8

Priorities for Maintaining and Enhancing the Geodetic Infrastructure

The previous chapters discussed the role of the geodetic infrastructure, its current state, and future requirements for answering selected science questions from the Decadal Survey (NASEM, 2018). The geodetic infrastructure required to help answer each of these science questions is given in Appendix A. This chapter identifies priority improvements to the geodetic infrastructure that would facilitate advances across those science questions (Task 4). These improvements are organized into five themes: (1) accuracy and stability of the terrestrial reference frame (TRF), (2) accuracy and stability of satellite orbits, (3) accuracy of the low-degree geopotential harmonics, (4) augmentation of the Global Navigation Satellite System (GNSS) station network, and (5) analytical support for an enhanced geodetic infrastructure.

ACCURACY AND STABILITY OF THE TRF

Questions on sea-level rise, terrestrial water cycle, and geological hazards require improvements in the accuracy and stability of the TRF. The most stringent science requirements are driven by sea-level science needs, which are quantified in terms of allowable errors in the rates of sea-level rise. We thus describe these limits in terms of reference frame accuracy and drift (see definitions in Box 1.2).

The sea-level science questions require a TRF accuracy of 1 mm and drift in the origin of the TRF of less than 0.1 mm/yr (or less than 0.02 ppb/yr in scale-rate equivalent). Meeting these requirements would allow global sea-level rise to be determined to an accuracy of better than 0.5 mm/yr over the course of a decade (Objective C-1a) and regional sea-level rise to within 1.5–2.5 mm/yr over the course of a decade (Objectives C-1d and S-3a). The TRF should be free of deformations due to ancient and modern ice melt that might cause errors in the regional patterns of sea-level change. The signals in the motion of the Earth’s center of mass are expected to vary by as much as 50 mm in the next 100 years. There must be commensurate stability of the reference points for metrology at the fundamental sites, such as the invariant points of Satellite Laser Ranging (SLR) telescopes or Very Long Baseline Interferometry (VLBI) dishes, or the GNSS monumentation. This may require studies on the stability and longevity of monumentation and drifts or stability of the tracking equipment. Finally, the tide gauge record must be maintained to validate the satellite altimetry data in order to achieve 0.1 mm/yr accuracy in the altimeter measurements averaged over a decade (Objective C-1a).

The TRF accuracy and drift requirements are somewhat less stringent for the terrestrial water cycle and geological hazards questions. The water cycle questions require that the center of mass drift rate be maintained to better than 0.2 mm/yr (Objectives H-2b and H-2c). Monitoring surface deformation associated with geological hazards requires that the TRF be maintained at an accuracy of 0.5 mm/yr globally (Objectives S-1a, S-1b, S-2a, and S-2b). The current accuracy of the TRF is sufficient for the Decadal Survey weather and climate and ecosystems science questions.

Three areas of improvement in the geodetic infrastructure are needed to meet the above requirements.

First, despite long-standing efforts, the next-generation VLBI and SLR systems have either not been installed or have not been fully tested (see Chapter 2). Deployment of the new systems, particularly in the southern hemisphere, is critical for maintaining the highest accuracy of the TRF. The definition of the Earth’s center of mass, especially in the Z-component, is especially dependent on successful tracking of SLR in the southern hemisphere. The National Aeronautics and Space Administration (NASA) will need to continue coordinating with international partners to improve the balance of fundamental stations with VLBI, SLR, and GNSS between the northern and southern hemispheres. Second, modeling of the center of mass motions expected over the next 100 years, due to the melting of the Greenland and Antarctic ice sheets, produces a large drift that can be monitored only if the monumentation of the fundamental sites remains stable over that time period (see Chapter 3). An assessment of the long-term stability of the monumentation may be required. Third, moving toward a fully time-dependent TRF would accommodate long-term (10–100 years) variations in the center of mass due to ice sheet melting, seasonal variations in the center of mass due to redistribution of water over the Earth, and short-term variations in the center of mass caused by large earthquakes and their postseismic deformation (Altamimi et al., 2019).

ACCURACY AND STABILITY OF SATELLITE ORBITS

The Decadal Survey science questions place different requirements on the accuracy and stability of satellite orbits. The highest accuracy of orbit determination is needed for low-Earth orbiting radar and laser altimetric satellites used to measure and interpret sea-level change and ice-sheet elevation changes (Objectives C-1a, C1-b, and C1-d). The requirements for their precision orbit determination act in concert with the requirements for TRF stability, particularly for measuring the rate of sea-level rise. Altimetric measurements with an accuracy of 20 mm or better and a stability of less than 0.5 mm/yr over a decade are required (Objectives C-1a and S-3a). The associated orbit determination requirements are 10–20 mm radial position accuracy. Three-dimensional orbit accuracy of better than 0.1 m is required for ice-sheet flow-rate measurements using Interferometric Synthetic Aperture Radar (InSAR; Objective S-3a).

The orbit determination and clock requirements for the weather and climate questions (Objectives C-2b, W-1a, and W-1b) are less stringent. For integrated water vapor, the GNSS orbits need a three-dimensional root mean square (RMS) accuracy of better than 50 mm in near-real-time and better than 25 mm post processing. For radio occultation, the low-Earth orbiting satellites need clock estimates every 30 seconds, with a velocity accuracy better than 0.5 mm/s RMS in near-real-time and better than 0.07 mm/s RMS post processing. Orbital accuracies need to be better than 0.21 m in real-time and better than 0.12 m post processing.

Orbit determination requirements for InSAR satellites are driven by terrestrial water cycle (Objectives H-2c, S-6a, and S-6b) and geological hazards (Objectives S-1a and S-1b) questions. Answering these questions requires sub-cm deformation measurements with high spatial density (<100 m), which can be achieved through a combination of GNSS stations having a spacing of better than 40 km and weekly InSAR coverage being provided by Sentinel-1 and soon NASA-ISRO Synthetic Aperture Radar (NISAR). The orbit accuracy requirements are similar to the requirements for satellite altimetry, with an accuracy of 20 mm radially and 60 mm along-track.

Enhancements to the geodetic infrastructure will be needed to meet the related requirement of bare-earth topography for geological hazards, vegetation structure, and carbon and water fluxes (Objectives E-1a, E-1b, E-2a, E-3a, S-1b, S-1c, S-2c, and S-4a), with 0.1 m vertical accuracy over selected tectonic areas and the attendant need for local GNSS ground stations for differential GNSS aircraft positioning better than 50 mm. For ecosystems science questions, maintenance of the current geodetic infrastructure is essential for delivering the current capability of 20 mm orbit accuracy and 40–70 mm along-track orbit position of lidar imaging (Objectives E-1a, E-1b, E-1c, E-1d, E-2a, E-3a, and S-4a).

ACCURACY OF THE LOW-DEGREE GEOPOTENTIAL HARMONICS

The same geodetic infrastructure and data (GNSS and SLR tracking) that provide the orbits for GNSS and altimeter satellites also enable determination of

the long-wavelength components of the Earth’s time-variable gravity field. The long-wavelength gravity field is needed for Decadal Survey questions related to determination of ocean mass (Objective C-1a), changes in ice sheets (Objective C-1c), temporal variations in total water storage of midsize basins (>200 km; Objective S-4a), and gravity change for large subduction zone earthquakes (Objectives S-1a and S-1b). The geodetic infrastructure enables a unique determination of the geocenter or degree-1 harmonics, provides validation for other long-wavelength components of the gravity field from dedicated gravity missions, and helps fill the gaps when no dedicated gravity missions are flying. Maintenance of the current geodetic infrastructure is essential for the continued availability of measurements of large-scale mass exchange in the Earth system.

AUGMENTATION OF THE GNSS STATION NETWORK

The stations of the GNSS network that define the global terrestrial reference frame must meet the highest standards for data quality, site design, stable monumentation, and metadata definition and dissemination. Global, national, and regional reference frame needs require a high-density network of such stations operating continuously, with a free and open dissemination of data with low latencies.

For the Decadal Survey science questions, the geographic coverage and density of the GNSS network are driven by the need to characterize large-scale plate boundary deformation and plate motions with an accuracy of 0.5 mm/yr (Objectives S-1a and S-1b). Tide gauges need to have co-located GNSS receivers that are part of this network. Reflectometric GNSS receiver installations can augment traditional tide gauges by simultaneously measuring sea level and vertical land motion (Objective C-1a). At regional and smaller scales, increased GNSS density is needed to calibrate InSAR and lidar techniques for terrestrial water cycle and geological hazards science questions (Objectives H-2a, H-2c, H-4a, S-1a, S-1b, S-1c, S-6a, and S-6b).

The terrestrial water cycle, geological hazards, and ecosystems chapters discuss the need for an increase in the density of core GNSS stations in the United States with good monument stability, long-duration time series (>10 years), and high data rate (~1 Hz). These stations would improve measurements of the elastic response of the Earth to changes in water loading (Objectives H-2b), provide measurements for correcting the long-wavelength errors in InSAR due to unmodeled atmospheric and ionospheric errors (Objectives E-1a, H-2b, S-1a, and S-1b), and allow estimation of soil moisture, snow water equivalent, and vegetation water content using reflectometry (Objectives E-1d and W-2a). In coordination with the International GNSS Service (IGS), these stations could become a permanent U.S. contribution to the global geodetic infrastructure.

In addition, having additional core GNSS receivers on remote islands or GNSS buoys would support climate change questions (Objective C-2b), and having them on ocean platforms would support seafloor geodesy and tsunami forecasting (Objective S-1d).

SUPPORTING SOFTWARE, MODELS, DATA, AND EXPERTISE

To gain the full benefit of enhancements to the geodetic infrastructure discussed above, software, models, open data archived to scientific specifications, and a skilled workforce have to be maintained.

Open Data, Cyberinfrastructure, and Workforce

Geophysical data analyses supporting Decadal Survey science questions must be able to utilize the International Terrestrial Reference Frame, which requires free, open, and timely access not only to the source geodetic data but also to high-quality software tools and automated processing. For example, GNSS applications connected to the terrestrial water cycle, geological hazards, atmospheric monitoring, and ecosystems require access to software for modeling or utilizing high-quality GNSS clocks and orbits, antenna phase center calibrations, software for GNSS reflections, and, in some cases, access to automated processing services. All software systems used by geodesists require high-quality metadata standards, which allow users to properly model changes at a site caused by changes in the equipment, firmware, or in some cases, the site itself (e.g., an earthquake). Similarly, InSAR applications connected to geological hazards, terrestrial water cycle, and ecosystems require open access to raw Synthetic Aperture Radar (SAR) data, accurate orbital information, and two or more open software developments

to continue advancing InSAR as a geodetic tool similar to GNSS. An important component of both the GNSS and InSAR infrastructure is the development of new software delivery tools to make these data available seamlessly to more users. The dramatic improvement in satellite orbits and clocks has enabled automated processing of very large sets of repeated observations (e.g., SAR, optical, radar altimetry, and lidar) that was not possible just a few years ago. This advance is important because the data sets are too large for a human to be in the processing loop. A continued linkage between accurate orbits, models, and automated software will enable the improvement of climate models in the coming decades. Developments in cyberinfrastructure will require an evolving workforce that can maintain institutional knowledge and technical capabilities of the geodetic infrastructure and also work in close collaboration with the high-performance computing community.

Ancillary Corrections and Models

A significant component of the geodetic infrastructure is the ancillary corrections and models used to achieve cm-level accuracy for all the geodetic methods. These models need to be maintained for the continued accuracy of the TRF, but they also need to evolve as the time series are extended and the measurements improve. Three types of models are important: time-variable gravity, time-variable surface deformation, and atmospheric and ionospheric propagation models. A time-variable gravity model is needed to maintain the TRF as well as achieve the cm-level accuracies of the low-Earth orbiting geodetic satellites. Continuation of Gravity Recovery Climate Experiment (GRACE)-type missions is needed to augment the low-degree gravity variations that are determined from SLR analysis. Maintenance of time-variable gravity models is needed for sea-level change (Objectives C-1a, C-1c, C-1d, and S-3a), terrestrial water cycle (Objectives H-2b, H-2c, and S-6b), geological hazards (Objectives S-1b, S-1c, and S-2c), and ecosystems (Objective E-2c).

Time-variable surface deformation models are associated with numerous processes, including plate motions, large earthquakes, elastic loading from ocean tides, ice loss, redistribution of surface water, atmospheric pressure variations, and viscous rebound associated with glacial cycles. As discussed in Altamimi et al. (2019), these models are used to constantly update the TRF, so there is a close connection between TRF accuracy and model accuracy. Improving these models requires collaboration between the scientists who develop the models to understand Earth processes, and the geodesists who maintain the TRF.

Atmospheric and ionospheric propagation models are needed to correct path delays of all of the main components of the geodetic infrastructure: VLBI, SLR, GNSS, and Doppler Orbitography and Radiopositioning Integrated by Satellite. As discussed in Chapter 6 (weather and climate), the GNSS geodetic infrastructure is used directly to measure path-delay variables, such as integrated water vapor and total electron content of the ionosphere. Accurate atmospheric models (for altimeters, GNSS, and InSAR) are needed to maintain the accuracy of the TRF, which again requires close collaboration between the scientists and TRF geodesists.

Finally, an important enhancement to the GNSS infrastructure is to upgrade the global IGS sites (hardware and products) to achieve Global Positioning System (GPS)-like accuracies for the other constellations (e.g., Galileo, Glonass, and Beidou). A significant upgrade would result in a dramatic increase in the number of radio occultations for weather and climate applications. This requires the support of multiple GNSS analysis software systems within the United States and moving from current GPS-only orbit and clock production.

SUMMARY

All of the active satellite systems recommended by the Decadal Survey (e.g., SAR, radar altimetry, lidar, and radio occultation) rely on very accurate three-dimensional orbital information to obtain the required measurement of range change; the accuracy of the range-change measurement is directly related to the accuracy of the orbit. While some passive satellite systems do not need decimeter or better orbital accuracies to achieve their imaging requirements, the availability of accurate orbits has enabled fully automated processing and accurate geolocation, which increases the exploitation of the large data sets being collected by Decadal Survey missions.

The accuracy and stability of satellite orbits relies on the accuracy and stability of the TRF, which is derived from the geodetic infrastructure. The committee

identified three areas of improvement in the geodetic infrastructure needed to help answer the Decadal Survey science questions:

1. Finalize deployment and testing of next-generation VLBI and SLR systems and complete deployment of multi-GNSS to achieve a balance of geodetic measurement techniques between the northern and southern hemispheres, document the errors in the systems, and improve our ability to estimate their positions accurately and automatically.
2. Increase the capabilities for measuring the center of mass motions expected over the next 100 years, due to the melting of the Greenland and Antarctic ice sheets.
3. Work with the international community to implement a fully time-dependent TRF that will accommodate annual as well as sudden changes in the locations of the fundamental stations.

The most stringent requirements for enhancements to the accuracy and stability of the TRF are driven by science questions related to sea-level change, ice-mass loss, and land-surface deformation associated with (a) the movement of water over the surface of the land, cryosphere, and oceans; and (b) the elastic and viscoelastic response of the solid Earth to water loading, earthquakes, and volcanic eruptions. If any of the associated flagship missions of the current NASA program of record (e.g., NISAR; Ice, Cloud, and land Elevation Satellite 2; GRACE-Follow On; and Surface Water Ocean Topography) had a failure of its on-board GNSS systems, it is not clear that the ground-based SLR tracking network (mostly international) would have sufficient capacity to handle the increased load.

Ground-based GNSS is essential for achieving the Decadal Survey science objectives related to sea level, cryosphere, terrestrial water cycle, weather, climate, geological hazards, and ecosystems. The density of core GNSS stations needs to be increased in high priority regions, including plate boundary zones to capture the earthquake cycle, coastlines to capture land motion that could affect sea-level impacts and coastal ecosystems, and regions with substantial terrestrial water storage. In addition, the United States will need to work with the International GNSS Service to deploy additional GNSS sites in remote, rapidly deforming areas, such as the perimeters of the ice sheets that deform by changes in mass loading. Such sites need good monument stability, long duration, and high data rate and availability. The U.S. stations should be considered part of the U.S. geodetic infrastructure, open to everyone, and thus have long-term financial support. Many of these stations already exist, but they are supported mainly through the National Science Foundation and thus long-term funding is not guaranteed.

Maintaining and enhancing the geodetic infrastructure to compute the TRF, satellite orbits, and other products requires complex software systems developed over decades by teams of scientists and engineers. The software systems ingest both the raw measurements from the geodetic infrastructure and models for the steady and tidal deformation of the Earth and for propagation of the electromagnetic waves through the ionosphere and atmosphere. The most important aspects of this activity are that all of the raw data are completely open and that there is cross-checking by at least two independent groups using largely independent and open software. Needless to say, this relies on a skilled geodetic workforce. Unfortunately, several federal agencies noted the difficulty of finding scientists and engineers with the skills needed to replace the pool of aging geodesists. On-the-job training of graduate students is becoming increasingly important for agencies involved with the geodetic infrastructure.

REFERENCE

This page intentionally left blank.