3.2.1 Preflight Characteristics and Expected Performance

Compared to the ERS AMI-SAR, ASAR offers five different modes of operation, dual polarization and a total of seven beams with a combined ground-range of 485km covering incidence angles from 15° to 45°. In spite of this substantial increase in functionality, the intention from the outset was that there would be no significant degradation in the various performance parameters such as spatial and radiometric resolution, sensitivity and ambiguity suppression compared to ERS.
As a result and in order to be convinced that this high goal would be met, performance prediction analyses have been repeatedly made and updated since the earliest days of the project using principally the specially developed PEAS software (Performance Evaluation and Assessment Software).

Confidence in the results obtained using PEAS is high, thanks to independent corroborative analyses and comparison, where possible, of the predictions and measurements made on the FM instrument.

Performance Objectives are shown in Level 1B Accuracy 2.6.1.3. .

Performance Prediction Methodology

The ASAR antenna with its 320 T/R modules requires a different philosophy when determining performance than say, ERS, since instead of a single point failure it has 320. That being said, it is reasonable to assume that not all of the modules will survive launch and four years of continuous operation. This forms the basis of the performance analysis philosophy the aim of which is to predict the worst case expected performance at end of life following a period of so-called graceful degradation.

To achieve this, two main issues with respect to the T/R modules are considered, namely: module failure and phase and amplitude setting errors.

Module Failure

From assessments of component reliability a figure of 6% failure rate has been estimated by industry for the T/R modules at end-of-life. This corresponds to about 20 failed modules.

Module Setting Errors

Each T/R module can be set to provide a certain phase and amplitude in order to generate the required elevation beam patterns. Six bits are available the required phase from 0° to 360° and a further six to specify the amplitude in the range 0 to -20dB. From measurements made at component and instrument level it has been determined that the phase setting accuracy is approximately ±5° and the amplitude setting accuracy approximately ±0.5dB..

Generating "End-of-Life" Antenna Patterns

In order to model accurately the impact of T/R module failures and setting errors on the ASAR performance it is possible to create a failure/error matrix representing the antenna and use it in producing representative simulated elevation beam patterns. Each pattern thus generated would have a slightly different shape to the one desired with some sidelobes becoming larger and nulls appearing at different look angles. The problem with this approach is clearly the extremely large number of possible permutations of failures and setting errors.

Instead, the method chosen was to calculate a disturbance level which could be applied linearly to the calculated elevation patterns with the effect of increasing sidelobe levels and "filling-in" nulls.

Ideally T/R modules would fail in an evenly distributed random pattern across the antenna. However, the possibility that failures might be clustered could not be ignored. This was accounted for by calculating the variance in the sidelobe level as a result of failures and basing the disturbance level to be applied on a 90% probability (two standard deviations) of the antenna being able to generate the required pattern within the sidelobe levels set (figure3.21 ).

Figure 3.21 Effect of Disturbance Level on Elevation Beam Pattern. Black: FM measured pattern, Red: sidelobe disturbed pattern at 90% probability

Sigma Nought Model

A similar "worst case" approach was taken in order to determine the distributed target ambiguity suppression ratio with respect to the normalized radar cross section of the in-swath target region and the ambiguous regions. From measurements provided in the literature ( Ref. [3.1 ] ) on C-band backscatter over land and sea or ice two graphs were produced. In swath, the sigma nought would be assumed to be the lower of the two curves while outside the swath, the ambiguous regions, would be assumed to have a sigma nought corresponding to the higher curve (see figure3.22 ).

Figure 3.22 ASAR sigma nought model showing the position of the IS3 swath

At low incidence angles this can lead to more than 15dB difference between target and ambiguous region sigma nought which puts extra constraints on the elevation pattern sidelobe levels.

Predicted Performance

The predicted performance of ASAR has been determined using the ASAR PEAS software which makes use of algorithms based on those used successfully to predict the performance of the ERS SARs. All FM instrument measured characteristics have been used as input parameters to PEAS together with the sidelobe disturbed FM beam patterns to produce the most accurate possible performance predictions for end-of-life operation.

Figure3.23 shows the sensitivity expressed as noise equivalent sigma nought for all image beams (IS1-IS7) compared to the ASAR sigma nought model. There is several dB margin everywhere for all beams compared to the sigma nought curves with the exception of the far range of IS5 over land which in any case overlaps with the near range of IS6.

Figure 3.23 Noise Equivalent Sigma Nought for Image Mode beams IS1-IS7

In figure3.24 the range distributed target ambiguity suppression for beams IS1-IS7 is shown. For each beam the measured FM pattern has been used in the calculation to show the start-of-life situation (lower set of curves) as well as the sidelobe disturbed version of the FM patterns to demonstrate the worst case (90% probability) expected suppression at end-of-life (upper set of curves). This shows that at start-of-life the DTRAR is better everywhere than about -25dB and even at end-of-life it should never be worse than about -20dB.

Figure 3.24 Distributed Target Range Ambiguity Ratio

A summary of all the end-of-life performance parameters and how they compare to those calculated or estimated for ERS is given in table 2.

Conclusions

Throughout the pre-flight assessment of the performance of the ASAR instrument a worst case scenario has been assumed including all component related issues. The result of this approach is that there is now high confidence that the ASAR will meet all the requirement specifications throughout its planned life. This means that the high standards which were set by the ERS AMI SARs will be maintained allowing ASAR to become a valuable instrument in terms of its potential for applications.

REFERENCES

Ref 3.1

R. Torres, C. Buck, J. Guijarro and J-L. Suchail (ESA-ESTEC, NL). "ESA's Ground Breaking Synthetic Aperture Radar: The ENVISAT-1 ASAR Active Antenna". APS'99.

Ref 3.2

J-L. Suchail, C. Buck, J. Guijarro and R. Torres (ESA-ESTEC, NL). "The Development of ENVISAT-1 Advanced Synthetic Aperture Radar". RADAR'99.

Ref 3.3

P. Mancini, J-L. Suchail, R. Torres, J. Guijarro and C. Buck (ESA-ESTEC, NL). "The ENVISAT-1 Advanced Synthetic Aperture Radar. The Development Status". CEOS'98.

Ref 3.4

J. Guijarro, C. Buck, P. Mancini, J-L. Suchail and R. Torres (ESA-ESTEC, NL). "The Development of the ENVISAT-1 ASAR". IGARSS'96.