Conference Agenda

please check the attached PDF document

The National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA) are both developing future L-band SAR missions to address key science questions and application needs relevant to solid Earth, ecosystems, cryosphere, and hydrology. The NASA-ISRO SAR (NISAR) mission is a dual-frequency L-/S-band SAR satellite scheduled for launch in late 2023 and recently identified as a pathfinder for the NASA Earth System Observatory. NISAR will acquire global dual-polarimetric L-band data with 20 MHz range bandwidth every 12 days (6 days ascending and descending), delivering unprecedented dense time-series at L-band and new Geocoded Single Look Complex products. The Radar Observation System for Europe at L-band (ROSE-L) is one of the ESA’s High-Priority Candidate Missions scheduled for launch after 2028 with the goal of augmenting the Copernicus constellation to address important information gaps and enhance existing Copernicus services and related applications. In the current design, ROSE-L is a two-spacecraft system that will operate in the Sentinel-1 orbit and be phased to achieve a repeat interval of 6 days.

Both NISAR and ROSE-L are designed to make repeated observations from a narrow orbital tube in order to generate time-series with nominal zero interferometric baselines. While this design choice has several benefits, it cannot address some of the measurements recommended by the 2017-2027 Decadal Survey for Earth Science and Applications. Two of these measurements are (1) 3D surface deformation vector and (2) vegetation vertical structure, for which long along-track and cross-track baselines, respectively, are required. NASA has been conducting dedicated studies to develop science and application traceability matrices (SATMs) as well as identify technology gaps and candidate architectures for Surface Deformation and Change (SDC) and Surface Topography and Vegetation (STV) measurements [1].

This paper analyzes the performance of concepts involving satellites flying in formation with satellites such as NISAR and ROSE-L in order to augment the observation capabilities of these missions through denser coverage, multi-squint or multi-baseline measurements. Receive-only co-fliers are attractive thanks to their simplified hardware architecture and to the ability to coherently combine their images without relying on a tight cooperation with the mothership SAR satellite. The talk addresses challenges and opportunities of proposed free- and co-flier concepts for NISAR and ROSE-L by leveraging previous and current studies being conducted at NASA (e.g., DARTS [2]) and ESA (e.g., SAOCOM-CS).

The analysis is carried out by varying the radar instrument and formation parameters (e.g., the bistatic angles, the perpendicular baseline) in pre-defined distributed formations with one to six satellites with non-zero along-track and/or across-track baselines. Three multi-static modes are considered for each scenario: SISO (single-inputs-single-output), SIMO (single-input-multiple-output), and MIMO (multiple-input-multiple-output). Performance is derived from closed-form equations where available, as illustrated in Figure 1, as well as from point-target and distributed-target simulations using the DARTS Trade Study Tool (TST). The DARTS TST provides the ability to analyze the global system performance taking into account orbital, radar signal, and scene characteristics that would be too complex for compact and closed-form analytical models. We plan to present the SDC/STV retrieval performance for various realistic co-flier formation configurations. Other aspects specific to NISAR and ROSE-L, such as the cooperation of the co-fliers with the sweep-SAR or scan-on-receive imaging modes, will be also discussed.

[1] A. Donnellan, D. Harding, P. Lundgren, K. Wessels, A. Gardner, M. Simard, C. Parrish, C. Jones, Y. Lou, J. Stoker, J. Ranson, B. Osmanoglu, M. Lavalle, S. Luthcke, S. Saatchi, and R. Treuhaft, “Observing Earth’s Changing Surface Topography and Vegetation Structure: A Framework for the Decade,” NASA Surface Topography and Vegetation Incubation Study, Mar. 2021.

[2] M. Lavalle, I. Seker, J. Ragan, E. Loria, R. Ahmed, B. Hawkins, S. Prager, D. Clark, R. M. Beauchamp, M. S. Haynes, P. Focardi, N. Chahat, M. Anderson, K. Matsuka, V. Capuano, and Soon-Jo, “Distributed Aperture Radar Tomographic Sensors (DARTS) to Map Surface Topography and Vegetation Structure,” in 2021 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), July 2021.

As discussed in [1], synthetic aperture radar (SAR) interferograms computed from small baseline acquisitions incorporate biases due to the coherent nature of the ambiguous energy, which is especially critical in systems with small antennas. Recent investigations within the scope of ESA’s Earth Explorer 10 Harmony mission [2, 3, 4] lead to the question on how coherent ambiguity removal algorithms must be applied on highly accurate ocean data products to satisfy user needs. The Harmony system design comprises of two receive-only satellites flying e.g. in a stereo formation several hundred kilometers ahead and behind a Sentinel-1 to obtain single-platform short-baseline along-track interferometric synthetic aperture radar (AT-InSAR) [5] maps. The companion satellite systems feature reduced ambiguity suppression capabilities compared to their full-performance counterpart. Proposed techniques for removal of those coherent ambiguity biases on interferogram level rely on summing or subtracting different interferometric looks (varying spectral Doppler support) [4, 6]. The short-time behavior of the interferometric signature of the ocean scene might break down the underlying assumptions of the aforementioned algorithms, i.e., the dynamics of the sea and ocean surfaces in scales of several synthetic apertures must be thoroughly understood for the tuning of the algorithms. We propose in this contribution to evaluate the impact of fast variations of interferometric signatures with the help of TanDEM-X spotlight-mode data.

In SAR one benefits from coherent processing of radar pulses to generate high-resolution images in two spatial dimensions. When distributed scenes incorporate random-like movement the coherence property of radar backscatter is vanishing over time [7, 8, 9, 10]. This mechanism is quantified using the coherence time of the underlying random processes, which usually refers to the width of the auto-correlation function of the radar backscatter [10]. Using AT-InSAR to measure surface velocities one obtains the backscatter-weighted contributions of all velocity components, i.e. Bragg velocities, phase and orbital velocities of gravity waves and current-induced movement [11, 12]. The former is a small-scale (electromagnetic wavelength) and the latter a large-scale effect (mesoscale) with respect to the imaging resolution. Since the image data for SAR is acquired over time (not instantaneously), it captures the state of the ocean and its movement over a period of time rather than taking a snapshot. The most relevant impact on the acquired data will have waves with periods in the order of the processed synthetic aperture time [8, 9, 13].

In a first step, data acquired in spotlight mode is investigated. The mode offers a long acquisition time of the same scene and can be used to observe a time-series of the ocean interferometric signature by forming subapertures or sublooks on the scene. Subapertures are generated by partitioning the full synthetic aperture into smaller consecutive subapertures, where each aperture time is below the scene coherence time (< 100 ms). Sublooks are generated from consecutive bands of Doppler frequencies, while obeying the same subaperture time to acquire the bandwidth of one look. The analysis is also expected to provide insight on the improvement of interferometric processing techniques for retrieval of velocities or wave spectras from a broader band of Doppler frequencies beyond the coherence time of the surface, e.g. [14, 15, 16].

First processing results of spotlight data show that the AT-InSAR signatures are contaminated by changes of the ocean surface and system- and calibration-related systematic effects.

InSAR is a powerful technology that provides high-resolution, wide area, and high accuracy ground surface deformation measurements. One of its limitations is that it only measures one-dimensional motion along a single radar line-of-sight (LOS). The lack of 3D information can lead to challenging interpretation of results, limited assessment of geo-hazards and ambiguous modelling processes. To retrieve the three components of deformation, at least three non-coplanar LOS vectors are required. This can be achieved by utilizing multiple InSAR stacks acquired from different pass directions, incidence angles, and left and right looking observations.

Currently, almost all operational spaceborne SAR sensors acquire images from near-polar orbits, mostly using right-looking modes. The InSAR measurements derived from these geometrical configurations provide accurate vertical and east-west deformation measurements, however the north-south component remains poorly resolved. When the left-looking geometry is incorporated, some improvement in the accuracy of the north-south component can be achieved, depending on latitude. Unfortunately, few missions have the capability to acquire images in left and right looking modes.

Some studies have successfully used multiple geometries from right looking modes to retrieve 3D displacement, but this is limited to regions at high latitudes where angular diversity increases. Others studies have utilized the pixel offset method and multi-aperture interferometry to overcome this problem; but these methods have their own sensitivity limitations.

Utilizing SAR images acquired from an inclined orbit enables 3D deformation monitoring since it increases the angular diversity compared to current missions. This will be possible with MDA’s CHORUS mission which is currently under development and consists of a multi-sensor SAR constellation that includes a C- and a trailing X-band SAR sensor. Both satellites will provide repeat-pass InSAR capability with their Stripmap and Spotlight modes. The satellites will operate in a mid-inclination orbit (53.5 °) that will provide more frequent coverage over mid to low latitudes (± 62.5° latitude). The inclined orbit of CHORUS will provide a wider diversity of lines of sight and will enable better monitoring of 3D deformation.

In this study, we demonstrate via simulations the improvements that CHORUS-C will provide in estimating 3D deformation. Our findings show that high quality three components of deformation can be obtained by using CHORUS-C alone or in combination with SAR sensors in near-polar orbits. Figure 1 shows an example of simulated ground deformation and 3D decompositions using different geometries from RADARSAT-2 and CHORUS-C at a mid-latitude. Measurement accuracy is evaluated at different latitudes and using multiple combinations of viewing geometries. Modeling of different geophysical sources of deformation are used to demonstrate this capability and to understand the impact of noise and the temporal offset between acquisitions.