RADARSAT-2 is a jointly-funded satellite mission of CSA (Canadian Space Agency) and MDA (MacDonald Dettwiler Associates Ltd. of Richmond, BC), representing a Canadian government/industry partnership [or PPP (Public Private Partnership)] in a commercial venture. In Feb. 1998, CSA awarded a contract to MDA to built RADARSAT-2. The contract calls for MDA to develop, own and operate RADARSAT-2 and related infrastructure (including data distribution). CSA provides a fixed financial contribution to MDA (about 75%), in exchange for imagery allocation from the S/C to government agencies. 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12)
US space policy, regulations related to the export of satellite and rocket technology as well as to the distribution of high-resolution imagery, caused CSA in January 2000 to cancel an existing satellite bus contract with OSC (Orbital Sciences Corporation) of Dulles, VA, and to award a new contract to Alenia Spazio of Rome, Italy.
RADARSAT-2 is an advanced state-of-the-art technology follow-on satellite mission of RADARSAT-1 with the objective to:
• a) continue Canada's RADARSAT program and to develop an Earth Observation satellite business through a private sector-led arrangement with the federal government
• b) provide data continuity to RADARSAT-1 users and to offer data for new applications tailored to market needs.
• c) the key priorities of the mission respond to the challenges of:
- Monitoring the environment
- Managing natural resources
- Performing coastal surveillance.
Figure 1: Overview of the RADARSAT-2 development structure (image credit: MDA)
Figure 2: Artist's rendition of the RADARSAT-2 satellite in orbit (image credit: MDA)
The S/C consists of a bus module, a payload module, and the ESS (Extendible Support Structure). Alenia Spazio, EMS Technologies, and AEC-Able were respectively awarded the subcontracts (from MDA) for the spacecraft bus, payload, and ESS. The SAR antenna itself is supported by ESS, which is used to deploy the antenna and to provide a rigid support in the deployed position. Note: As of 2006, the former EMS Technologies Canada Ltd. was bought by MDA and is now referred to as MDA Satellite Subsystems.
The S/C bus is based on PRIMA, a reconfigurable bus developed for the Italian Space Agency (ASI) with bus dimensions of 3.7 m (height) by 1.36 m (diameter). The phased array SAR antenna, built by EMS of Montreal, has 8192 radiating elements fed by 512 T/R modules, two subapertures are provided offering a limited GMTI (Ground Moving Target Indication) capability based on along-track interferometry. The overall SAR antenna size is 15 m by 1.5 m with a mass exceeding 700 kg.
The S/C is three-axis stabilized (zero-momentum satellite). ACS (Attitude Control Subsystem) is using two star trackers for precision pointing (provided by Galileo Avionica S.p.A. of Milan, Italy). Attitude knowledge is ±0.02º, attitude control is ±0.05º (3σ in each axis). The GPS receiver (LAGRANGE of Laben SpA) provides real-time position knowledge of ±60 m. Image location knowledge is < 300 m at downlink, and < 100 m for processed imagery. RADARSAT-2 is yaw-steered (unlike RADARSAT-1). The yaw steering, combined with the improved attitude control of RADARSAT-2, simplifies image processing and improves image quality. Yaw steering provides a measure of Earth rotation compensation and thus brings the data Doppler centroids (DC) near zero Hz. 13) 14)
S/C mass at launch is 2200 kg, power of 2.4 kW at EOL (two solar panels with dimensions: 3.73 m x 1.8 m, each), one Nickel-hydrogen battery with 89 Ah. The S/C design life is seven years. RADARSAT-2 provides several new imaging modes, such as polarimetric imagery (retrieval of full vectorial polarization information), ultra-fine (3 m resolution) beams, in addition to preserving all RADARSAT-1 modes.
RADARSAT-2 uses a monopropellant hydrazine fuel. The S/C has six 1 N thrusters which are used initially to maneuver the spacecraft into its operational orbit (correcting any launch dispersions). The thrusters then maintain the spacecraft's orbit to keep the ground track within a strict tolerance range (better than ± 5km, with a goal of ± 1km) during its operational lifetime.
- Four thrusters of the propulsion system are in the nadir direction (+Z)
- Two thrusters are in the velocity direction (-X)
- They are used alternatively for dragmake-up and COLA (Collision Avoidance) maneuvers.
- The original fuel budget did not include anything for COLA maneuvers
- The fuel budget of October 2016 is enough to for extensive COLA maneuvers and de-orbit services.
Table 1: Comparison of RADARSAT-1 and -2 system capabilities
Launch: A launch of RADARSAT-2 took place on Dec. 14, 2007 on a Soyuz launch vehicle (launch provider: Starsem) from Baikonur, Kazakhstan.
Orbit: Sun-synchronous polar orbit, mean altitude = 798 km, inclination = 98.6º, period = 100.7 min (14 7/34 orbits/day), LTAN (Local Time on Ascending Node) = 18:00 hrs ±15 min (dusk-dawn orbit), repeat cycle = 24 days (343 orbits). RADARSAT-2 is in the same orbit as RADARSAT-1, separated by 30 minutes (and having the same ground track and repeat cycle as RADARSAT-1). The spacecraft orbit control system is capable of maintaining ground-track repeatability to within at least ± 5 km (with a goal of ±1 km), at any point in the orbit. This facilitates proper ground station scheduling. In addition, the tandem flight configuration of RADARSAT-1 and -2 provides a wealth of interferometric applications support.
Orbit determination: The POD (Precise Orbit Determination) software in combination with the onboard GPS receivers permits accurate real time orbit information. The LAGRANGE receiver is a 12-channel dual frequency receiver that produces pseudorange and Doppler phase measurements. The POD software is a GPS based onboard orbital filter, which uses an orbital filter to combine the GPS measurements with high fidelity orbit models to significantly improve onboard orbit knowledge, under the constraints imposed by the bus on the onboard computing resources. Fast-delivery position knowledge is ± 60 m (3σ in each axis), the post-processed position knowledge is ± 15 m (3σ in each axis). 15)
RF communications: Two onboard SSR (Solid-State Recorder) each with a capacity 150 Gbit BOL provide recording of the source data outside of the receiving station range. They can accept data at rates up to 400 Mbit/s. BAC (Block Adaptive Quantization) data compression is used to encode signal data with a selectable wordlength (normally 4 bits I + 4 bits Q).
Agreements between CSA and MDA make MDA the S/C owner and operator. The RADARSAT-2 ground segment is also owned and operated by MDA. This includes the re-use of the existing RADARSAT-1 infrastructure where possible. CSA's investment will be recovered through the supply of imagery to a number of Canadian government agencies during the mission lifetime. RSI (as well as others) are a commercial distributors of imagery.
• June 2018: The satellite continues to operate well without any degradation in image quality, supporting operational imaging with 15 Stripmap modes, 4 ScanSAR modes, and a Spotlight mode (Figure 3), with options for different polarization selections and both left- and right-looking satellite orientation. 16)
- A novel feature of the RADARSAT-2 system is that both the payload and the ground system were designed to sup-port the addition of new modes during ongoing operations. This feature has been well exploited to address evolving government and commercial market needs and opportunities. New modes that have been added include:
a) Spotlight mode, 2008
b) Five wider coverage stripmap modes, 2010
c) Two higher resolution ScanSAR modes for maritime surveillance, 2015
d) Extra wide swath 5 m resolution mode (Extra Fine), 2015.
Figure 3: RADARSAT-2 SAR beam modes (image credit: MDA)
- To illustrate how the usage of RADARSAT-2 has evolved over the last decade the annual coverage for each RADARSAT-2 mode is plotted in Figure 4. Several trends can be identified:
1) ScanSAR Wide mode provides the greatest coverage as it is the workhorse for daily sea ice monitoring. The peak during 2013-2014 was due to RADARSAT-2 supporting ESA needs for SAR data following the loss of ENVISAT and prior to the start of the Sentinel-1 mission.
2) Rollout of the wider coverage stripmap modes in 2010 led to considerable uptake with usage outstripping the original modes.
3) Usage of the new Maritime Surveillance modes ramped up quickly following their rollout with a corresponding decrease in usage of the heritage ScanSAR modes. These new modes were specifically developed for the Polar Epsilon maritime surveillance system and initially they were only available to Government of Canada users but since 2015 they have also been avail-able commercially.
4) The Extra Fine mode (5 m resolution, 125 km swath) responds to the need for wide area monitoring of anthropogenic activity, particularly over the tropics were optical imaging is often obscured by cloud cover. The RADARSAT-2 background mission has shifted from acquiring Multi-Fine mode to Extra-Fine mode over tropical areas. An extensive deep archive of Extra Fine data is now available (Figure 5) and is being exploited for a variety of change detection applications.
Figure 4: Annual coverage of RADARSAT-2 modes between 2008 and October 2017 (image credit: MDA)
Figure 5: Stack depth of the RADARSAT-2 SAR Extra Fine archive (image credit: MDA)
- MDA continues to make the RADARSAT-2 archive more accessible to users. Two recent initiatives include:
a) Leveraging cloud environments for big data analytics on RADARSAT-2 data. The Extra-Fine archive is now available via DigitalGlobe's GBDX (Geospatial Big Data) platform (https://platform.digital-globe.com/gbdx/) along with workflows for applying amplitude change detection. Several projects are exploring how to expand RADAR-SAT-2 exploitation via GBDX.
b) Making the definitive catalogue available online (https://gsiportal.mdacorporation.com). This portal supports search and discovery of RADARSAT-2 data against areas of interest de-fined on a detailed web map interface and complements the existing ordering process (Figure 6). Search results are presented both as quick looks/polygons on the map and tabular format. Results can quickly be filtered by date, mode, polarization and imaging geometry.
• June 2018: Having the world's longest coastline, Canada has a great interest in maintaining and advancing its maritime surveillance capabilities. Launched in 2007, the MDA-built RADARSAT-2 has been providing invaluable means for monitoring Canada's coasts and the surrounding waters, alerting government users on suspicious and polluting vessels. 17)
- Protecting Canada's coastal waters, Exclusive Economic Zone (EEZ) and approaches, including the North is an important part of Canada's national security. This involves monitoring maritime traffic in these areas to detect, identify and track suspicious vessels, an activity co-ordinated by the MSOCs (Marine Security Operations Centers) in Halifax and Esquimalt. MSOC is an inter-departmental agency involving the Department of National Defence (DND), the Royal Canadian Mounted Po-lice, Canada Border Services Agency, the Canadian Coast Guard, the Department of Fisheries and Oceans and Transport Canada, that leverages their joint capability, capacity and authority to enhance Canada's marine security through collaborative detection, assessment, and warning of security threats, incidents and events.
- There are two critical inputs to the RMP (Recognized Maritime Picture) maintained by MSOC; the first is ship detection reports from the DND Polar Epsilon system, which uses wide swath SAR images from the RADARSAT-2 satellite to provide daily surveillance of Canada's coastal waters, EEZ and approaches. Ship detection reports are produced by the OceanSuite software, which detects and characterizes vessels in SAR imagery. The Polar Epsilon system is currently being updated to support expanded use with the RCM (RADARSAT Constellation Mission), which is due to be launched in 2018. The second critical input to the RMP is reports from ships listing their identity, location, seed and heading, which broadcast via the AIS (Automatic Identification System) or the LRIT (Long Range Identification and Tracking) system, which provides identity and location.
- These information sources are complementary. Polar Epsilon provides independent confirmation of ship location along with estimates of ship length, width and heading which supports limited identification. AIS/LRIT provides detailed ship identification and characteristics including size, speed, heading, cargo and destination. Correlation of RADARSAT-2 ship detection and AIS/LRIT self-reporting potentially allow "compliant" emitting vessels to be distinguished from "suspicious" non-emitting vessels, also known as ‘dark ships'. However, there are some challenges. RADARSAT-2 ship detection performance is limited by various factors including ship RCS (Radar Cross Section), which is a function of ship size and design, ocean backscatter, which is a function of sea state, and imaging geometry and radar parameters including resolution and sensitivity. Not all ships broadcast AIS or LRIT and even when they do the information may be incorrect either due to operator error or intentional spoofing.
- Significant effort has been made in the last several years to refine the tools that are used to extract vessel detection reports from SAR imagery, thereby providing more confidence to the information contained in these reports. Additionally, new capabilities for maritime surveillance in challenging environments are being developed.
• December 2017: Launched in December 2007, Canada's Earth observation satellite has now been on orbit for 10 years! During those 10 years, radar imagery from RADARSAT-2 has been used to monitor the impacts of climate change, protect our environment, manage resources and support disaster relief. 18)
- The spacecraft and its payload are active after 10 years on orbit. This project represents a unique collaboration between government and industry. As of October 2017, Maxar Technologies Ltd. [former MDA (MacDonald, Dettwiler and Associates Ltd.)] owns and operates the satellite and ground segment. The Canadian Space Agency (CSA) helped fund the construction and launch of the satellite and recovers this investment through the supply of RADARSAT-2 data to the Government of Canada during the lifetime of the mission. 19)
- RADARSAT-2 demonstrates the Canadian space community's leadership in developing remote sensing technology and applications. With state-of-the-art technology, it is one of the world's most advanced commercially available Earth observation radar image providers and offers users around the world an expanded range of high-quality data products for hundreds of applications.
- A key priority of the Canadian Space Program (CSP) is responding to the twin challenges of monitoring the environment and managing natural resources. The hardy, versatile RADARSAT Earth Observation satellites are a major data source for commercial applications and remote sensing science.
• November 2016: The RADARSAT-2 spacecraft and its payload are operating nominally.
RADARSAT-2 was launched in December 2007 with no formal collision avoidance strategy. The orbit is now in one of the most populated debris zone.
- 2007: The FengYun-1C satellite was destroyed (2600 pieces of trackable debris)
- 2009: Iridium and Cosmos collision (1250 pieces of trackable debris)
- 2015: The US weather satellite, DMSP-13, exploded (147 pieces of trackable debris - and RADARSAT-2 identified as one of ten at-risk satellites).
Figure 7: NASA UNOOSA report 2011 of spatial density in LEO (image credit: NASA)
- First conjunction alert at MDA was received in March 2009 from CSA (Canadian Space Agency)
- Email communication with the JSpOC (Joint Space Operations Center) of the USAF to confirm the orbit data
- Initial effort made (2009) to develop effective procedure for analyzing and assessing the risk.
Table 2: Collision Avoidance Strategy
Table 3: As of October 6, 2016: 11 COLA maneuvers performed for RADARSAT-2 (Ref. 20)
• As of April 24, 2016, RADARSAT-2 is entering its 8th operational service year (RADARSAT-2 was declared operational on April 24, 2008). Numerous enhancements have been added to the original capabilities both on the ground and on the space segments. The operational performance is well within the specification with an acquisition success rate above 98% (acquisition successfully executed vs acquisition loaded on the spacecraft for execution) and a percentage of availability of 99.95% (hours of outage vs total hours in a year).
• Oct. 1, 2015: MDA's Information Systems group has released two new RADARSAT-2 imaging modes for commercial use. These modes are designed specifically for maritime monitoring and will enhance the capabilities of the RADARSAT-2 satellite: 22)
1) Ship Detection mode: A mode that provides superior ship detection capability for ships 25 meters in length or greater, over areas as large as 450 km x 500 km in a single RADARSAT-2 scene. The Ship Detection mode is ideal for applications such as monitoring illegal fishing and sovereignty protection.
2) Ocean Surveillance mode: A mode that provides maritime monitoring capabilities for a variety of maritime applications. The Ocean Surveillance mode can image areas as large as 530 km x 500 km in a single scene. This mode also delivers ship detection capability, but includes monitoring of other ocean features such as detecting oil on water, ice and wind analysis and wake detection.
• August 2015: RADARSAT-2 continues to perform well and system performance targets continue to be met or bettered. There has been no measurable performance degradation during operations. The RADARSAT-2 mission remains the most operationally-focused commercial SAR mission available. 23)
- The most significant anomaly to-date is the SAR antenna CDU (Column Drive Unit) 12 unit failure in 2012. Payload performance is nominal using the CDU unit redundancy and a plan was completed in 2014 to mitigate impacts of any further failures. Other on-board anomalies have been relatively minor. Two heater intermittent failures on the propulsion system resulted in the preparation of a flight software patch which will be implemented in the year 2015 but there is no impact on mission as a temporary work-around solution was implemented in the meantime. - Attitude performance continues to be excellent. All AOCS equipment Main and Redundant are available.
- The new solar array in Sun tracking mode configuration (implemented in 2014) was proven efficient and resulted in a significantly lowered number of battery discharge and charge cycles. The maximum Depth Of Discharge is around 1% outside of eclipse and 8% during eclipse and remains well below expected usage.
- Thermally, some systems exhibited higher than previous temperatures following the change in solar array configuration (in sun tracking). As of April 2015, all temperatures are still well within operating range and monitored regularly.
- The quantity of fuel consumed for drag make up maneuvers is low and the available margin large. There is still more than 120 kg of fuel on board the spacecraft. The average consumption per burn is about 0.011 kg.
- To improve the quality of high-precision SAR products generated from RADARSAT-2 such as interferometry (InSAR), geolocation, and digital elevation mapping, an effort was made by MDA to develop and implement a new orbit determination (OD) tool, termed EDOT (Enhanced Definitive Orbit Tool). Development and testing of EDOT occurred throughout 2014 and by the end of the year EDOT was incorporated into the daily RADARSAT-2 operations at St-Hubert. The definitive orbit files produced by EDOT now replace the previous definitive orbit files distributed by MDA. In addition to generating all future definitive orbit files, EDOT has also been used to reprocess orbit telemetry for the entire mission resulting in a full catalogue of enhanced definitive-orbit files. Therefore newly acquired products and historical products can benefit from enhanced orbit accuracy when ordered after the EDOT rollout. Figure 8 illustrates the EDOT in the RADARSAT-2 context (note that "ODMP" is the current Orbit Determination Software).
Figure 8: EDOT in the RADARSAT-2 Orbit Determination context (image credit: MDA)
Because EDOT uses raw GPS measurements, knowledge of the GPS constellation ephemerides at the time of the measurements is necessary for obtaining orbit solutions. Three types of GPS ephemerides are made available from the IGS (International GNSS Service) data centers, the UltraRapid, Rapid, and Final, which have an average latency of about 4 hours, 17 hours, and 12 days respectively. For EDOT, the Rapid ephemeris is used and as a result there is an increase in the latency of the definitive orbit files generation. EDOT is run once per day shortly after 17:00 UTC and each run generates orbit files for a period of 24 hours. Because of the delay in Rapid GPS ephemeris, EDOT orbit files will have a latency of between one and two days, i.e., the earliest orbit file each run will have a latency of about 48 hours while the latest orbit file will have a latency of about 24 hours.
One important advantage of EDOT is that the satellite attitude is used to further resolve the position of the actual SAR antenna. Neither of the other two OD tools is capable of this. Therefore the reference position on the spacecraft for the EDOT orbit is the center of the SAR antenna. For the current definitive orbit generated by ODMP, the reference was not as well defined, but on average tended towards RADARSAT-2's center of mass, which is approximately 1 m above the center of the SAR antenna. By including the attitude in the orbit determination process, the accuracy in terms of the SAR antenna position increases because the antenna shifts position by up to one meter during attitude slews.
EDOT was run for entire orbit history of mission. As shown in Figure 4-2, estimated uncertainty in OD process is below 1 m.
Figure 9: Relative uncertainty (1σ) estimated for EDOT position (image credit: MDA)
Geolocation testing provides another measure of the orbit accuracy, albeit not a perfect one because geolocation measurements also include many other potential sources of error (e.g. payload timing, processing, ground truth knowledge, atmospheric delays) that are difficult to control. Overall, the RMS absolute geolocation errors observed were between 1 and 2 m depending on the site, while on exact repeat passes the standard deviations of the errors after atmospheric corrections varied from 0.2 to 0.4 m in the along-track direction, and 0.1 to 0.3 m in the slant-range direction, depending on the site and incidence angle (Ref. 23).
• August 2015: The EMOC (Enhanced Management of Orders and Conflicts) process is the means by which all GoC (Government of Canada) RADARSAT-2 data ordering and planning is handled. EMOC is the responsibility of the CSA (Canadian Space Agency) Satellite Operations group, and is performed and managed by the CSA Order Desk – referred to as GRDS (Government RADARSAT Data Services) and includes the establishment of guidelines, procedures, and policies developed in collaboration with other government departments and users. EMOC, as it operates today, is the result of an evolutionary process over many years dating back to 2004 when the concept of a coordinated approach to RADARSAT-1 data ordering was conceived. Originally EMOC was focused on coordinating orders for maritime GoC users, but evolved to encompass all GoC orders for RADARSAT-2 data. 24)
Table 4: Background and history of EMOC
- EMOC process: The EMOC process is performed on a monthly basis and handles RADARSAT-2 requests from GoC authorized users over the Canadian Government AOI (Area of Interest). The main outcome of the process is an acquisition plan, produced every month, of RADARSAT-2 acquisition requests for the GoC, whereby conflicting acquisitions have been eliminated and compatible requests have been merged into shared single requests to maximize the use of the satellite resource. EMOC functions on the principle of transparency whereby each user has visibility into all of the GoC acquisition plans, thereby facilitating collaboration among users over areas of common interest. Five major sub-processes are involved including: 1) the collection of acquisition plan files from GoC clients, 2) the conflict resolution of incompatible overlapping requests, 3) the coordination of compatible requests that can be shared among federal departments, 4) the creation and publication of a coordinated conflict-free monthly government-wide acquisition plan, and 5) the submission of each department's requests to the RADARSAT-2 Order Handling System upon review and acceptance of the conflict-free acquisition plan by the requesting departments.
- For the 2014 calendar year, well over 1000 RADARSAT-2 acquisition requests were handled on a monthly basis within the EMOC process for all of the GoC departments with a significant proportion of these presenting conflicts to be resolved or requiring coordination to convert them into shareable single requests. The number of conflicts was observed to increase during the summer months because of increased demand from the many environmental and agricultural applications. These applications were conflicting mostly with those of the operational marine domain, such as ship detection and pollution monitoring, along the coasts and over the inland lakes. Figure 10 shows the results of the EMOC process for the month of August which had the most conflicts in 2014 with a total of 395. February had the lowest number of conflicts with a total of 202.
Figure 10: Results from the EMOC process for the month of August 2014 (image credit: CSA)
- EMOC is helping the GoC to maximize its use of the RADARSAT-2 satellite resulting in a more sustainable consumption of its data allocation credit while maintaining the continuity of RADARSAT data until the RCM (RADARSAT Constellation Mission) becomes operational. Furthermore, with the loss of RADARSAT-1 in March 2013 increasing the pressure on RADARSAT-2 data ordering, the EMOC process is proving to be an essential mechanism to ensure that more users are served in a fair and transparent manner and that their data requirements are met; thus increasing user satisfaction.
- The EMOC process has also eliminated the need for the use of higher priority programming levels thus avoiding incurring high programming fee costs that impact the GoC RADARSAT-2 data credit. Because the process is transparent and collaborative, there is no longer a need for GoC authorized users to be first-in to submit their request at a very high priority level to guarantee or increase their chances of keeping the desired acquisition against another unknown user.
- Another key cost saving measure that was recently introduced by the CSA Order Desk is the creation of shared single requests over common areas of interests on the Atlantic and Pacific coasts of Canada for maritime domain awareness applications. Before EMOC, ship detection and marine pollution monitoring applications (both of which acquire imagery on multiple common passes daily) would have submitted individual requests into the RADARSAT-2 system without any prior coordination causing a double deduction against the GoC RADARSAT-2 data credit for essentially similar requests. Because of coordination done through EMOC, they now consist of a single request with a dual downlink and delivery to their respective receiving stations for near-real time processing.
- While EMOC has only been fully operational for just over a year, it is already largely considered a success story among its many GoC users who depend on it every month to help them obtain their required RADARSAT-2 data.
• Feb. 2015: On Dec. 14, 2014, RADARSAT-2 completed 7 years on orbit. RADARSAT-2 continues to operate above its mission specified parameters in most areas. RADARSAT-2 was built with reliability, operational capacity, flexibility as key criteria and continues to meet the growing demands of its global user base and is expected to operate at this level for many years. 25)
- The future RCM (RADARSAT Constellation Mission) is the evolution of the RADARSAT Program with the objective of ensuring data continuity, improved operational use of SAR (Synthetic Aperture Radar) and improved system reliability. It is a constellation of three satellites providing around-the-clock coverage. Information obtained from RCM can include repeat imaging of the same area at different times of day, dramatically improving the frequency of monitoring coastal zones, northern territories, Arctic waterways and other areas of strategic and defence interest. RCM will also incorporate automated identification system technology, which when combined with the powerful radar images, supports the immediate detection and identification of ships worldwide. The three-satellite configuration will provide complete coverage of Canada's land and oceans offering an average daily revisit, as well as daily access to 95% of the world to Canadian and International users. The mission development has begun in 2005, with satellite launches planned for 2018.
• July 2014: Now in its seventh year of flight operations, the RADARSAT-2 satellite continues to perform well and system performance targets continue to be met or bettered. The most significant anomaly to date is the SAR antenna CDU (Column Drive Unit) 12 unit failure in 2012. The payload performance is nominal using the CDU unit redundancy and work is in progress to mitigate impacts of any further failures. Other on-board anomalies have been relatively minor. 26)
Figure 11: Average SAR on-time (minutes) per orbit (image credit: MDA)
- The longest outage in the last year, of approximately 12 hours, arose from a battery overvoltage determined as caused by an uncommon race condition between the ABC (Automatic Battery Charge) algorithm BCS (Battery Control System) overvoltage monitor and the last image end-time. Future occurrences have been prevented by a small change to the algorithm.
- Some transient attitude instabilities after roll slews (for left/right imaging) were observed, and investigation revealed that small changes were needed to the on-board attitude control logic to account for normal aging of the STT (Star Tracker), and also that a correction was needed to the slew commanding sequence. The changes have been implemented and nominal performance confirmed.
- A study on battery usage resulted in the decision to remain in sun tracking configuration of the solar arrays for the remainder of the mission. This is the "as designed" configuration, though power margin coupled with SAR on-time previously made it unnecessary, and it is intended to decrease battery charge/discharge cycling and optimize battery life.
- The quantity of fuel consumed for drag make up maneuvers is low and the available margin large. Average consumption per burn is about 0.009 kg. Figure 12 shows the fuel consumption since launch.
Figure 12: Fuel mass evolution since launch (image credit: MDA)
- Almost a decade after much of the RADARSAT-2 specific ground segment development was completed and two decades after completion of the TT&C (Telemetry, Tracking, and Command) stations originally installed for RADARSAT-1, sustaining engineering to maintain availability and ensure maintenance capability over the long-term is a mission priority. High availability is being achieved. Recent ground segment enhancements have focussed on Planning software improvements for more flexible and efficient SAR data downlink planning, SAR processor upgrade, for new and improved image products, and significant ground processed orbit determination accuracy improvement. The RADARSAT-2 SAR data handling capability for reception in Canada has improved with the introduction reception and processing at Inuvik, North West Territories, and the global network of reception stations able to receive RADARSAT-2 data continues to expand.
- The very large operational volumetric capacity of imaging in the RADARSAT-2 system is ultimately constrained by data downlink capacity, and downlink usage can be optimized through flexibility in downlink planning. The RADARSAT-2 planning software has been upgraded with flexibility to allow sharing of alternate downlinks between users.
- The RADARSAT-2 processor at MDA's Data Processing Facility (CDPF), located at the Gatineau Satellite Station in Quebec, continues to be maintained as a "state-of-the-art" SAR data processor, able to generate the latest upgraded products. The CDPF now also remotely controls an MDA processor located at the Aurora Research Institute in Inuvik processing data downlinked to DLR's reception facility. Products may be delivered directly from Inuvik or transferred to Gatineau for value-added processing.
- The RADARSAT-2 global network of X-band SAR data reception stations now extends to 26 sites.
- RADARSAT-2 Routine Phase mission operations comprises the daily order handling, planning, satellite monitoring command and control, SAR data reception, processing, and delivery, and overall system maintenance and management to ensure continuing performance. The mission operations also needs to accommodate special projects and enhancements driven by business needs. Recent special projects have included implementation and trials of the TOPS (Terrain Observation by Progressive Scans) mode for ESA (Sentinel-1), a new GMTI (Ground Moving Target Indicator) trial for DRDC, and NRT (Near Real-Time) value-added production exploiting the Inuvik Facility. Recent enhancements include routine control of the mean satellite ground-track to within 150 m of nominal under most conditions, for interferometric SAR analysis purposes, and automation of TT&C station control activities. Most recently, the new Extra-Fine (XF) mode, 5 m resolution 125 km swath width, has been released for external clients.
- The RADARSAT-2 system continues to evolve with the mission operations team balancing the need to upgrade the ground segment for long-term maintainability and security of existing capability, with the business need for improved features and performance, and the overriding requirement to safeguard long-term health and performance of the satellite.
- In the next year, the enhanced definitive orbits, will be made available routinely for SAR data processing. High incidence angle Ultra-Fine and Spotlight-A is planned to be rolled out. Enhancements to benefit ship detection as part of MDA's strong maritime surveillance product and service offerings will be explored. RADARSAT-2 is also expecting to benefit from the introduction of CCMEO's (Canada Centre for Mapping and Earth Observation) upgraded antenna infrastructure in Canada, with opportunities for combined X-band data reception and S-band TT&C pass operations (Ref. 26).
• June 2014: RADARSAT-2 image quality remains stable and well within operating objectives. Seasonal fluctuations are relatively minor but are monitored carefully and compensated through adaptive processing and calibration adjustments as needed. 27)
- NRT geolocation accuracy: RADARSAT-2 continues to deliver excellent geolocation accuracy in its NRT (Near Real Time) products, thanks to ephemeris data provided in the downlink by on-board POD (Precision Orbit Determination) software. Figure 13 shows the latest results for the subset of modes that are used to monitor calibrated corner reflectors.
Figure 13: NRT ground range locations estimates (left), NRT azimuth location error estimates (right), image credit: MDA
- Radiometric accuracy: RADARSAT-2 radiometric accuracy remains stable and within expectations. Figure 14 shows that measured ratios between actual and reference mean image levels over the Amazon are nearly always within ±1 dB, where the reference levels are based on the Amazon backscatter model. The mean values over the past year were approximately -0.1 dB on ascending passes and +0.15 dB on descending passes. The difference of approximately 0.25 dB is consistent with expectations. The population standard deviation per pass is 0.3 dB.
Figure 14: Mean radiometric error over the Amazon, all modes (image credit: MDA)
Figure 15: Amazon reference backscatter vs incidence angle (image credit: MDA)
- Resolution: RADARSAT-2 ground range and azimuth resolutions remain stable and continue to meet expectations derived from system requirements. Figures 16 and 17 show the latest monitoring results obtained in commercial single beam and Spotlight SAR modes by analysis of returns from antenna dishes and corner reflectors. Measured resolutions have remained essentially constant with time, except for occasional outliers due to imperfections in the measurements such as snow and other external influences, as well as a small known dependence of azimuth resolution on the incidence angle in Spotlight mode. Note that range resolutions are normalized to an incidence angle of 35º.
Figure 16: Normalized ground range resolution (SGX products), image credit: MDA
Figure 17: Azimuth resolution (SGX products), image credit: MDA
- Noise levels: RADARSAT-2 noise levels remain stable and consistent with reported levels. Figure 18 shows that noise measurements obtained in receive-only noise monitoring modes (symbols) are within approximately 1 dB of the calibrated noise level of each pulse (horizontal lines). Noise levels reported in product metadata are also verified explicitly from time to time by comparing them with cross-polarized returns in calm ocean scenes.
Figure 18: Measured (symbols) and calibrated (lines) noise levels for each pulse (image credit: MDA)
- Beam pointing accuracy: Elevation beam pointing knowledge accuracy remains stable with a mean error near zero. Figure 19 shows elevation pointing angle error estimates derived by comparing the modulation of Amazon rainforest data by the elevation beam pattern with the expected modulation derived from knowledge of the beam pattern and the downlinked spacecraft roll angle. The vast majority of error estimates are within ±1 mrad (standard deviation 0.4 mrad). The impact of a 1 mrad error is beam dependent, but translates for example into a radiometric error of ±0.5 dB or less at the edges of the regular Ultra-Fine mode swath at 35º incidence and ±0.5 dB or less between beams in ScanSAR Wide A mode.
Figure 19: Elevation pointing error estimates over the Amazon, all modes (image credit: MDA)
In the azimuth (Doppler) direction, beam pointing knowledge accuracy upon downlink continues to vary seasonally due to various factors, particularly temperature fluctuations in cables carrying RF signal to and from the SAR antenna. Currently the seasonal pattern is going through a period of change, due to increased SAR usage and related increased use of solar array tracking to maintain available power margins. However, in most cases this has no influence on image quality because the Doppler centroid frequency is determined adaptively from the SAR echo data during SAR processing. Figure 20 shows the mean difference between these adaptive Doppler centroid estimates and the Doppler centroid derived from downlinked orbit and attitude alone, for Amazon rainforest data.
Figure 20: Monthly mean and standard deviation of differences between adaptive and orbit & attitude Doppler centroid estimates over the Amazon (image credit: MDA)
- Dual-receive aperture balance: Some RADARSAT-2 modes (including Multi-Look Fine, Ultra-Fine and Spotlight) use dual receive apertures to double the effective pulse repetition frequency without needing to sacrifice swath width. Phase imbalances between fore & aft receive apertures can arise due to differences in component temperatures between wings, among other factors. This is a main reason why the azimuth beam pointing fluctuates seasonally, as discussed in the previous section.
Most differences between receive apertures are compensated adaptively during SAR processing. The only exception in the commercial modes is that phase imbalances can be observed as small residual radiometric ripples in Spotlight images over the Amazon rainforest. These observations are represented by symbols in Figure 21. Imbalances are kept small by adjusting special fore & aft wing phase calibration factors at discrete intervals in H&V receive polarizations. These adjustments, represented by the solid lines in the figure, minimize the impact of inter-wing phase differences on image quality.
Figure 21: Inter-wing phase calibration summary (image credit: MDA)
- Polarimetric accuracy: As discussed above, RADARSAT-2 polarimetric imbalances fluctuate seasonally but the fluctuations remain well within specifications, subject to annual calibration adjustments to correct for longer-term drifts (Ref. 27).
• March 17, 2014: The interferogram of Figure 22 shows the Petermann Glacier grinding towards the sea along the northwestern coast of Greenland. Two RADARSAT-2 TOPS images, acquired 24 days apart, were used to generate it. RADARSAT-2 was programmed specially by MDA to work in an experimental imaging mode called TOPS (Terrain Observation by Progressive Scans) in azimuth to match the way ESA's Sentinel-1 will image Earth. 28)
InSAR (Interferometric SAR) is a technique where two or more satellite radar images acquired over the same area are combined to detect large-scale surface changes. Small changes on the ground cause changes in the radar signal phase and lead to rainbow-colored fringes in the interferogram.
This image shows some stationary and relatively slowly moving features, as well as some large areas of much faster moving ice. The interferometric fringes are widely spaced in the stationary areas and closer together in the center of the glacier where the ice is moving much faster.
Figure 22: Petermann Glacier on the move (image credit: ESA, MDA)
• The RADARSAT-2 spacecraft and its payload are operating nominally in 2014. On Dec. 14, 2013, RADARSAT-2 completed 6 years on orbit.
• December 13, 2013: MDA's Information Systems group announced that it has signed an agreement with the European Space Agency (ESA) to provide RADARSAT-2 information to user groups under ESA's Third Party Mission Program. The RADARSAT information will be used to perform research and develop applications and services such as ice monitoring, pollution monitoring, disaster management, agricultural monitoring and forest management. 29)
• October 21. 2013: MDA announced that it has signed two contracts for the provision of RADARSAT-2 information and services. 30)
- The Norwegian Space Center has extended the contract, signed in January 2003, for MDA to provide RADARSAT-2 information to the Norwegian government for a further three years for use in ice mapping, landslide monitoring, oil spill detection, and ship detection services.
- The second contract is an amendment for MDA to extend its provision of RADARSAT-2 imagery until November 2014 in support of Europe's Copernicus program. The RADARSAT-2 imagery will be used to provide mission critical information for sea ice monitoring of the Baltic Sea, Arctic Ocean, and Antarctic Ocean throughout the ice seasons, improving the safety of maritime navigation and supporting environmental monitoring as part of the Copernicus program.
• May 2013: RADARSAT-2 was reprogrammed by MDA to match the way the Sentinel-1 spacecraft of ESA will be operated. A remarkable achievement is the fact that RADARSAT was able to emulate the way Sentinel-1 images Earth's surface using a method called TOPS (Terrain Observation with Progressive Scan) SAR operations support mode, thus providing a quality of image almost exactly the same as Sentinel-1. 31) 32)
The very first results are promising, as can be seen in the images acquired over Vancouver harbor on the west coast of Canada. The city is clearly visible, as are the ships docked in the harbor, the coastline and the nearby mountain ranges.
The acquisition of more images over specific test sites are planned to demonstrate the suitability of Sentinel-1 for classifying sea-ice, for applications using ocean winds and waves, and for detecting ships, thereby preparing users for the uptake of data.
Figure 23: RADARSAT-2 TOPS image of Vancouver harbor, Canada, released on May 6, 2013 (image credit: MDA, ESA)
Figure 24: RADARSAT-2 TOPS image of Vancouver, Canada, released on May 6, 2013 (image credit: MDA, ESA, Ref. 31)
• The RADARSAT-2 spacecraft and its payload are operating nominally in 2013. 33)
• 2012: DRDC (Defense Research and Development) of Ottawa Canada, and Fraunhofer FHR (Wachtberg, Germany) are developing new operation modes and signal processing methods to enhance and optimize the traffic monitoring capability of the satellite RADARSAT-2.This method enables efficient detection of moving objects and accurate estimation of their parameters and does not require any knowledge of the street network. 34) 35)
• On July 20, 2012, the operator of RADARSAT-2, MDA (MacDonald, Dettwiler and Associates Ltd.) announced that it had signed a contract amendment with CSA and ESA to increase its provision of RADARSAT-2 satellite imagery to Europe's GMES (Global Monitoring for Environment and Security) program. The additional RADARSAT-2 imagery addresses the gap in data availability created by the recent loss of ESA's Envisat spacecraft in April 2012. The agreement fulfills ESA's maritime monitoring needs until the full operational capacity of the Sentinel-1A satellite is available, which is expected around mid-2014. 36)
The RADARSAT-2 imagery will be used to provide mission critical information for sea ice monitoring of the Baltic Sea, Arctic Ocean, and Antarctic Ocean throughout the ice seasons, improving the safety of maritime navigation and supporting environmental monitoring as part of the GMES program. GMES was established to provide users in Europe with access to accurate and timely information services to better manage the environment, understand and mitigate the effects of climate change and ensure civil security.
• The RADARSAT-2 spacecraft and its payload are operating nominally in 2012. RADARSAT-2 has now completed four years of routine phase operations. A thorough mid life review of all spacecraft subsystems has been completed. The spacecraft and ground segment continue to perform well and the operations team strives for best achievable performance. 37)
Some performance items of interest are listed below:
- System outages: A transient event (SEU) caused a malfunction of one of the gyroscope in March 2012. The transient caused an attitude divergence, leading to a Processor Module (PM) restart. Payload and many bus units were switched off resulting in several hours recovery. This anomaly had been observed by the manufacturer on another spacecraft.
- Satellite operations: The frequency of orbit maneuvers to maintain a ±500 m ground track have recently increased from once every 1.5 months to once every 3 weeks due to higher solar activity (more drag). The burn frequency is expected to increase in relation with solar max activity in 2013. Burn activities are conducted without impact on payload activities. Slew maneuvers from right looking to left looking and back are performed in average 150 times a month. - Two collision avoidance maneuvers were performed to prevent risk of collision with other objects. One without user impact, the second one required a few hours outage to perform a retro burn in order to maintain the ground track.
- Alternate downlink: During the peak activity season of summer 2011, bottle necks were identified in the overall downlink time available with the Canadian receiving stations. The current GS systems only allow a user to specify a single downlink location [e.g. Gatineau Receiving Station (GSS)]. The change introduced with the alternate downlink project would let the user specify up to four alternate sites to downlink the data (e.g. Gatineau Station or Prince Albert Station), thus introducing more flexibility for the planning system to find the earliest downlink opportunity and improving the overall downlink capacity usage.
- New and Enhanced Beam Modes: Mode development continues with the next mode in line being, Extra Fine. The mode will offer the best compromise between wide swath (105 to 170 km), and resolution (5 m). New product types offering different resolutions and number of looks from the same mode are being made available. - Work is in progress to extend the coverage offered by Ultrafine and Spotlight from 49.54º to 54.2º, bringing the range resolution down to 1.9 m.
- PDHT timeouts: One of the recurring payload anomalies used to be a timeout issue that occurs at end of imaging and results in a partial lost image if the image is taken in pass through mode (downlink occur while the satellite is still imaging). In depth investigation has been carried out, and changes made to reduce the occurrence of such issue. The result shows a decrease of occurrence by about 40 %. Only a few modes are still occasionally affected and effort to reduce the impact is continuing (Ref. 37).
• August 18, 2011: Imagery acquired from the Canadian satellite RADARSAT-2 has enabled the landmark discoveries announced today by UCI (University of California at Irvine) researchers. - Previously unmapped glaciers of Antarctica have been charted by accessing imagery collected from Canadian, European and Japanese satellites. Using NASA technology, the researchers have discovered unique terrain features that indicate the direction and velocity of ice in Antarctica. This will provide invaluable insight into ice melt and future sea rise due to climate change.
The full continental coverage of Antarctica was made possible due to the unique capabilities of RADARSAT-2 to image left and capture data and information over the central part of the continent. This capability allowed the capture of data over the full land mass, from the South Pole to the coast, imagery that is at the heart of the discovery made by the UCI researchers (the project was funded by NASA). 38) 39) 40) 41)
Figure 25: First map of ice velocity over the entire continent of Antarctica derived from radar interferometric data, (image credit: NASA/JPL, UCI, ESA, CSA, JAXA)
Legend to Figure 25: The map is mainly derived from Envisat, Radarsat-2 and ALOS SAR data, with some contribution of ERS-1/2 and Radarsat-1 data. These new findings are critical to measuring the global sea-level rise resulting from ice flowing into the ocean. The Antarctica ice velocity map is based on SAR data acquired during the International Polar Year (IPY 2007-8) in coordination between ESA, CSA and JAXA. The map, which was created by scientists from the University of California Irvine and NASA's Jet Propulsion Laboratory, reveals not only the flow of the large glaciers, but also their tributaries – effectively rivers of ice – that reach thousands of kilometers inland. 42)
The color-coded satellite data are overlaid on a mosaic of Antarctica created with data from NASA's MODIS (Moderate Resolution Imaging Spectroradiometer) instrument on the Terra spacecraft. The pixel spacing is 300 m. The thick black lines delineate major ice divides. Subglacial lakes in Antarctica's interior are also outlined in black. Thick black lines along the coast indicate ice sheet grounding lines (NASA/Caltech-JPL).
• In Sept. 2011, MDA signed a contract with the US NGA (National Geospatial-Intelligence Agency) to provide SAR data (of RADASAT-2) to be used in creating ice charts and for maritime surveillance to improve the safety of maritime navigation. 43)
• In the summer of 2011, RADARSAT-2 has completed over three years of routine phase operations. The spacecraft and ground segment continue to perform well and the operations team have successfully implemented a number of system enhancements and improvements. 44) 45)
- The system outage history for the last two years of Routine Phase operations is shown in Figure 26. The above 97 % availability requirement has been consistently maintained. Despite the growing demand of RADARSAT-2 SAR data, the success rate of the image acquisitions has been maintained at a high level.
Figure 26: System outages in the period March 2009-March 2011 (image credit: MDA, Telesat)
- The RADARSAT-2 spacecraft has continued to meet performance specification with margin throughout the Routine Phase. There have been no unusual trends in the spacecraft bus subsystem performance. The battery performance in the annual eclipse season has been nominal with battery depth of discharge of less than 8% of the battery capacity.
- During the last year there have been outages caused by recurring anomalies, a few with new signatures, and also some new anomalies with the payload sensor electronics, some of which caused loss of images. Many of these new anomalies are self recovering. These anomalies have been generally attributed to SEU (Single Event Upsets) and managed through monitoring and recovery using pre-prepared and, in some cases, automatic recovery procedures. - As the solar activity rises towards the maximum in the cycle, the rate of SEU occurrence will continue to be monitored, and the management measures will be modified and enhanced, if necessary, to minimize performance impacts.
- The image quality has been maintained or improved in all aspects over the last year of operations. The performance monitoring campaign, to track systematic variations in the system, and to apply periodic updates to various calibration factors, which was established in the beginning of operations, is continuing and has allowed identification of issues and potential performance enhancements, and for adjustments, in addition to the seasonal adjustment to the calibration factors.
- The RADARSAT-2 image quality function is closely coupled with overall system maintenance and the resolution of problems appearing at system level. As examples, the occasional appearance of darker bands in the VV channel of some ScanSAR images has been resolved through payload configuration adjustment. The tracking of intermittent abortion of image data downlinks has highlighted some critical factors in the timing of the acquisitions and has been addressed through beam-modes adjustments. The development of experimental modes triggered a particular SAR payload behavior which caused inconsistent beams in some highly specialized ScanSAR modes. This behavior is now characterized and beam consistency achieved through beam design.
- RADARSAT-2 mission planning has been at the forefront of several ordering and planning system improvements to continuously improve the RADARSAT-2 ordering experience and to optimize the use of the satellite resources. Some of the key improvements made over the last year include:
3) Feasibility checks function now available from the Acquisition Planning Tool (APT).
4) Order desk template functionality to streamline order creation for customer with repeated order parameters.
5) Acquisition and Reception Planning Software (ARPS) upgraded to allow sharing of the same reception segment between two stations with overlapping mask.
6) System upgrades to support the development and operation of HMA (Heterogeneous Mission Access) interface with ESA. The catalog service interface has been completed.
7) Implementation of the ordering capabilities of several new or updated versions of beam modes.
- New network (receiving) stations were added to give a total of 12 foreign partners operating 22 X-band receiving antennas.
- As of March 2011, RADARSAT-2 has now completed 42 orbit maintenance maneuvers. One collision avoidance maneuver was performed in June 2010, followed by a retro grade burn to keep the ground track within its operational target bounds.
- Figure 27 shows an overall growth in data acquisition over the last year, split between government and commercial users, and also the remaining background and image quality/calibration acquisitions. On average over 4,200 acquisition segments have been programmed per month over the last year.
Figure 27: Minutes of acquisition time used by major user group (image credit: (image credit: MDA, Telesat)
- Satellite operations: While orbital collision risk is still assessed to be low, the Iridium/COSMOS incident in 2009 has increased the risk and made LEO satellite operators more aware of the need to adopt mitigation strategies. RADARSAT-2 satellite operations continues to pay increased attention to orbital collision avoidance and the incorporation of collision avoidance procedures into the operations.
Naturally, the RADARSAT-2 project is taking advantage of information and services available from national and international agencies. The inputs feed the collision avoidance procedures, which include quantitative criteria for avoidance maneuver planning. RADARSAT-2 is working with CSA, and the RADARSAT-1 mission, to refine the procedures and tools for risk assessment and avoidance maneuver decision.
- Antarctic mosaic: As a successor to the RADARSAT-1 Antarctic Mapping Mission (AMM) in 1997, and as part of the International Polar Year, the continent was imaged to provide Pole to coast dual polarization mosaic coverage and 3 cycles of interferometric coverage for ice velocity mapping of the interior. RADARSAT-2's agility to slew to "left-looking" to avoid the coverage gap beyond 78º S in the normal "right-looking" orientation was used extensively with 620 slews for the mosaic campaign and 570 slews for the interferometric coverage.
Satellite slewing between the imaging orientations involves a 10 minutes imaging outage for the maneuver and stabilization. So as to minimize the impacts on other users, the satellite slew constraints and the planning were reviewed and modified. Better than specification in-orbit performance permitted the satellite slew constraints (number per orbit) to be relaxed. Planning for command uplinks and slew timing were arranged to allow for the greater command volume and to minimize conflicts.
Figure 28: RADARSAT-2 Mosaic of Antarctica (image credit: MDA. Telesat)
- Enhancements for Project Polar Epsilon: Under the Canadian Department of National Defence (DND) Polar Epsilon NRT (Near Real Time) ship detection project, several systems improvements have been made, particularly to order handling and mission planning. These improvements allow for the introduction of the two new Polar Epsilon reception stations, and facilitate downlink sharing between government reception stations. The Polar Epsilon changes have now all been rolled out into operations.
- Ordering Interfaces for HMA Project: Under the HMA project contract with ESA, MDA has been developing an infrastructure to support a standardized way of querying the catalog, reporting on data acquisition, and ordering data. In this context, a machine-to-machine interface for fulfilling these functions is being developed. Web services are being created from more traditional applications, such as a web application in the case of the OD (Order Desk) system, or a standalone tool in the case of the APT (Acquisition Planning Tool).
An APT web server has already been developed which currently supports a catalog query service as well as receiving acquisition coverage from the catalog. - An Order Desk web server is being developed with new web services such as creating an order, submitting an order, or obtaining status on an order element (Ref. 44).
• The RADARSAT-2 spacecraft and its payload are operating nominally in 2011.
- All impulse response measures have remained consistently with specification
- Radiometry has remained stable since start of Operational Phase. No overall system calibration adjustments have been made since the last CEOS meeting.
- Periodic polarimetric calibration adjustments keep the accuracy well within original goals
- The geolocation accuracy is currently assessed as < 10 m for all single beam modes.
- Additional Quad-Pol beams have been made available operationally. There are now 31 swath positions available, spanning 20º - 49º incidence.
RADARSAT-2 is providing new and better opportunities for spaceborne monitoring of vessel traffic and fishing activities because of better flexibility in the choice of resolution, polarization and look direction. Due to user demand, five new modes were installed on the satellite in 2011. The new modes give the possibility to order images with better resolution and quad-polarisation (for some of the modes). This may be better for some scenarios in operational ship detection due to the increased swath width and coverage area for the new modes. Prior to the installation of the new modes, high resolution and quad-polarisation images only covered small areas. For ship detection these images have mainly been used for research purposes, harbor areas or over a known small area of interest.
The RADARSAT-2 SAR sensor is very flexible. It is possible to reprogram the sensor for example according to resolution and swath width. Five new modes have now been made due to user demand. The new modes open up new opportunities to use images with good resolution and relatively wide swath width at the same time. The Wide Fine Quad-Pol and Wide Standard Quad-Pol modes give the opportunity to choose quad-pol images with wider swath widths compared to the original modes.
Table 5: New additional beam modes of RADARSAT-2
Table 6: Description of the new beam modes (Ref. 48)
• The RADARSAT-2 spacecraft and its payload are operating nominally in 2010. Commercial spacecraft operations are conducted by Telesat (Ottawa) for CSA/MDA (the mission comprises a large system and a number of entities, organizations, and stakeholders inside and external to MDA.). 51) 52)
• In late April of 2010, RADARSAT-2 has successfully completed its second year of routine phase operations. The mission operations team has continued to operate and maintain the system to meet and exceed performance specifications. In particular, the system outage trend has decreased. In the last year orbital collision risk mitigation measures have become more important and RADARSAT-2's first avoidance maneuver was successfully executed as a precaution. The team continues to implement system and operations improvements, including new and improved beam modes. A significant upgrade to support the DND Polar Epsilon project is underway. The system will continue to evolve to respond to customer and market needs. 53)
- In parallel with the commercial operations, RADARSAT-2 supported a series of MODEX (Moving Object Detection Experiments) conducted by Defence R&D Canada (DRDC), involving special SAR antenna configurations and beam/modes. The initial trials work is now largely complete.
- The RADARSAT-2 spacecraft has continued to meet performance specification with margin throughout the routine phase. There have been no unusual trends in spacecraft thermal or power performance. Battery performance in the annual eclipse season has been nominal with battery depth of discharge of less than 10% of the battery capacity (Ref. 53).
• In January 2010, MDA provided free imagery acquired by the RADARSAT-2 satellite over Haiti in support of disaster recovery and reconstruction efforts by the Canadian Government and relief agencies through the Canadian Space Agency's participation in the International Charter "Space and Major Disasters". 54)
• RADARSAT-2 is a contributing mission to GMES (Global Monitoring for Environment and Security) of the EU. In early 2010, MDA is completing a project to implement the HMA (Heterogeneous Mission Access) catalog interface capability -- to accomplish a coherent access to archives for the support of scientific exploitation. 55)
• In May of 2008, RADARSAT-2 polarimetric SAR test data for Natural Resources Canada became available.
• RADARSAT-2 was declared operational on April 24, 2008 when the commissioning phase ended with the Commissioning Review. - Since that time, additional modes such as Spotlight and MODEX (i.e. GMTI) have been introduced, and work has been undertaken to optimize performance of all modes beyond the original specifications. 58) 59) 60) 61) 62)
• Left or right looking: Left- and right-looking modes are available on RADARSAT-2. This permits routine Antarctica mapping and, in emergency situations, the choice of beam mode and position can be set to ensure the greatest repeat coverage of the region of interest.
• In late March 2008, all the pre-defined modes of RADARSAT-2 had demonstrated performance that met requirements in all aspects of image quality. Initial radiometric, polarimetric and geolocation calibration had been completed for all modes.
• The very first images of RADARSAT-2 were acquired on Dec. 18, 2007 just four days after launch.
Figure 29: The first Quad Pol image of RADARSAT-2 on Dec. 18, 2007 showing the region of the Sermilik Fjord in Greenland (image credit: MDA)
Figure 30: RADARSAT-2 Quad Pol image of Devon Island in the Canadian Arctic Archipelago (image credit: MDA)
The payload module consists of the SAR instrument, the SAR antenna, and specific support equipment required to perform such functions as timing and control of the payload, signal distribution, signal detection, thermal control, data storage, and X-band downlink.
SAR (Synthetic Aperture Radar) instrument:
• Operationally, the SAR instrument is identical to that of RADARSAT-1 (with regard to mission control, mission planning, and data collection and processing)
• Fully integrated into the RADARSAT-1 ground segment
• Supports tandem mission operations with RADARSAT-1
• All RADARSAT-1 image quality specifications are met or exceeded
• A 3 m ultra-fine resolution is provided (improved object detection)
• A selectable incidence angle and the choice of left- and right-looking imaging capability is provided with the ability to maneuver the S/C on command. This feature offers much quicker revisit times for handling user requests. It also doubles FOR (Field of Regard) for event monitoring support applications.
• Provision of fully polarimetric imaging modes. Polarimetry modes (amplitude and phase information) increase the information content of the measurement.
• An increased downlink power permits the use of 3 m antenna dishes for ground receiving stations
• The encrypted downlink science data stream maintains service confidentiality
• The utilization of onboard GPS receivers offers real-time position knowledge within 60 m
• The functional exploitation of solid-state recording technology permits the provision of new customer services (selection of image scenes and receiving station)
The most fundamental change in the instrument design from RADARSAT-1 is the introduction of an active phased array antenna on RADARSAT-2 providing two-dimensional beamforming and beamsteering to produce the many (over 200) beams required for the various imaging modes. This replaces the passive slotted waveguide antenna with one-dimensional elevation beamforming of the earlier S/C. The full set of modes is given in Table 8, and they are illustrated in Figure 31. All modes are available in both left- and right-looking satellite orientations. 64) 65)
The AIU (Antenna Interface Unit) and 16 CDU (Column Drive Units), one on each column, perform the command & control and the power distribution functions of the antenna.
Table 7: SAR instrument parameters
Table 8: SAR imaging modes of RADARSAT-2
SAR instrument calibration: The design of the RADARSAT-2 payload has made it possible to operate in fully polarimetric quad-polarization (quad-pol) modes, which introduce the requirement for polarimetric calibration on top of the radiometric calibration of all other modes. In addition to these two primary forms of calibration, there are also a number of other forms of image measurement and payload characterization work which must be undertaken in order to ensure that the required image quality is achieved and maintained. An IQS (Image Quality Subsystem) has been established within the RADARSAT-2 ground segment to perform these tasks, and the key tools required for analysis of data and generation of calibration information have been built into a suite of software provided in several IQS workstations. 66) 67)
Figure 31: Some observation geometries and coverage modes of RADARSAT-2 (image credit: MDA)
SAR instrument overview:
The SAR payload consists of the SAR Antenna and the Sensor Electronics (SE) as shown in Figure 34. The Antenna includes a planar distributed phased array antenna having 512 subarrays each interfaced to one solid state transmit / receive (T/R) module. The long, or horizontal, dimension (15 m) of the array is directed along the direction of flight. The short, or vertical, dimension (1.4 m) is oriented such that the normal to the physical antenna aperture is inclined towards the Earth's surface at an angle of 29.8º with respect to the vertical between the spacecraft and the ground. 68) 69)
The subarrays are each provided with dual ports to enable transmission and reception in horizontal polarization (HP) mode (electric field vector aligned with long dimension of the array) and vertical polarization (VP) mode (electric field vector aligned with the short dimension of the array). The 512 T/R modules and their associated radiating subarrays are arranged in 16 columns uniformly spaced over the length of the array. Each column comprises 32 T/R modules and subarrays uniformly spaced over the height of the array.
Low level transmit signal generation and on-board receive signal handling is performed by the SE (Sensor Electronics) subsystem. Low power transmit signal pulses at C-band are generated by a waveform generator within the SE and distributed to each T/R module via a PAA (Power Amplifier Assembly) and a corporate signal distribution network comprising APDN (Azimuth Power Dividing Networks) and EPDN (Elevation Power Dividing Networks). Radar echo pulses are amplified by the T/R modules and collected by two independent sets of corporate combining networks, one for each polarization. The receive path combining networks have the same physical and electrical properties as the transmitting network, but operate in the reverse direction.
The output signals from each of the combining networks are routed via the SM (Switch Matrix) to the two receiver channel inputs of the SE. The SM controls the routing of received signals from the antenna wings to the SE input ports as required by the current imaging mode, and allows a low power reference sample of the transmitted pulse to be inputted to the SE during the transmit pulse time.
Figure 32: Line drawing of the RADARSAT-2 spacecraft (image credit: MDA)
The antenna radiating elements, T/R Modules, CDUs and associated power, control and RF signal interconnecting harnesses and distribution networks are mounted on the SAPA (SAR Antenna Panel Assembly) as shown in Figure
Figure 33: The SAPA layout (image credit: MDA)
Figure 34: Overview of the SAR instrument architecture (image credit: MDA)
T/R module architecture (Figure 35): The T/R module provides the transmit RF power amplification and receive signal amplification for the antenna array. A T/R module includes two transmit power amplification channels and two receive channels. The two receive channels may be operated simultaneously for reception of V and H polarized signals. One transmit channel can be selected depending on the setting of the RF switches. If required, the transmit operation can alternate between H and V on alternate pulses.
A nominal peak pulse power of +6 dBm is applied to the 50 ohm transmit input port of each T/R module. The center frequency is 5.405 GHz, and maximum bandwidth, dependent on operating mode, is 100 MHz. The T/R module sends the Transmit RF pulse by the HP or VP port to the subarray. Nominal peak pulse output power levels are +40 dBm (10 watts) in high power mode and +38 dBm (6.3 W) in low power mode, into 50 ohm. Only one transmitter pulse output can be active at any one time. The same ports provide HP and VP 50 ohm RF receive interfaces with the subarray.
A control board, containing digital ASIC and discrete devices, controls the operation of T/R module RF functions. Two identical ASIC's, one dedicated to each RF channel, provide serial interface control, modulator drive, timing and phase and amplitude data memory functions for their associated RF channel.
Figure 35: Block diagram of the T/R module configuration (image credit: MDA)
MODEX (Moving Object Detection Experiment):
MODEX represents an experimental GMTI (Ground Moving Target Indication) capability of the SAR system on RADARSAT-2. During MODEX operation, the SAR antenna is partitioned into two subapertures along the satellite track to sequentially observe the same scene from the same spatial point. The technique allows a wide variety of operating modes and parameters to be tried. In one version of the experiment, similar to ultra-fine mode, MODEX makes use of the dual-receive capability of the RADARSAT-2 antenna. This dual-receive capability provides two apertures aligned in the along-track direction, which is suitable for detecting moving objects. By processing the received echo data using along-track interferometric techniques and DPCA (Displaced Phase Center Antenna) techniques, objects with non-zero radial velocities can be detected and their radial velocities can be estimated. 70) 71) 72)
MODEX has been implemented through collaboration between DND, the Canadian Space Agency (CSA), and the satellite builder, MDA Corporation, Ltd. Three GMTI detection algorithms were implemented: 73)
• The first one is the classical DPCA (Displaced Phase Center Antenna), which is based on using two separate signal channels that observe the same scene at different times from the same point in space. The presence of moving targets is detected in the differential data by comparing the magnitude of the signal difference against a computed CFAR (Constant False Alarm Rate) threshold.
• The second detector is a new algorithm called the Hyperbolic detector, which is based on the combination of two independent metrics. The first metric is used to detect targets with a high probability of detection while allowing a rather large number of false alarms. A second metric is then applied to only the targets which exceeded the detection threshold of the first metric to reduce the false alarms while preserving the high probability of detection. The two independent metrics are combined into one two-dimensional metric. The resulting detection threshold from the combined metric is a hyperbolic curve instead of a classic CFAR threshold value.
• The third detection algorithm is called the HATI (Histogram Along-Track Interferometry) detector. This algorithm differs from the other detectors in that it uses an adaptive nonparametric CFAR detection scheme that does not involve the theoretical modeling of clutter statistics. Although the HATI algorithm uses ATI for moving target detection, the adaptive histogram techniques to calculate CFAR thresholds can also be applied to other detectors such as DPCA.
Figure 36: MODEX-1 processor architecture of the RADARSAT-2 mission (image credit: DRDC)
In addition to the dual-receive mode of operation, RADARSAT-2 also supports an alternating-transmit mode where pulses are transmitted alternately from each wing, and received alternately on each wing. This toggle mode allows greater separation of the two-way phase centers in the along-track direction. The transmitter toggling approach (between fore and aft sub-apertures) has the advantage of maintaining the same phase-center distance as the dual-channel case, which is 3.75 m for RADARSAT-2, and is capable of generating three phase centers.
Six sets of MODEX data were collected during RADARSAT-2 commissioning trials: three in the MODEX-1 mode and three in the MODEX-2 mode (or the toggle mode). Preliminary MODEX results showed that the sensor is capable of detecting and measuring both cross-track and along-track velocity components as low as 5 m/s. The accuracy of the measurements is about ±0.3 m/s in the cross-track direction and ±4.0 m/s in the along-track direction (Ref. 73).
Note: MODEX-1 represents the MODEX processor implementation at launch which includes a preprocessor to prepare the data for GMTI processing as well as two-antenna techniques to detect and estimate moving target parameters. MODEX-2 is a processor upgrade which includes toggle mode data processing and waveform diversity processing as well as three-antenna detectors and estimation methods.
The MODEX experiment is being funded by DND (Department of National Defense) of Canada for technology demonstration purposes. DND aims to develop a SAR-GMTI processing system to investigate the military and commercial utility of space-based moving target measurements. MODEX is being carried out by DRDC (Defense Research and Development Canada).
RADARSAT-2 data: RSI (RADARSAT International), a subsidiary of MDA, will market and distribute RADARSAT-2 data on a commercial basis. 74)
In addition, Canada is going to use the RADARSAT-2 data in the following programs:
• SOAR (Science and Operational Applications Research for RADARSAT-2), a joint partnership program between the Canadian government and MDA through CSA and CCRS (Canada Center of Remote Sensing) of NRC (Natural Resource Canada).
- The SOAR program provides access to RADARSAT-2 data only for research and testing purposes.
- The SOAR program provides an opportunity to explore the enhanced capabilities of RADARSAT-2 and their potential contributions to applications through a loan of limited amounts of RADARSAT-2 data to research projects.
Now that the satellite is fully operational, the Government of Canada will develop specific R&D initiatives under the SOAR program umbrella.
• Project Polar Epsilon: A surveillance and ship detection initiative within Canada's Defence Program. RADARSAT-2 wide-area data will be used to enable all-weather, day/night persistent surveillance of Canada's Arctic region and ocean approaches (Canada has the longest coastline in the world at 243,772 km and a corresponding marine area of responsibility of over 11 million km2). The goal in marine security is domain awareness: to know what is happening and where it is happening in the marine domain or ocean approaches to the borders.
Project Polar Epsilon invests in applications and ground segment infrastructure to receive and process RADARSAT 2 information. Polar Epsilon will deliver four main capabilities: near real-time ship detection; arctic land surveillance; environmental sensing; and maritime surveillance radar beam optimization. The near real-time ship detection capability will include local RADARSAT-2 satellite reception, processing and applications in support of the emerging Marine Security Operations Centers (MSOCs) on both of Canada's east and west coasts at or near Halifax and Esquimalt. 75) 76) 77) 78)
The RADARSAT-2 imagery analysis by the MSOCs will be complemented with optical imagery of the MODIS (Moderate-Resolution Imaging Spectroradiometer) instrument on various US missions. The provision of ocean color information from the MODIS sensors will assist the MSOCs with operational use of maritime patrol aircraft, ships, submarines and sonar performance prediction.
The RADARSAT-2 surveillance must be seen as complementary to future AIS (Automated Identification System) services onboard ships, mandated by the IMO (International Maritime Organization). AIS is a mixed ship and shore-based broadcast transponder system, operating in the VHF maritime band, which sends ship identification, position, heading, ship length, beam, type, draught and hazardous cargo information, to other ships as well as to shore. Of significance is that AIS can be monitored from satellites.
An AIS device will not be part of RADARSAT-2. RADARSAT-2 will only perform the ship detection function with its SAR imagery. However, Canada is planning to fuse commercial space-based AIS with coincidental RADARSAT-2 passes - this will help to confirm non-compliant AIS reports or no reports.
The follow-up system of RADARSAT-2, a three-spacecraft mission referred to as RCM (RADARSAT Constellation Mission), with launches in the timeframe 2012-14, is planning to combine SAR and AIS on the same spacecraft platform.
The ground segment architecture is broken down into several subsystems as shown in Figure 37 (modular distributed system). Where possible, the existing RADARSAT-1 ground segment facilities and infrastructure are being reused. However, almost all software and computer hardware has been replaced by new systems. The use of up to date technology allows lower maintenance and operational cost while providing improved functionality. 79) 80) 81)
The RADARSAT-2 ground segment was upgraded from the RADARSAT-1 infrastructure and includes a Mission Control Facility at the CSA Headquarters at St. Hubert (near Montréal), Québec, and TT&C stations at St. Hubert and at Saskatoon, Saskatchewan. The CSA order desk will coordinate all government user requests for RADARSAT-2 data. MDA is responsible for the sale and distribution of RADARSAT-2 data to all commercial users. In May 2006, an MDA MOC (Mission Operations Center) was installed in St. Hubert.
Figure 37: Overview of RADARSAT-2 ground segment architecture (image credit: MDA)
Figure 38: Ground segment facility locations (image credit: MDA)
The RADARSAT-2 mission components are shown in Figure 39. In addition to internal and external system components and mission operations, the mission involves the data users, MDA business management, and the regulatory authority. The Canadian Government is a major mission stakeholder and receives data in return for its investment in the system development. RADARSAT-2 mission management concerns management of and/or interfacing with these components to ensure orderly conduct of the mission to meet mission objectives (Ref. 51).
RADARSAT-2 operations are characterized by intensive, highly automated, intermittent TT&C pass operations, supported by complex planning cycles to convert client orders into imaging and data downlink plans and to prepare on-board command schedules and upload command sequences. A multi-mission approach for operations provides cost and risk savings for participating missions.
Figure 39: RADARSAT-2 mission components (image credit: MDA)
RADARSAT-2 routine phase mission operations are functionally arranged as shown in Figure 40. These provide for the day-to-day end-to-end system operations and maintenance of the satellite and ground segment, including overall system and operations management, planning of satellite and ground reception activities in response to client orders, satellite command and control, and Canadian SAR data reception, archiving, cataloguing, processing and distribution. Operations development established plans for these functional areas.
Figure 40: RADARSAT-2 mission operations functions (image credit: MDA)
For a commercial mission, operations costs are an important metric. Early during development a cost model was established. This was updated throughout development so that cost metrics could be reviewed at design milestones and for comparison of operations approaches. MDA senior management took, and continue to take, a particular interest in reviewing the status and planning for mission operations costs in view of their impact on RADARSAT-2 business success. This led to frequent corporate reviews during development and to the setting of cost budget targets and goals which drove planning and implementation decisions (Ref. 51).
Regulatory environment: During RADARSAT-2 development a new regulatory environment was introduced in Canada governing commercial satellite remote sensing operations – the "Remote Sensing Space Systems Act". This has had the effect that key mission and system requirements and plans, such as operational orbit parameters and control, performance of products generated from the SAR Payload, and the need to maintain and execute an end-of-life decommissioning plan, have become legal as well as mission responsibilities. An additional system activity reporting burden has also been imposed at mission level.
A system impact of this new regime concerned the changes needed to order handling to automate the otherwise operator intensive and potentially error-prone new "access control" rules for imagery imposed on clients by the regulations. A further mission management impact concerns the need for formal permission to enhance system capability in areas which improve licensed performance (Ref. 51).
In the spring of 2010, RADARSAT-2 is starting in its third year in orbit and in its third year routine phase operations. Despite the system complexity, commissioning was completed relatively smoothly. System and operations performance in the routine phase has met and surpassed system requirements. A maintenance policy was adopted early in development and its implementation has proved to be successful. The mission operations processes and system were adapted in response to a new regulatory environment introduced shortly before launch. A mission operations risk management process has been adopted and includes orbital collision avoidance measures. A number of operations and system enhancements have been implemented in parallel with continuing operations. The system and operations will continue to evolve to meet business needs (Ref. 51).
Polar Project Epsilon (first attempts):
Leveraging the programmability of the RADARSAT-2 sensor, MDA developed and assessed two new ScanSAR beam modes: a 450 km wide ship detection optimized beam mode in the HH polarization channel, and a 530 km wide multi-purpose beam mode in the HH and HV polarization channels. These new beam modes are significantly better, offering nearly uniform ship detection of much smaller vessels across the full swath width, when compared to the existing RADARSAT-2 ScanSAR beam modes. 82)
Approach: To aid in the development of improved SAR images, MDA first constructed a tool which takes as input the relevant characteristics of a proposed SAR beam mode and outputs a statistically probable Minimum Detectable Ship Length (MDSL) metric at a 90% confidence level. This tool makes use of an empirical relationship between ship length and RCS (Radar Cross Section).
The tool makes the simplifying assumption that the entire RCS of the vessel is contained within a single resolution cell. This is valid since our objective is to assess the detectability of small ships which do satisfy this assumption. The tool is likely to underestimate the length of larger ships in the case where the RCS is spread over several resolution cells. Due to the complexity associated with ship orientation and length-to-width ratio, no attempt has been made to model larger ships to a higher fidelity, and therefore caution must be used when interpreting the absolute value of the MDSL metric reported.
The tool simulates, in a statistical sense, vessel detection being performed by a Constant False Alarm Rate (CFAR) filter using a K distribution. Here, the simulation is set to allow a maximum of 2 false alarms per 3600 km along-track. While a plethora of notable ship detection algorithms are currently available, this simplified tool allows to make useful relative comparisons between the existing RADARSAT-2 beam modes and these new beam modes to guide the beam mode design.
Recognizing that maritime domain awareness applications are not solely about vessel detection, the project developed two classes of beam modes: those which are optimized for vessel detection performance at the expense of other maritime applications, and multi-purpose beam modes which provide improved vessel detection performance, relative to the traditional beam modes, while continuing to support other maritime applications.
• The vessel detection optimized beam mode is 450 km wide, comprised of seven sub-beams (in the HH polarization channel only) spanning incidence angles ranging from 34º to 57º, and which makes use of the "dual-receive aperture" capability of the RADARSAT-2 sensor. The high data rate associated with the 50 MHz pulse, the high PRFs (Pulse Repetition Frequencies) and the dual-receive necessitated compressing the data using 1-bit block adaptive quantization (BAQ) encoding ito achieve real-time imaging and downlinking.
• The multi-purpose beam mode is 530km wide, comprised of eight sub-beams (in the HH and HV polarization channels) spanning incidence angles ranging from 20º to 50º. The high PRFs and the applications-driven need for low noise and dual-polarization with a minimum of 2-bit BAQ encoding made it necessary to use the 17 MHz pulse in order to achieve real-time imaging and downlinking.
Figures 41 and 42 illustrate the 90% confidence level MDSL metric versus incidence angle for the case of wind at 8 m/s at an angle of 45º relative to the beam. These plots compare the ScanSAR Wide beam mode (red lines) with the new ship detection optimized and multi-purpose beam modes (blue lines). Since both of these new beam modes have only one look in the azimuth direction, there will be a two-dimensional variation in the SAR sensor noise, leading to a corresponding variation in the MDSL metric (each of the blue lines represents a different position in azimuth within a burst) that has been described as an "egg carton."
Figure 41: Relative comparison of the MDSL metric for the 500 km wide ScanSAR wide (HH polarization) and the 450 km wide vessel detection optimized beam modes (image credit: MDA)
Figure 42: Relative comparison of the MDSL metric for the 500 km wide ScanSAR wide (HH polarization) and the 530 km wide multi-purpose beam modes (image credit: MDA)
Table 9: Relative comparison of the MDSL metric for the new and existing ScanSAR beam modes
The two newly designed ship-detection application-specific ScanSAR beam modes have demonstrated that both beam modes offer marked improvement over the existing ScanSAR beam modes in terms of ship detection performance. While more detailed analysis remains to be carried out, the multi-purpose beam mode has been shown to be capable of supporting other ocean surveillance applications such as ice analysis, oil and pollution monitoring, ocean wave analysis, and wind retrievals.
1) D. De Lisle, L. Brûlé, M. Reza Sahebi, B. Hawkins, K. Murnaghan, W. Branson, G. Staples, "Pre-Launch Activities in Preparation for RADARSAT-2," Proceedings of EUSAR 2006, Dresden, Germany, May 16-18, 2006
2) L. Brûlé, D. De Lisle, H. Baeggli, J. Graham, "RADARSAT-2 Program Update," Proceedings of IGARSS 2005, Seoul, Korea, July 25-29, 2005
3) A. Hillman, D. Comi, W. Branson, P. Rolland, "Countdown for RADARSAT-2 System Operations," Proceedings of IGARSS 2005, Seoul, Korea, July 25-29, 2005
5) P. Fox, "The RADARSAT-II Mission," Proceedings of IGARSS'99, Hamburg, Vol. III, June28 - July 2, 1999, pp. 1500-1502
6) H. Baeggli, L. Brûlé, J. Smyth, "RADARSAT-2 Program Update," Proceedings of IAC 2004, Vancouver, Canada, Oct. 4-8, 2004, IAC-04-B.2.01
7) L. C. Morena, K. V. James, J. Beck, "An introduction to the RADARSAT-2 mission", Canadian Journal of Remote Sensing, Volume30, No. 3, June 2004, pp. 221-234
8) L. Brûlé, H. Baeggli, "RADARSAT-2 Program Update," Proceedings of IGARSS 2002, Toronto, Canada, June 24-28, 2002
9) A. A. Thompson, D. Racine, A. P. Luscombe, "RADARSAT-2 Antenna Calibration Using Ground Receivers/Transmitters," Proceedings of IGARSS 2002, Toronto, Canada, June 24-28, 2002
10) M. Farhat, F. Lauzon, A. Trudeau R. Fiset, "InSAR End-to-End Simulation Environment," Proceedings of IGARSS 2002, Toronto, Canada, June 24-28, 2002
11) P. Fox, A. P. Luscombe, Z. Ali, "The Radarsat-2 Mission, New Modes and Techniques," Proceedings of IAC (International Astronautical Congress), Houston, Tx, Oct. 10-19, 2002, IAC-02-B.3.07
12) Daniel De Lisle, "Capabilities of RADARSAT-2 and use by the Canadian Government," Proceedings of the 4th International POLinSAR 2009 Workshop, Jan. 26-30, 2009, ESA/ESRIN, Frascati, Italy, URL: http://earth.esa.int/workshops/polinsar2009/participants/122/pres_6_DeLisle_122.pdf
13) Pierre D. Beaulne, Charles E. Livingstone, "Evaluation of RADARSAT-2 Yaw Steering for SMTI Applications," Proceedings of EUSAR 2010, 8th European Conference on Synthetic Aperture Radar, June 7-10, 2010, Aachen, Germany
14) C.E. Livingstone, I. Sikaneta, C. Gierull, S. Chiu, P. Beaulne, "RADARSAT-2 System and Mode Description," Proceedings of RTO-MP-SCI-150, paper 15, Neuilly-sur-Seine, France, 2006, URL: http://www.dtic.mil/dtic/tr/fulltext/u2/a469927.pdf
15) A. Ray, G. Goldman, A. Albanese, "Precise Onboard Orbit Determination for RADARSAT-2," NAVITEC 2001, 1st Workshop on Satellite Navigation User Equipment Technologies, Dec. 10-12, 2001, ESA/ESTEC, Noordwijk, The Netherlands
16) Ron Caves, "Evolution of RADARSAT-2 Usage Over a Decade of Operation," EUSAR 2018 (12th European Conference on Synthetic Aperture Radar), Aachen, Germany, June 4-7, 2018
17) Evgeniy Lebed, Ron Caves, "Recent and Ongoing Developments in using RADARSAT-2 for Maritime Surveillance in Canada," EUSAR 2018 (12th European Conference on Synthetic Aperture Radar), Aachen, Germany, June 4-7, 2018
18) "December 14, 2017 – 10-year anniversary of RADARSAT-2 — '2017 in review – 20 highlights for space, science and the CSA'," CSA, URL: http://www.asc-csa.gc.ca/eng/blog/2017/12/22/2017
19) "RADARSAT-2," CSA, 14, Dec. 2017, URL: http://www.asc-csa.gc.ca/eng/satellites/radarsat2/default.asp
20) Casey Lambert, Camille Decoust, Bryan Cooke, " RADARSAT-2 and RCM Conjunction Analysis and Mitigation Operations," SSA Operators' Workshop, Denver, Colorado, Nov. 3-5, 2016, URL: https://advancedssa.com/assets/img/workshop/presentations
21) Bryan Cooke, Casey Lambert, "RADARSAT-2 Conjunction Tales," SSA Operators' Workshop, Denver, Colorado, Nov. 3-5, 2016, URL: https://advancedssa.com/assets/img/workshop
22) "MDA Sees Better At Sea... Improves Ocean Applications Imaging Modes For RADARSAT-2," SatNews Daily, Oct. 1, 2015, URL: https://www.satnews.com/story.php?number=706086251
23) M. Chabot, C. Decoust, C. Lambert, P. Rolland, P. Ledantec, D. Williams, "RADARSAT-2 system operations and performance," Proceedings of the IGARSS (International Geoscience and Remote Sensing Symposium) 2015, Milan, Italy, July 26-31, 2015
24) Dean G. Flett, Eric J. Arsenault, Ron W. Pietsch, Daniel Bujold, Stephen Schaller, Satish K. Srivastava, "Enhanced Management of Orders and Conflicts (EMOC) process:coordinating RADARSAT-2 data orders for Canadian government users," Proceedings of the IGARSS (International Geoscience and Remote Sensing Symposium) 2015, Milan, Italy, July 26-31, 2015
25) Information provided by Wendy Keyzer of MDA, Richmond, BC, Canada.
26) M. Chabot, C. Decoust, P. Ledantec, D. Williams, A. Hillman, P. Rolland, R. Périard, "RADARSAT-2 System Operations and Performance," Proceedings of IGARSS (IEEE Geoscience and Remote Sensing Society) 2014, Québec, Canada, July 13-18, 2014
27) Dan Williams, Pierre LeDantec, Marielle Chabot, Anthony Hillman, Kenny James, Ron Caves, Alan Thompson, Cathy Vigneron, Yan Wu, "RADARSAT-2 Image Quality and Calibration Update," Proceedings of EUSAR 2014 (10th European Conference on Synthetic Aperture Radar), Berlin, Germany, June 3-5, 2014
28) "Petermann on the move," ESA, March 17, 2014, URL: http://www.esa.int/spaceinimages/Images/2014/03/Petermann_on_the_move
29) "MDA Joins ESA Third Party Mission Program," GEO Informatics, Dec. 13, 2013, URL: http://www.geoinformatics.com/blog/latest-news/mda-joins-esa-third-party-mission-program
30) "MDA increases provision of RADARSAT-2 information and services in Europe," Oct. 21, 2013, URL: http://www.newswire.ca/en/story/1246103/mda-incre
31) "International effort helps users get ready for Sentinel-1," ESA, May 6, 2013, URL: http://www.esa.int/Our_Activities/Observing_the_Earth/
32) G. Davidson, V. Mantle, B. Rabus, D. Williams, D. Geudtner, "Implementation of TOPS mode on RADARSAT-2 in Support of the Sentinel-1 mission," CEOS-WGCV-SAR Subgroup, October, 16, 2013, URL: http://sarcv.ceos.org/site_media/media/documents/16h00_CSA_ASAR_RTOPS_submit.pdf
33) Information provided by Alfred Ng of CSA, Canada.
34) "Traffic monitoring with RADARSAT-2," URL: http://www.fhr.fraunhofer.de/en/businessunits/Air-and-space
35) Delphine Cerutti-Maori, Ishuwa Sikaneta, Christoph Gierull, "Optimum SAR/GMTI Processing and Its Application to the Radar Satellite RADARSAT-2 for Traffic Monitoring," IEEE Tansactions on Geoscience and Remote Sensing, Vol. 50, Issue 10, Oct. 2010, pp. 3868-3881
36) "MDA to increase its provision of RADARSAT-2 imagery in support of Europe's Global Monitoring for Environment and Security program," MDA, July 20, 2012, URL: http://www.mdacorporation.com/corporate/news/pr/pr2012072001.cfm
37) Phillipe Rolland, Marielle Chabot, René Périard, Pierre Ledantec, Camille Decoust, "RADARSAT-2 Mission Operations Status," Proceedings of IGARSS (International Geoscience and Remote Sensing Symposium), Munich, Germany, July 22-27, 2012
38) "RADARSAT-2 data contributes to a new understanding of Antarctic terrain," CSA, August 18, 2011, URL: http://www.asc-csa.gc.ca/eng/media/news_releases/2011/0818.asp
39) "New NASA satellite map shows complete Antarctic ice flow," URL: http://blog.al.com/space-news/2011/08/new_nasa_satellite_map_shows_c.html
40) Bernd Scheuchl; Jeremie Mouginot; Eric Rignot, "Antarctic Ice Velocity Mapped from Space: A Game Changing New Product," September 2011, URL: http://www.waisworkshop.org/presentations/2011/Session4/Scheuchl.pdf
41) "NASA Research Yields Full Map of Antarctic Ice Flow," NASA/JPL, August 18, 2011, URL: http://www.jpl.nasa.gov/news/news.cfm?release=2011-256
42) E. Rig not, J. Mouginot, B. Scheuchl, "Ice Flow of the Antarctic Ice Sheet," Science, Sept. 9, 2011, Vol. 333, No 6048, pp. 1427-1430, DOI: 10.1126/science.1208336
43) "MDA to provide radar information in support of maritime safety to the U.S. National Geospatial-Intelligence Agency," MDA, Sept. 23, 2011, URL: http://www.mdacorporation.com/corporate/news/pr/pr2011092304.cfm
44) A. Hillman, P. Rolland, M. Chabot, R. Périard, P. Ledantec, N. Martens, "RADARSAT-2 Mission Operations Status," Proceedings of IGARSS (International Geoscience and Remote Sensing Symposium), Vancouver, Canada, July 24-29, 2011
45) Daniel De Lisle, "RADARSAT-2 Government of Canada Data Utilization," Proceedings of IGARSS (International Geoscience and Remote Sensing Symposium), Vancouver, Canada, July 24-29, 2011
46) Anthony Luscombe, Pierre leDantec, "RADARSAT-2 Image Quality and Calibration Update –Use of Database Information," Proceedings of the CEOS SAR Cal/Val Workshop, Zürich, Switzerland, Aug. 25-27, 2010
47) A. Hillman, P. Rolland, R. Périard, T. Luscombe, M. Chabot, C. Chen, N. Martens, "RADARSAT-2 Continuing System Operations and Performance," Proceedings of IGARSS (IEEE International Geoscience and Remote Sensing Symposium) 2010, Honolulu, HI, USA, July 25-30, 2010
49) Bob Slade, "RADARSAT-2 Product Description," MDA, RN-SP-52-1238, Issue ⅛: April 15, 2011, URL: http://gs.mdacorporation.com/includes/documents/RN
50) Marco van der Kooij, "Examples of InSAR and other land applications RADARSAT-2," Remote Sensing – The Synergy of High Technologies, Moscow, Russia, April 25-27, 2012, URL: http://www.sovzondconference.ru/upload/medial
51) Anthony Hillman, "RADARSAT-2 Mission Management – experience from commercial remote sensing flight operations," Proceedings of the SpaceOps 2010 Conference, Huntsville, ALA, USA, April 25-30, 2010, paper: AIAA 2010-1950
52) Jeff Hurley, "Operational Review: RADARSAT-1 & -2," SEASAR Workshop 2010, January 25-29, 2010, Frascati, Italy, URL: http://earth.eo.esa.int/workshops/seasar2010/8_Hurley.pdf
53) A. Hillman, P. Rolland, R. Périard, A. Luscombe, M. Chabot, C. Chen, N. Martens, "RADARSAT-2 continuing system operations and performance," Proceedings of IGARSS (IEEE International Geoscience and Remote Sensing Symposium) 2010, Honolulu, HI, USA, July 25-30, 2010
54) "RADARSAT-2 images used to monitor Haitian disaster to assess and support rescue," Jan. 29, 2010
55) Ahmed Mahmood, "RADARSAT-1, RADARSAT-2 and RCM," GSCB ()Ground Segment Coordination Body) Workshop, June 18-19, 2009, ESA/ESRIN Frascati, Italy, URL: http://earth.esa.int/gscb/papers/3.3_Mahmood.pdf
56) A. Hillman, P. Rolland, R. Périard, A. Luscombe, M. Chabot, C. Chen, N. Martens, "RADARSAT-2 Initial System Operations and Performance," Proceedings of IGARSS (IEEE International Geoscience and Remote Sensing Symposium) 2009, Cape Town, South Africa, July 12-17, 2009
57) Anthony Luscombe, "Image Quality and Calibration of RADARSAT-2," Proceedings of IGARSS (IEEE International Geoscience and Remote Sensing Symposium) 2009, Cape Town, South Africa, July 12-17, 2009
58) Information provided by Daniel De Lisle of CSA, St. Hubert, Quebec, Canada
59) D. De Lisle, "RADARSAT-2 Program Update," Proceedings of EUSAR 2008, 7th European Conference on Synthetic Aperture Radar, June 2-5, 2008, Friedrichshafen, Germany
60) A. A. Thompson, A. Luscombe, K. James, P. Fox, "RADARSAT-2 Mission Status: Capabilities Demonstrated and Image Quality Achieved," Proceedings of EUSAR 2008, 7th European Conference on Synthetic Aperture Radar, June 2-5, 2008, Friedrichshafen, Germany
61) A. Luscombe, P. LeDantec, K. James, A. Thompson, P. Fox, "RADARSAT-2 SAR Imaging Performance and Calibration," Proceedings of EUSAR 2008, 7th European Conference on Synthetic Aperture Radar, June 2-5, 2008, Friedrichshafen, Germany
62) Luc Brûlé, Jill Smyth, Daniel DeLisle, Michael Manore, Mario Lagrange, Jean-Marc Chouinard, "RADARSAT-2: Capabilities and Benefits for the Canadian Government," Proceedings of the 59th IAC (International Astronautical Congress), Glasgow, Scotland, UK, Sept. 29 to Oct. 3, 2008, IAC-08.B1.5.11
63) G. C. Staples, J. Hornsby, "Turning the Scientifically Possible into the Operationally Practical: RADARSAT-2 Polarimetry Applications," Proceedings of IGARSS 2002, Toronto, Canada, June 24-28, 2002
64) P. Arsenault, , C. Grenier, I. Barnard, A. Baylis, "RADARSAT-2 Antenna Measured Beam Pattern Performance and Comparison with Software Predictions," Proceedings of EUSAR 2006, Dresden, Germany, May 16-18, 2006
65) C. Grenier, I. Barnard, P. Arsenault, "The RADARSAT-2 Synthetic Aperture Radar Phased Array Antenna Performance Analysis Methodology," Proceedings of EUSAR 2004, Ulm, Germany, May 25-27, 2004
66) A. Luscombe, A. Thompson, K. James, P. Fox, "Calibration Techniques for the RADARSAT-2 SAR System," Proceedings of EUSAR 2006, Dresden, Germany, May 16-18, 2006
67) R. Touzi, R.K. Hawkins, S. Côté, "Data Quality Assessment of Polarimetric Radarsat-2: Preliminary Results," Proceedings of the 4th International POLinSAR 2009 Workshop, Jan. 26-30, 2009, ESA/ESRIN, Frascati, Italy, URL: http://earth.esa.int/workshops/polinsar2009/participants/265/pres_5_Touzi_265.pdf
68) A. Brand, C. Grenier, I. Barnard, "RADARSAT-2 T/R Module Development," Proceedings of the 13th Canadian Astronautics Conference, ASTRO 2006, Montreal, QC, Canada, organized by CASI (Canadian Astronautics and Space Institute), April 25-27, 2006
69) S. Riendeau, C. Grenier, "RADARSAT-2 Antenna," Proceedings of the 2007 IEEE Aerospace Conference, Big Sky, MT, March 3-10, 2007
70) S. Chiu, C. Gierull, "Multi-Channel Receiver Concepts for RADARSAT-2 Ground Moving Target Indication," Proceedings of EUSAR 2006, Dresden, Germany, May 16-18, 2006
71) P. D. Beaulne, C. H. Gierull, C. E. Livinstone, I. C. Sikaneta, S. Chiu, S. Gong, M. Quinton, "Preliminary design of a SAR-GMTI processing system for RADARSAT-2 MODEX data," Proceedings of IGARSS 2003, Toulouse, France, July 21-25, 2003
72) P. D. Beaulne, C. E. Livingstone, "An Experiment Plan to Test RADARSAT's-2 GMTI Capabilities," Proceedings of EUSAR 2006, Dresden, Germany, May 16-18, 2006
73) Shen Chiu, Chuck Livingstone, Ishuwa Sikaneta, Christoph Gierull, Pete Beaulne, "RADARSAT-2 Moving Object Detection Experiment (MODEX)," Proceedings of IGARSS 2008 (IEEE International Geoscience & Remote Sensing Symposium), Boston, MA, USA, July 6-11, 2008
74) T. Luscombe, "RADARSAT-2: Early Results," EOBN 2008 (Earth Observation Business Network), May 13-14, 2008, Richmond, BC, Canada
75) R. J. Quinn, "Transformation in Maritime Domain Awareness: Project Polar Epsilon and Automated Identification System," Proceedings of the 13th Canadian Astronautics Conference, ASTRO 2006, Montreal, QC, Canada, organized by CASI (Canadian Astronautics and Space Institute), April 25-27, 2006
76) P. J. Butler, "Project Polar Epsilon: Joint Space-based Wide Area Surveillance and Support Capability," Proceedings of IGARSS 2005, Seoul, Korea, July 25-29, 2005
77) J. Howes, "Polar Epsilon: Joint Space-Based Wide Area Surveillance and Support Capability," EOBN 2008 (Earth Observation Business Network), May 13-14, 2008, Richmond, BC, Canada
78) P. J. Butler, "Project Polar Epsilon: Joint Space-Based Wide Area Surveillance and Support Capability," Proceedings of IGARSS 2005, Seoul, Korea, July 25-29, 2005, Vol. 2, pp. 1194-1197
79) S. K. Srivastava, P. Rolland, "Meeting Global Customers Needs of RADARSAT-2 Data," 58th IAC (International Astronautical Congress), International Space Expo, Hyderabad, India, Sept. 24-28, 2007, IAC-07-B1.4.04
81) P. Meisl, C. Pearce, D. Comi, "RADARSAT-2 ground segment," Canadian Journal of Remote Sensing, Vol. 30, No 3, 2004, pp. 295-303
82) K. Beckett, A. Thompson, A. Luscombe, G. Stirling, "Optimization of RADARSAT-2 SAR Imagery for Vessel Detection Applications," Proceedings of the SPIE Remote Sensing Conference, Toulouse, France, Vol. 7826, Sept. 20-23, 2010, paper: 7825-10, 'Remote Sensing of the Ocean, Sea Ice, and Large Water Regions 2010,' edited by Charles R. Bostater Jr., Stelios P. Mertikas, Xavier Neyt, Miguel Velez-Reyes, doi: 10.1117/12.865058
The information compiled and edited in this article was provided by Herbert J. Kramer from his documentation of: "Observation of the Earth and Its Environment: Survey of Missions and Sensors" (Springer Verlag) as well as many other sources after the publication of the 4th edition in 2002. - Comments and corrections to this article are always welcome for further updates (email@example.com).