Conference Agenda

Capella Space is the first American company to design, build and operate a constellation of small Synthetic Aperture Radar (SAR) satellites. Its first commercial satellite was launched in August 2020. The Capella Space constellation has been growing since then, as well as the technical capabilities of the system. Among others, a novel repeat tasking pattern was introduced in 2022, in support for change-detection applications and as a first step to assess the interferometric SAR (InSAR) performance of the system by employing opportunistic interferometric collects.

The Capella 100-kg class spacecraft and its radar system have been designed to support interferometry, as the radar system has an ultra-stable oscillator and the spacecraft has a propulsion system.

The very high resolution provided by the Capella agile satellites and the diversity of observation geometries will enable new possibilities for interferometry. On one hand the high resolution of the system will enhance a wide range of applications, which require a high level of detail, as the monitoring of open pit mines, infrastructure, urban areas, etc. On the other hand, the diversity of the observation geometries, which can be obtained with the Capella system, will allow obtaining 3-D high sensitive ground deformation estimations, by properly combining deformation from each radar line of sight.

The goal of this contribution is to present the on-going repeat-pass InSAR demonstration activity, which approaches the attainment of interferometry for the Capella satellites in a systematic way. Different aspects will be discussed, related to satellite repeat ground tracks (RGT), expected InSAR performances, observation geometries, interferometric results obtained with our in-house built processor, etc.

GaoFen-3 (GF3) is the first Chinese civilian high-resolution C-band SAR satellite and part of the CHEOS (China High Resolution Earth Observation System) project to provide high-resolution observations and disaster monitoring. The first satellite was launched in August 2016. The recent ones were launched successively in November 2021 and April 2022, forming a three-satellite constellation. The first GF3 satellite was initially designed for marine science, with its primary users being the State Oceanic Administration (SOA). However, several studies have already demonstrated that GF3 is capable of doing interferometry. Starting from 2023, with GF3B and GF3C now in operation mode, InSAR-related applications will be planned as part of the daily missions for GF3 constellations.

Currently, the InSAR community lacks a comprehensive overview of GF3's capability in doing InSAR and multi-temporal analysis for a few reasons. First, with the first satellite's primary mission as in marine science, there were insufficient repeat-pass data over land, partly limiting the data source for InSAR applications. Secondly, some InSAR-related technical issues remain to be solved for GF3. For example, some research mentioned the orbital error as a limiting factor. Others mentioned that the spatial baseline for GF3 might be causing significant spatial decorrelation. Providing a quantitative evaluation of GF3's interferometric capability and performance is still essential. Last but not least, there are not enough open-source platforms currently supporting the InSAR and multi-temporal persistent scatterers interferometry (MT-PSI) processing of GF3. Among those few supporting platforms, as we have tested, some still give bugs, preventing the science community from using GF3 freely for their applications.

In our study, we carried out a number of interferometry and multi-temporal analysis for the GF3 constellation data, aiming at providing a first glimpse at GF3's performance. Specifically, the following topics are studied. First, we investigated the interferograms and coherences for approximately 30 repeat-pass images, aiming at giving a quantitative analysis between the coherence and GF3 system parameters, such as the baselines. Second, we studied the interferograms between the three satellites to understand the InSAR performance for the constellation. At last, we performed MT-PSI analysis for GF3 stacks and compared its outcome with Sentinel-1 (S1) results.

Our study shows that the GF3 constellation has very good interferometric capability. What is more, this InSAR capability is routine, not random. Among the five tracks and more than 60 GF3 constellation data we received, most showed considerable good coherence. Spatial baselines are mostly well controlled, and the spatial decorrelation is acceptable. We also generated interferograms for the newly launched GF3B and GF3C, as well as the interferograms between GF3B/C and previously launched GF3. All InSAR products demonstrated relatively good coherence, making it possible to carry out more domain-specific InSAR applications.

For the first time, an MT-PSI analysis for GF3 using 23 images in 2 years was carried out, giving a very promising result. For benchmarking, we ran an S1 processing using similar parameters (for example, number of images, time span) for the same AoI. The estimated velocity, time series, and height for most of the delivered points were highly consistent for both datasets. In our AoI of 11km*3km, after carefully selecting a reasonable threshold based on the statistics of temporal coherence of the PS points, S1 returned 72,582 valid PS time series. On the other hand, GF3 returned 301,499 PS time series, equivalent to a density of 9,136 points per km squared. Consider S1's pixel spacing (14.1mx2.3m) to be roughly 5.6 times the pixel spacing of GF3's stripmap FSI mode (2.5mx2.3m), and the fact that the number of delivered points above the threshold for GF3 is 4.15 times the numbers for S1, then the GF3 result in terms of point density is already quite close to the theoretical upper bound using S1 as a benchmark. The result demonstrated excellent MT-PSI performance for the GF3 data stack and revealed great potential for future MT-PSI applications using the GF3 constellation.

Along with our study, we addressed some issues in the InSAR processing chain for the GF3 dataset. First, we have implemented a network-based orbital ramp removal method based on FFT and frequency modulation estimation. In the second place, we have also implemented several small processing steps for GF3, including a common band filter. We have also tried several ways of coregistration and evaluated the accuracy of orbit state vectors for GF3. Finally, we are implementing all our work in open-source InSAR processing software. The InSAR part is implemented in RIPPL (Radar Interferometric Parallel Processing Lab, the TU Delft's next-generation DORIS), and the MT-PSI part is implemented in GECORIS (GEodetic COrner Reflector InSar toolbox). We want our work to be reproducible for the InSAR community and facilitate future InSAR applications for the GF3 constellation.

We suggest to present a mission proposal submitted to the Earth Explorer 12 call that aims to address and quantify dynamic processes in cold environments by measuring the static and dynamic topography. This information is essential for understanding, modelling and forecasting the dynamics and interactions within the different elements of the cryosphere and with other Earth system components. The mission proposal will provide very accurate high-resolution, multi-temporal topographic data that will make it possible to derive mass balances and structural changes in the cryosphere, with a focus on permafrost areas as well as glaciers and ice caps, ice sheets and sea ice. At the same time, the mission proposal will enable unprecedented measurements of volume change processes in the geosphere, including volcanic, landslide and seismic activities. In addition, the mission proposal will generate a global digital elevation model (DEM) of about one order of magnitude better, in terms of resolution and height accuracy, then the current reference provided by TanDEM-X.

The instrument consists of a cross-platform Ka-band radar interferometer with two spacecraft that fly in a reconfigurable formation and can dynamically adapt to the needs of scientific observation. Cross-track SAR interferometry is an established remote sensing technique for large-scale measurements of static and dynamic topography and the use of Ka-band minimizes systematic biases and errors that would be caused at lower frequencies due to wave penetration into semi-transparent media. The mission proposals unique ability to provide time series of highly accurate surface topography measurements allows the mission’s primary scientific objectives to be optimally fulfilled, namely

a) the monitoring of permafrost degradation by means of DEM acquisitions with short repetition intervals and estimates of volume changes in time,

b) the measurement of snow topographic changes to observe different snow regimes to feed hydrological models for a more precise prediction of water availability, and

c) the measurements of glaciers, ice caps, ice-clad volcanoes and their mass balance and modelling of ice dynamics and ice/climate interactions.

At the same time, the SKADI measurements allow to serve a number of secondary science objectives related to floating ice and geosphere applications such as

a) the measurement of sea ice and fresh water ice topography to define the surface-air-interface,

b) the monitoring of geohazards involving large deformations and volume changes caused by landslides, glacier lake outbursts, rockfalls, mining, landfill, volcanic activities and seismic events,

c) the measurement of a global DEM with unprecedented resolution and accuracy.

The mission proposal will moreover complement and fill critical observation gaps of the current Copernicus and Earth Explorer missions (e.g., Sentinel, Cryosat) by providing frequency diversity and enhanced spatial resolution, while at the same time offering the Earth Observation community and future ESA missions (e.g., Aeolus, EarthCARE) a global topographic reference of superior accuracy and resolution, which enables a major step forward in improving the quality and interpretation of a vast amount of past, present, and future Earth observation data. A first order performance of the mission products reveal that very accurate DEM change products can be expected, with accuracies in the order of decimetres to centimetres.

In comparison to previous SAR missions, the short wavelength in Ka-band allows a reduction of the size and weight of the antennas and spacecraft and enables the joint launch of two radar satellites with a single medium-sized launch vehicle like Vega C. In this regard, the mission proposal space segment offers also a unique platform to explore and demonstrate new bi- and multistatic SAR techniques, technologies and applications which are expected to shape the future of radar remote sensing.

The mission concept and the associated space segment have been developed in two Pre-Phase 0 studies in close collaboration with Airbus DS and OHB. Both industry partners proposed innovative Ka-band SAR instrument architectures and showed the feasibility of the current mission proposal within the programmatic constraints and the cost cap provided by ESA in its call for Earth Explorer 12 mission ideas. We believe that the versatility and technological innovation of the mission proposal are an important complement to its unique scientific objectives, thereby increasing its impact on societal welfare.

The consolidated mission objectives, the well-defined mission products, the highly accurate performance and the innovative instrument design will be presented. Following the encouraging recommendations provided in the ACEO (Advisory Committee for Earth Observation) report from the EE-11 call, the mission proposal team is working towards the submission of a revised version of the mission proposal for the ESA Earth Explorer 12 call.

In recent years, the availability of low-cost small satellites and the innovation of constellations have resulted in an increasing number of commercial companies who have established business models to provide information services fed by their own satellite systems. These new space players are now playing an important role in the EO international strategy. Some of these new missions are already part of, or potential candidates, for the Earthnet Third Party Missions (TPM) programme of the European Space Agency (ESA). The TPM programme allows the access of European users to a large portfolio of EO data in addition to the ESA-owned EO missions.

The Earthnet Data Assessment Project (EDAP+) is a continuation of its predecessor EDAP (2018-2021) who’s main goal is to assess the quality and suitability of Earth Observation (EO) missions included or being considered for ESA’s Earthnet TPM. The key objective of ESA's EDAP+ is thus to take full advantage of the increased range of available data from non-ESA operated missions and to perform an early data quality assessment for various missions that fall into one of the following instrument domains:

• Optical missions
• SAR missions
• Atmospheric missions
• AIS (Automatic Identification System) & RF (Radio Frequency) missions

The SAR mission quality assessment is based on specific guidelines and usually covers the following aspects:

• Data Provider Documentation Review: the assessment covers the products information, metrology, and products generation topics. The goal of this assessment is to evaluate the quality of the documentation provided to the users in terms of products formats, generation and calibration, and of the availability and accessibility of the SAR products.
• Independent validation of the data quality by analyzing ad hoc datasets of the third-party missions over calibration sites (e.g., point target calibration sites or homogeneous areas) in order to verify the overall data quality in terms of Impulse Response Function characteristics, spatial resolution, radiometric calibration, geolocation accuracy and noise level.

In the past, the focus of the EDAP activities was the assessment of L1 products quality, delegating to the users the assessment of the suitability of the products for interferometric applications. The goal of the EDAP+ project is to start defining a framework for the assessment of the quality of interferometric products [1] [2], which could be generated operationally in the future.

The assessment of the InSAR quality of a SAR mission will address:

• The availability of L1 products for the generation of InSAR stacks (based on operational acquisition planning or users’ acquisitions tasking)
• The quality and suitability of L1 data to InSAR applications.

The latter assessment includes the following quality parameters:

1. Interferometric baseline computed from the orbits annotated in the products.
2. Doppler Centroid annotated in the products.
3. Interferometric coherence from interferograms generated applying co-registration from orbit only.
4. Interferometric coherence from interferograms generated applying co-registration refinement from data, e.g., enhanced spectral diversity (ESD) or incoherent speckle tracking.
5. For quad pol data, comparison of the HH and VV coherence.

The present contribution provides an overview of the EDAP+ activities and focuses on the quality assessment of the interferometric products of two SAR missions: ICEYE and SAOCOM.

ICEYE is a commercial SAR satellite manufacturer and service provider founded in Finland in 2014. As of the beginning of 2023, the ICEYE constellation includes more than 20 X-band SAR satellites. Over the next years, ICEYE will continue to grow its constellation capacity in specialized orbital planes designed to provide persistent monitoring capabilities and high-resolution view of the Earth's surface. Currently, the satellites support operation in the imaging modes called 'Strip', 'Spot' and 'Scan'.

SAOCOM is an L-band twin-satellite SAR constellation operated by the Argentinian space agency (CONAE). The two satellites; SAOCOM 1A and 1B, were launched in October 2018 and August 2020, respectively. Together they allow a revisit time of 8 days. SAOCOM is the first L-band mission implementing the TopSAR acquisition mode. However, burst synchronization is not performed, and therefore the TopSAR data is not ideal for interferometry.

The InSAR quality assessment for EDAP+ includes data of the Strip and Spot imaging modes of ICEYE, and data of the StripMap imaging mode of SAOCOM. In this contribution, the methods and the results of the InSAR data assessment of the ICEYE and SAOCOM missions within the EDAP+ project will be presented.

References

[1] Marinkovic, P., Ketelaar, VBH., van Leijen, FJ., & Hanssen, RF. (2008). InSAR quality control: Analysis of five years of corner reflector time series. In H. Lacoste, & L. Ouwehand (Eds.), Fifth International Workshop on ERS/Envisat SAR Interferometry, `FRINGE07', Frascati, Italy, 26 Nov-30 Nov 2007 (pp. 1-8). ESA Communication Production Office.

[2] Geudtner, D., Prats, P., Yague-Martinez, N., Navas-Traver, I., Barat, I., & Torres, R. (2016). Sentinel-1 SAR Interferometry Performance Verification, Proceedings of EUSAR 2016: 11th European Conference on Synthetic Aperture Radar, Hamburg, Germany (pp. 1-4).

Synspective, a Japanese startup founded in February 2018, has been developing small X-band SAR satellites and aims to build a constellation of 30 satellites by the late 2020s. The basic technology of the small SAR satellite was developed under a Japanese government program called ImPACT (Impulsing PAradigm Change through disruptive Technologies) from 2015 to 2018. The constellation of 30 satellites will enable us to observe any area on Earth within two hours of a natural disaster, helping us to quickly understand the hazard situation and spatial extent in detail from space. Frequent data collection and surface monitoring in normal times will also contribute to risk assessment and disaster prevention in human society. We have also developed and provided solution services using SAR data, such as land displacement monitoring.

The StriX constellation consists of small SAR satellites in the 100 kg class with a 4.9 x 0.7 m antenna. Its central frequency is 9.65 GHz (X-band). The satellites can observe either the right or left direction. The polarization is VV. The off-nadir angle can range from 15 to 45 degrees. Two observation modes are available: Stripmap and Sliding Spotlight. Stripmap mode has ~3 m spatial resolution (75 MHz bandwidth), 20 km nominal swath width, and 50-70 km azimuth length. Sliding Spotlight mode has a spatial resolution of ~1 m (300MHz bandwidth) and a nominal swath width of 10 x 10 km. Available product formats are SLC (CEOS or SICD) and GRD. SARscape, SNAP, and GAMMA can process StriX SLC data (as of February 2023).

Synspective launched the first demonstration satellite, StriX-α, on December 15, 2020, and successfully acquired the first image on February 8, 2021. The second demonstration satellite, StriX-β, was launched on March 1, 2022, and the third satellite, StriX-1, which is also the first commercial prototype satellite, was launched on September 16, 2022. Many images have been acquired and delivered to customers by the three satellites. The revisit period of StriX-β and StriX-1 is one day, allowing for continuous daily monitoring. Some high-coherence interferograms have also been obtained.

In this presentation, we will report on the current status of the StriX constellation, the first InSAR results, and future plans.