The pre-launch Envisat Calibration and Validation Plan for all instruments, including ASAR, can be found below, whereas a more detailed ASAR specific pre-launch Cal/Val plan is available.
The results of the Envisat commissioning phase activities were presented during two workshops: the Envisat Calibration Review held at ESA/ESTEC from 9 to 13 September 2002 and the Envisat Validation Review held at ESA/ESRIN from 9 to 13 December 2002. Download proceedings from the Calibration Review .
In addition to the five imaging modes, ASAR could acquire data in two calibration modes: External Calibration (EC) and Module Stepping Mode (MSM). MSM was used to monitor the stability and behaviour of the ASAR antenna, whereas the EC related to the signal magnitude and how that is transformed to physical units (radiometric calibration).
Full details on the ASAR calibration activities throughout the mission are available in the paper presented at the ESA Living Planet Symposium in 2013 by Miranda et al. (2013): The Envisat ASAR Mission: A look back at 10 years of operation.
Information on the geometric validation of ASAR imagery is available from the Proceedings of the 2004 Envisat & ERS Symposium.
Cyclic reports relating to the global validation of ASAR Wave Mode products are available online. A summary of the quality assessment, calibration and validation activities of ASAR Wave Mode data can be found in the annual reports for 2010 and 2011.
The objective of the ASAR instrument internal calibration scheme was to derive the instrument internal path transfer function, and to perform noise calibration. This objective was realised by dedicated calibration signal paths and special calibration pulses within the instrument for making the required calibration measurements and by using these measurements to perform corrections within the ground processor.
Transmit / Receive modules temperature compensation
The Transmit/Receive (T/R) module amplitude and phase characteristics vary principally as a function of temperature. Therefore the instrument included a scheme to compensate for drifts over temperature. This scheme provided the antenna with a high degree of stability; however, it did not compensate for the aging effect or T/R module failures. Also, under conditions of rapid temperature variation, such as eclipse, the compensation performance was degraded. Therefore, it was necessary to include the active antenna components within the calibration loop.
The instrument calibration loop was used to perform three distinct functions. Firstly, it was used to characterize the instrument transfer function during the measurement modes. Secondly, it was used to characterize individual T/R modules. Finally, it was used in the special external characterization mode.
The calibration loop in ASAR was in fact comprised of a distinct calibration path to each of the 320 T/R modules. This enabled transmit pulses at each T/R module output to be sampled, and allows calibration pulses to be injected into the receiver front end of each T/R module. Effectively, the scheme provided a multi-pathed calibration loop that encompasses all the active electronics in the instrument transmit and receive paths. In particular, aging of T/R modules characteristics and T/R module failure could be sensed.
There was no active switching within this network in order to maximize its reliability and stability. The calibration distribution network acted as a combiner when the loop was being used to sense T/R module transmissions, and as a splitter when the loop was being used to inject pulses into the T/R module receivers. The antenna calibration port could be switched either to an auxiliary receiver or to an auxiliary transmitter, both of which are located within the instrument central electronics. These elements could be used to sense or inject calibration pulses at the antenna calibration port. The detailed use of the calibration loop was partly controlled by the operating states of the T/R modules themselves (i.e., ON/OFF, Tx/Rx, H/V), because there was no switching within the calibration network.
During normal operation in any of the ASAR measurement modes, a sequence of calibration pulses was interleaved with the normal radar pulses. These pulses characterized the active array, both on transmit and receive, on a row by row basis (i.e. only 10 modules along one row are activated, while the 310 remaining modules are off). For different pulses within the sequence, different rows were activated. The rationale for row by row characterization is that ASAR was essentially an elevation plane beam steering instrument. Thus, the amplitude and phase settings applied to the T/R modules along a row are nominally uniform, and the calibration signals from them were nominally coherent.
For each of the 32 rows, the antenna and the central electronics were characterized with 3 types of pulses. Pulse P1 characterizes the transmit chain of the instrument.
However, since T/R modules of the 4 adjacent rows share the same power supply, the 10 modules of the 'wanted' row were set to their nominal phase and amplitude settings for pulse P1, while the phase of the modules of the 3 'unwanted' rows were set so that their combined contribution out of the calibration network was nominally zero. Thus, their interference to the measurement of the 'wanted' row was minimized.
A second type of transmit pulse, referred to as pulse P1A, was added, in order to characterize the residual parasitic contribution of the 3 unwanted rows during P1. During P1A, the 3 unwanted rows were set as for P1, and the previously wanted row was now switched off. Even though the load conditions on power supplies were not exactly representative, the small error introduced on the estimation of P1A could be considered as small enough to be neglected. The receive path of the instrument was also characterized with a so called pulse P2, but, on receive path, no variation was expected from power supply load variations, and a row by row characterization was possible.
The central electronics transmit and receive paths are included in both P1/P1A and P2 characterizations. It was therefore necessary to characterize the central electronics independently by the use of the internal pulse P3.
Internal calibration processing
One consequence of row by row characterization was that the instrument transfer function cannot be simply calculated from a few pulses, as this was the case in the AMI SAR. Instead, the ground processor needed to utilize the calibration pulses from a complete cycle through the 32 rows to estimate the transfer function. Also, a replica pulse for the instrument had to be calculated from a complete row cycle.
As well as providing internal calibration during the measurement modes, ASAR included a special module stepping mode, in which individual T/R module characteristics could be measured. This mode could be used to investigate T/R modules failures and aging effects. In this mode, only one module was activated at a time, either on transmit or in receive.
The internal calibration scheme also included measurements of the instrument noise level. The measurements were included in the initial calibration sequences, at the beginning of a mode. In the modes which have natural gaps in their imaging sequence (i.e., wide swath and global monitoring modes), noise measurements were also made during nominal operation throughout the mode.
The internal calibration scheme monitored drifts in the transfer function of the majority of the instrument, excluding the passive part of the antenna, the calibration loop itself and the mechanical pointing of the antenna. As part of the overall calibration strategy to monitor these elements a dedicated mode of ASAR called External Characterisation Mode was used nominally every six months.
During this mode a sequence of pulses sent by each antenna row in turn was simultaneously sensed by the antenna calibration loop and recorded on ground by a special ground receiver built in the ASAR transponder.
From data recorded in the transponder and data down-linked from the instrument the relative phase and amplitude of the pulse from each row were compared in the ground processor. The relative amplitude and phase was used to characterise the row of radiating sub-arrays and the calibration path from the row.
The external calibration scheme with the objective to derive the overall calibration scaling factor used the successful methodology developed for ERS-1/2 for the narrow swath mode.
Three specially built high precision transponders with a radar cross section high enough compared to background backscattering coefficient and noise were deployed across the ASAR swath and observed several times during every 35 days orbit cycle. Images acquired over suitable area of the amazon rain forest were used to derive the in-flight elevation antenna pattern. Absolute calibration factors derived from transponder measurements and across swath correction derived from the radar equation were used to calibrate the final image product.
Ground processing calibration
As part of the processor Data Handling and Reformatting I/Q science data were uncompressed and were subject to an I/Q correction (bias, differential gains, non-orhogonality). Like ERS any non linearity correction could be applied in the Ground Processor using pre-launch instrument ADC characterisation.
As part of the Ground Processor Internal Calibration, the amplitude and phase of calibration pulses (P1, P1A, P2, P3) for each row were used. The amplitude and phase of P2 relative to P3 were calculated and P1A was vectorially substracted from P1 as discussed earlier. External characterisation data and the derived amplitude and phase values for the 32 rows on transmit and receive were used to measure any deviation of the instrument reference gain pattern from its ground characterised value.
The replica of the transmitted pulse was calculated from the P1, P1A, P2 and P3 measurements, the ground characterised row patterns and the external characterisation data. The constructed replica, tracks variations in all the transmit and receive circuits and was used to determine the range reference function for range compression processing.
The ground processor included a Doppler Centroid Estimator with specified accuracy of 50 Hz for image and wave mode like ERS and 25 Hz in ScanSAR modes in order to control azimuth radiometric errors.
Search with Keywords
- Bathymetry and Seafloor Topography
- Human Dimensions
- Imaging Radars
- Land Surface
- Land Use and Land Cover
- Natural Hazards and Disaster Risk
- Ocean Temperature
- Ocean Waves
- Sea Ice
- Snow and Ice
- Solid Earth
- Terrestrial Hydrosphere
- Volcanic Activity
- Weather Events