aimed specifically at societal benefit.

MAIA is the first competitively selected
NASA investigation
aimed specifically at
societal benefit.

MAIA implementation is now in the final design and fabrication phase.[2] Development challenges that have arisen since inception include several that are inherent to the Venture paradigm, some that result from the unique nature of the MAIA investigation, and some caused by unexpected disruptions beyond the project’s control. This paper describes examples of such challenges, how scientists and engineers worked together to address them, and lessons learned.[3]

Motivation for MAIA

According to a recent study, ambient particulate air pollution is the top environmental risk factor for worldwide deaths (GBD 2020). Epidemiological studies have linked particulate matter (PM) to adverse health outcomes such as chronic obstructive pulmonary disease, asthma, lower respiratory infections, lung cancer, ischemic heart disease, heart attack, stroke, cognitive impairment, kidney disease, diabetes, preterm birth, and low birth weight.

The US Environmental Protection Agency and its counterparts in other countries regulate PM2.5 and PM10, particles with aerodynamic diameters less than 2.5 and 10 µm.[4] But “One of the biggest gaps in our knowledge relates to what specific air pollutants, combination of pollutants, sources of pollutants, and characteristics of pollutants are most responsible for the observed health effects” (Pope and Dockery 2006, p. 730). Because of the myriad PM sources and typical atmospheric residence times of less than a week, tropospheric aerosols are complex mixtures of particles and exhibit a high degree of spatial and temporal variability. Researchers have cited the need to “find the particles that are most dangerous to health.… The mix—and its toxicity—varies…in ways that are not tracked, understood or managed” (Li et al. 2019, p. 437).

Surface-based monitors provide the most accurate means for measuring PM mass concentrations and chemical compositions. But, as noted in the United Nations Urban Air Quality Management Toolbook (UNEP 2005, p. 54), “The principal disadvantage of direct ambient air quality monitoring…is its cost in money, equipment and skilled manpower.… The financial and technical demands are usually the biggest constraint to ambient air quality monitoring in developing countries.” Reliance on a sparse distribution of central fixed-site (ground) monitors is thus a limiting factor in knowledge of which PM types are most harmful.

Furthermore, neighborhood-scale (0.5–4 km) measurements are needed to better represent the conditions in which people live and work (EPA 2013a,b). While low-cost PM sensors can increase the density of total PM mass concentration estimates, monitors capable of assessing PM speciation are scarce because of the complexity and cost of the needed sample collection and chemical analysis systems.

Observations from Earth-orbiting satellites can provide extensive spatial coverage to help address these issues. A 2005 workshop jointly organized by the ­National Institute of Environmental Health Sciences and EPA highlighted the value of remote sensing to “augment ground-based air quality sampling and help fill pervasive data gaps that impede efforts to study air pollution and protect public health” (Tinkle et al. 2007, p. 2).

Technical and Scientific Elements

A follow-up decadal survey (NRC 2018) listed among its most important objectives the estimation of global air pollution impacts on human health, including impacts due to speciated PM, and encouraged the use of combined remote and ground-based sensing and model data to meet this objective. Each data source brings complementary attributes in terms of accuracy, capability, and spatial and temporal coverage and resolution.

MAIA employs this combined approach to generate daily maps of total PM10, total PM2.5, and PM2.5 species (sulfates, nitrates, organic carbon, elemental carbon, and dust) at 1 km resolution over globally distributed target areas. The NRC (2018) report cites MAIA’s unique contributions to the program of record (POR) for aerosol observations and societal/health applications, showing how Venture class investigations can complement other missions in the POR and serve as pathfinders to components of comprehensive future programs, such as NASA’s new Earth System Observatory.[5]

Key elements of the MAIA investigation—made possible by recent advances in remote sensing and health science, computational and geostatistical modeling, sensor engineering and networking, and space qualification of novel measurement technologies—are briefly described in the following sections.

Satellite Instrument

Satellite-based techniques for estimating aerosol properties over land have advanced significantly over the past 2 decades (Dubovik et al. 2019; Kokhanovsky and de Leeuw 2009) and demonstrate the value of combining multiangular, multispectral, and polarimetric observations for mapping column aerosol optical depth, physical characteristics (e.g., particle size distribution), and optical properties (e.g., light absorption efficiency).

Boland and Diner fig 1.gif
FIGURE 1 The observational component of the MAIA investigation combines space- and surface-based data. Left: MAIA camera during assembly and testing at the Jet Propulsion Laboratory. Light enters from the right (the red protective cover will be removed before flight) and is brought to a focus at left rear. Right: Rooftop air quality station at JPL with several ground-based monitors. PM = particulate matter. Image credits: NASA/JPL.

Leveraging this heritage, the MAIA instrument contains a spectropolarimetric camera that is pointable in two axes (figure 1). The along-track axis enables observations of selected targets at multiple view angles as the satellite passes overhead; the cross-track capability enables revisits of these targets 3–4 times per week. Using technologies developed with support from NASA’s Earth Science Technology Office and Research Program (Diner et al. 2013) the targets can be imaged with moderately high spatial resolution in 14 spectral bands (three of which are polarimetric) from the ultraviolet to shortwave infrared.

As noted in a report on the impact of air pollution on children (UNICEF 2019, p. 9), “Integration of satellite-based estimates with reliable ground-based measurements is…necessary to…[enhance] understanding of local and regional air quality.” MAIA is designed to provide the required observational capabilities.

Surface Monitor Network

Satellite-based aerosol microphysical and optical attributes are only indirectly related to chemical composition. Transformations between satellite aerosol information and near-surface PM properties will be derived by generating geostatistical regression models (GRMs) that use colocated surface and MAIA satellite measurements. Once the coefficients and predictor variables are established (an ongoing process that improves in quality as the mission progresses), the models will be applied to the rest of the satellite imagery to produce spatial maps.

MAIA epidemiologists will use the project’s PM exposure maps to examine health outcomes in globally distributed target areas.

The surface network to be used by MAIA includes existing PM2.5 and PM10 monitors and filter-based chemical speciation samplers operated by government agencies and research organizations, supplemented by filter-based samplers, black carbon aethalometers, and low-cost sensors deployed by the MAIA project (figure 1).

Chemical Transport Model (CTM) and Other Ancillary Data

A CTM is a computerized simulation of the aerosol lifecycle, gas phase chemistry, and atmospheric flow. MAIA uses the mesoscale weather research and forecasting model coupled with chemistry (WRF-Chem) to generate estimates of PM10, PM2.5, and speciated PM2.5. The CTM is primarily used for spatial and temporal gap filling in places where satellite retrievals are unavailable (e.g., because of cloud cover) or on days when a particular target area is not observable from orbit.[6]

The CTM is also a source of meteorological information that, along with ancillary surface data such as human population and roadway density, will be used as PM predictors in the GRMs.

Health Studies

MAIA epidemiologists will use the total and speciated PM exposure maps generated by the project to examine health outcomes associated with different time scales of exposure in globally distributed target areas.[7] Acute (short-term) exposure is generally associated with premature mortality and increased hospital visits. Subchronic exposure studies are aimed at birth outcomes and pregnancy complications, such as low birth weight and preeclampsia. Chronic exposure studies track health effects resulting from PM inhalation over many years.

The study designs will be based on the predominant PM species present, types of health records available, and previous studies of the effects of particulate air pollution.

Boland table 1.gif

Boland and Diner fig 2.gif
FIGURE 2 MAIA target map showing geographic distribution of primary target areas (PTAs; blue), secondary target areas (STAs; green), STAs that might be converted to PTAs ­(purple), and ­calibration/validation target areas (orange). Designations are current as of July 2021 and subject to update as the project matures. Map data © 2001 Google.

Target Areas

To fit within Venture program resource constraints, the MAIA investigation was scoped to address investigation goals achievable within a discrete set of target areas, as described in table 1 and illustrated in figure 2. The ­criteria used to select the primary target areas (PTAs) are enumerated in table 2.

Boland and DIner table 2.gif

Opportunities, Challenges, and Lessons Learned

NASA’s Venture program provides frequent opportunities for PI-led cost-capped investigations to “focus on establishing new research avenues or on demonstrating key application-oriented measurements” (NRC 2007, p. 12). The EVI line supports projects for which the PI is responsible for the investigation, while access to space is coordinated through the ESSPPO.[8]

The 2017 decadal survey recognized both opportunities and challenges associated with small missions, ­citing the benefit of opening the door to “non­traditional vendors and providers [in] both the government and the private sector” while also identifying “risks that must be acknowledged and accepted, as would be the case with any evolving technology in its early stages of realization” (NRC 2018, p. 185).

MAIA’s development experience as part of EVI has elucidated several practical lessons, as described below.

Optimization vs. Flexibility

The hosting services paradigm trades optimization for flexibility. The MAIA instrument is planned for launch aboard a commercial spacecraft in late 2022. The ESSPPO awarded the MAIA hosting services contract in August 2018, 4 months after the instrument’s preliminary design review. The launch provider was only recently selected. The timing of these events relative to instrument development means that instrument engineering proceeded without knowing many technical details about the launch vehicle, host spacecraft, or orbit, requiring resilience to a wide range of scenarios. This implies that the EVI line’s quest for programmatic flexibility comes at the expense of engineering (and cost) optimization. It also requires a level of risk tolerance from engineers and project and program managers not typically associated with NASA flight projects.

Cost, Schedule, and Performance Trade-offs

Trade-offs between cost, schedule, and performance tilt toward cost. On any cost-capped project, implementation issues encountered during the project design phase (unless they are outside of the project’s purview, such as host platform interfaces[9]) must be handled within the project’s cost and schedule commitment, leaving performance trades as the remaining means of problem resolution.

As an example, during the design phase it became clear that the original plan to have two side-by-side cameras would lead to significant engineering challenges. Once the science team had selected most of the PTAs, however, analysis showed that critical requirements could be met with a single camera, though at the expense of cross-track coverage and engineering redundancy.[10] While it is not uncommon for cost-capped projects to implement descopes, this decision was made more difficult by not yet knowing the host spacecraft’s orbit altitude (which affects ground coverage and target revisit frequency). The science and engineering teams worked closely together to protect core mission science, meet high-level requirements, preserve engineering margins, and reduce project risk.

Resource Limitations

Limited resources encourage and even require creativity. Detailed design during the development phase can uncover limitations in the plans envisioned during the proposal phase. For a cost-capped project, this means either finding ways to meet the identified needs with existing resources or addressing them by other means. MAIA has taken advantage of both approaches.

For example, to ensure statistical robustness of the GRMs, the science team recommended increasing the surface monitor sampling frequency. The ­project addressed this by (i) reconsidering the equipment deployment plans to make use of samplers with different designs, capabilities, and costs; (ii) establishing a partnership with the US State Department to handle the logistics and costs of overseas shipments[11]; (iii) leveraging support from the US Agency for International Development to fund surface monitor deployments and filter chemical analyses in Africa[12]; and (iv) making arrangements in several PTAs to access speciated PM data from organizations that operate existing samplers.

Complexity

Coordination of multiple project elements adds complexity. The MAIA engineering effort includes the ­typical disciplines associated with implementation of a flight instrument, but adds the technical and ­logistical complexities of coordinating the deployment of PM monitoring equipment, access to health records, and study cohorts in numerous national and inter­national target areas. While these elements are somewhat decoupled from the instrument delivery schedule, they are linked to launch date, which in the early phases of the project was uncertain and still holds the prospect of fluidity.[13] Similarly, lining up suitable health data sources and collaborators requires adjusting plans in the event of a change in launch date or other unexpected events.[14]

Project Team Stability

Maintaining a stable core project team is essential. Minimizing disruptions to project staffing is necessary to ensure continuity of knowledge and avoid schedule delays. This can be challenging for small missions that do not have sufficient funding to maintain key staff at full-time support. MAIA benefits from individuals with flight mission experience who are accustomed to working across disciplinary boundaries, in combination with more junior personnel. Careful management has been required to deal with demands on their time from other projects.

Proactive Management

Contention for facilities requires proactive management. Competition for facilities can be a challenge for small missions, particularly at an institution where access to clean rooms, environmental test facilities, and prioritization in the delivery queues of commonly used subcontractors compete with the needs of larger, highly visible, and more expensive missions. As an example, the contractor responsible for the MAIA telescope received a high-priority order for rework of an optical system for another project that delayed the telescope delivery and incurred unexpected cost to MAIA.

Some requests from the user community had to be declined because of the project’s limited resources.

Expectations vs. Resources

User community expectations are not always consistent with resources. A number of candidate observing targets have been identified and proposed beyond those shown in figure 2, but enlarging MAIA’s target set might reduce the frequency of PTA observations because of scheduling conflicts and the need to manage onboard data storage. Consequently, some target requests from the user community had to be declined because of the resource-limited nature of the investigation. Such limitations, while undesirable, are necessary to maintain project integrity.

Conclusion

The 2007 and 2018 decadal surveys envisioned an Earth observation program that advances both basic scientific research and societal applications. The Earth Venture class has provided opportunities for PI-led investigations to contribute in novel ways.

MAIA is a small, PI-led, cost-capped mission that is also complex and ambitious. Enabled by years of scientific heritage and technology maturation, it breaks new ground for the NASA Earth science program with its emphasis on human health applications, direct involvement of epidemiologists on the science team, and engagement with the ESSPPO’s efforts to use commercial access to low Earth orbit.

Accomplishment of investigation objectives within the EVI cost cap, management of programmatic complexity associated with interfacing with a commercial host, and integration of a multicomponent system of space-based instrumentation, ground measurements, atmospheric models, and ancillary data sources all present a variety of challenges. Addressing them requires sustained focus on the investigation’s primary objectives, diligent and creative project management, and active communication and negotiation among the science, engineering, and user communities.

Acknowledgments

We thank the MAIA project manager, Kevin Burke, along with the entire MAIA team, our international and interagency collaborators, and the ESSPPO for their dedication and support. The MAIA project is managed by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. © 2021 ­California Institute of Technology. Government sponsorship acknowledged.

References

Diner DJ, Xu F, Garay MJ, Martonchik JV, Rheingans BE, Geier S, Davis A, Hancock BR, Jovanovic VM, Bull MA, and 3 others. 2013. The Airborne Multiangle ­SpectroPolarimetric Imager (AirMSPI): A new tool for aerosol and cloud remote sensing. Atmospheric Measurement Techniques 6(8):2007–25.

Diner DJ, Boland SW, Brauer M, Bruegge C, Burke KA, ­Chipman R, Di Girolamo L, Garay MJ, Hasheminassab S, Hyer E, and 12 others. 2018. Advances in multiangle satellite remote sensing of speciated airborne particulate matter and association with adverse health effects: From MISR to MAIA. Journal of Applied Remote Sensing 12(4):042603.

Dubovik O, Li Z, Mishchenko MI, Tanré D, Karol Y, Bojkov B, Cairns B, Diner DJ, Espinosa WR, Goloub P, and 30 others. 2019. Polarimetric remote sensing of atmospheric aerosols: Instruments, methodologies, results, and perspectives. Journal of Quantitative Spectroscopy and Radiative Transfer 224:474–511.

EPA [US Environmental Protection Agency]. 2013a. ­National Ambient Air Quality Standards for Particulate Matter: Final Rule. Federal Register 78(10), Jan 15, Rules and Regulations.

EPA. 2013b. Quality Assurance Handbook for Air Pollution Measurement Systems, Vol II: Ambient Air Quality Monitoring Program (EPA-454/B-13-003). Washington.

GBD [Global Burden of Disease] 2019 Diseases and Injuries Collaborators. 2020. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. Lancet 396:1204–22.

Kokhanovsky A, de Leeuw G, eds. 2009. Satellite Aerosol Remote Sensing over Land. Berlin: Springer-Verlag.

Li X, Jin L, Kan H. 2019. Air pollution: A global problem needs local fixes. Nature 570:437–39.

NRC [National Research Council]. 2004. Research Priorities for Airborne Particulate Matter: IV. Continuing Research ­Progress. Washington: National Academies Press.

NRC. 2007. Earth Science and Applications from Space: ­National Imperatives for the Next Decade and Beyond. Washington: National Academies Press.

NRC. 2018. Thriving on Our Changing Planet: A Decadal Strategy for Earth Observation from Space. Washington: National Academies Press.

Peri F, Volz S. 2013. Innovative approaches to remote sensing in NASA’s Earth System Science Pathfinder (ESSP) program. Proceedings, SPIE 8866, Earth Observing Systems XVIII, Aug 25–29, San Diego.

Pope CA III, Dockery DW. 2006. Health effects of fine particulate air pollution: Lines that connect. Journal of the Air & Waste Management Association 56(6):709–42.

Tinkle S, Gant M, Humble M, Foley G, Garcia V, Bond A. 2007. Integrated Earth Observations: Application to Air Quality and Human Health. Washington: US Environmental Protection Agency.

UNEP [United Nations Environment Program]. 2005. Urban Air Quality Management Toolbook. Nairobi.

UNICEF [United Nations Children’s Fund]. 2019. Silent ­Suffocation in Africa. New York.

 


[1]  Aerosol “type” refers to particulate matter mixtures composed of different proportions of particle sizes, shapes, and compositions.

[2]  Phase C in NASA parlance.

[3]  The reader is also referred to an ongoing study of the ­National Academies of Sciences, Engineering, and Medicine, Lessons Learned in the Implementation of NASA’s Earth Venture Class (https://www8.nationalacademies.org/pa/projectview.aspx? key=52158).

[4]  Sources of PM10 include grinding processes and windblown soil, while PM2.5 originates from combustion in motor vehicles, power plants, residential wood burning, forest fires, agricultural burning, and industrial processes.

[5]  https://science.nasa.gov/earth-science/earth-system- observat ory

[6]  Raw WRF-Chem estimates of PM concentrations will be bias corrected using surface monitor measurements in a GRM framework similar to that used with the satellite data. Gap-filled PM estimates will consist of weighted averages between the satellite-based retrievals (where available) and the CTM-based results.

[7]  While the satellite, surface monitor, and CTM data products and datasets will be publicly available, health records will be ­handled only by professionals trained in procedures for ­maintaining patient privacy and data security.

[8]  In contrast, Earth Venture Missions include spacecraft and launch services as part of the proposal package, and the Earth Venture Suborbital line enables campaign-based investigations typically involving aircraft platforms.

[9]  The ESSPPO maintains a separate “accommodations” budget that is outside of the PI-managed cost cap, helping defray costs associated with late-developing interfaces and mitigating associated risks.

[10]  The MAIA proposal included elimination of one camera as a potential descope, but evaluation of the science/applications impact was not possible until the primary target areas were ­selected.

[11]  The State Department is assisting with the shipment of MAIA equipment overseas via diplomatic pouch, and will host certain instruments at embassies or consulates.

[12]  USAID is funding these and capacity building in Ethiopia and South Africa.

[13]  Ground monitor deployments need to be accomplished well in advance of launch to (i) generate an initial set of GRMs (using ground-based aerosol sunphotometers as a proxy for MAIA) and (ii) give the operators ample time to become familiar with equipment and prepare for routine operation in synchrony with the satellite overpasses. Deployment too early consumes limited project resources; deployment too late risks unpreparedness for mission start.

[14]  The ESSPPO accommodations budget helps offset the impacts of certain circumstances external to the project (e.g., launch delays, government shutdowns—or coronavirus pandemic ­lockdowns).

About the Author:Stacey Boland is the MAIA project system engineer and David Diner the principal investigator, both at the Jet Propulsion Laboratory, California Institute of Technology.