ESA Earth Home Missions Data Products Resources Applications
    24-Jul-2014
EO Data Access
How to Apply
How to Access
Index
Credits
RA2/MWR Data Formats Products
Glossary
Frequently Asked Questions
Main RA-2 Level 2 Algorithms (Non-Ocean)
Main MWR Level 1b algorithms
Main RA-2 Level 1b algorithms
MWR Instrument
RA-2 Instrument
RA-2 and MWR Instruments
MWR Instrument Characteristics and Performance
In-flight performance verification
MWR Instrument Description
RA-2 Instrument Characteristics and Performance
In-flight Performance verification
RA2/MWR Products and Algorithms
RA-2/MWR Auxiliary files
Common Auxiliary Data Sets
Level 2 processing
Level 1b processing
Specific Topics Related To The Radar Altimeter
Orbit interpolation
Measurement Reference
Datation
Time Handling And Leap Seconds
RA-2/MWR Level 2 Products And Algorithms
RA2/MWR Level 2 Products
Averaged Ku chirp band
Ocean depth/land elevation
RA-2 ionospheric correction
RA-2/MWR Level 1b Products and Algorithms
RA-2/MWR Level 1b product
MWR Level 1b algorithms
Level 0 products
MWR Level 0 products
Products Evolution History
Definitions and Conventions
Organisation of Products
Data Handling Cookbook
Characterisation And Calibration
Latency, Throughput And Data Volume.
Product size
RA2/MWR Products User Guide
Further reading
How to use RA-2/MWR data
How To Choose RA-2/MWR Data Products
Summary of Applications vs. Products
Peculiarities of MWR
Peculiarities of RA2
Geophysical Coverage
Principles Of Measurement
Scientific Background
ENVISAT RA2/MWR Product Handbook
Services
Site Map
Frequently asked questions
Glossary
Credits
Terms of use
Contact us
Search


 
 
 


4.1 RA-2 Instrument

Question 4.1 : Which are the innovative characteristics of the Envisat Radar instrument?

The radar altimeter on board EnviSat is a nadir looking pulse limited radar altimeter functioning at the main nominal frequency of 13.575 GHz (Ku band), at which the instrument automatically detects, acquires and tracks the earliest part of the radar echoes from any surface without interruption. A secondary channel at a nominal frequency of 3.2 GHz (S band) is also operated to estimate the ionospheric range delay.

The new features of the RA-2, compared to previous altimeters are mainly:

  • the robustness of the on-board tracker, not particularised for ocean waveforms but with an alternative algorithm that generates a height error signal which is linear over the entire width of the range window, independently of the pulse shape
  • the capability of the instrument to autonomously adapt its resolution to be able to follow any type of surface

Question 4.2 : Why is it that RA-2 does not lose track whatever is the surface type underneath?

An innovative tracking algorithm, known as Model Free Tracker (MFT), has been designed on RA-2 such that the estimate of radar to surface range, echo power or other echo features is not so accurate, while the instrument concentrates its effort maintaining the earliest part of radar echoes within the tracking window, independently of its shape.

In particular, the MFT will decide whether the range window is using the adequate resolution, whether the resolution could be increased, or decreased, based on the Signal to Noise ratio (SNR) of the on-board waveform position compared to reference values stored in the on-board memory.

When the radar echo is about to move out of the tracking window, due for example to a sudden change in the surface elevation, the window is broadened to recapture the echo. This allows uninterrupted radar operation over all type of surfaces and its boundaries, and will avoid a dedicated mode change, commanded from ground. Ocean, ice sheets and land surface topography are measured to the highest possible accuracy, and new terrains are tracked

Question 4.3 : Why are there three different Ku resolutions on RA-2?

The RA-2 is provided with three different bandwidths:

  • 320 MHz, with resolution of 47 cm and tracking window width of 61 m
  • 80 MHz, with resolution of 190 cm and tracking window width of 243 m
  • 20 MHz, with resolution of 750 cm and tracking window width of 960 m

so that the resolution can be adapted to different scenarios (ocean, ice, ice-sheet, sea ice and land).

Question 4.4 : Is the resolution change done manually or automatically?

The change of the resolution is done autonomously by the instrument, through the so called resolution selection logic, to optimally tune the width of the tracking window to the topographic features observed.

That is; if the instrument is, say, on the maximum resolution (320 MHz bandwidth) and the surface type changes, the resolution will autonomously change (from 320 MHz to 80 MHz) in order to keep always the echo within the tracking window, avoiding, as it happened on ERS-2 RA, to lose track.

If at 80 MHz it's still difficult to maintain the echo within the window, the resolution shall be again changed from 80 to 20 MHz.

Only under critical conditions, over abrupted terrains, when the 20 MHz window width is not enough to hold the echo, the track is lost. This lasts about 0.6 s (12 Data Blocks) and is called Acquisition sequence, needed in order to properly reinitialise the tracking. During this period, a Noise power estimation phase, a detection phase and an AGC setting phase occur.

The tracking will restart after that, with the instrument at 20 MHz. Then, according to the SNR of the on-board waveform position, as explained in the previous paragraph, the instrument will decide if the resolution can be changed, first to 80 MHz and then to 320 MHz.

Question 4.5 : Which is the advantage of having a dual frequency instrument?

RA-2 is a nadir looking pulse limited radar functioning at the main nominal frequency of 13.575 GHz (Ku band).

A secondary channel at nominal frequency of 3.2 GHz (S band) is also operated to compensate the range error on altitude measurements caused by the propagation of the radar signals through the ionosphere.

Furthermore, a dual frequency instrument is used in the retrieval of the rain algorithm and is used for scientific research using the difference of backscattering coefficients for the two bands.

Question 4.6 : Why two extra DFT samples have been included on RA-2?

RA-2 provides 128 samples and 2 DFT Ku samples of the waveform compared to the 64 samples of ERS. For extremely low SWH conditions, range and SWH estimates can be further improved by including in the precision processing, the use of two extra echo samples, computed by the on board digital processor through a standard DFT algorithm and made available for each Ku band radar echo down linked to ground. The location of the two additional waveform samples is programmable from ground so that it can be optimally tuned with respect to the position of the tracking point in the FFT bank. This is a unique feature of the RA-2.

Question 4.7 : Why are individual echoes provided by RA-2?

RA-2 is able to transmit to ground In Phase and Quadrature components (i.e. raw data without any processing applied on board) of the echoes from 2000 consecutive Ku radar pulses. In this concept, the full-rate data are stored, for a short burst, in an internal buffer memory, in parallel to the normal averaging and other functions of the instrument. The buffered data are subsequently read out at a such lower rate and appended to the normal science data (RA2 Level 0 ISPs).Experimental processing of these individual echoes on ground can provide more insights on surface topography at the boundaries.

These individual echoes are available as a dedicated Data Set in the L1b products and in the Level 2 SGDR products.

Question 4.8 : Which are the quality flags to be used for editing RA2 and MWR data?

By looking at every single flag in the L2 RA2 MCD (field 8 in all L2 products) the following information and useful advices for the users can be derived:

a) Packet length error flag (bit 0 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 (bad) when the source packet length inside the packet does not coincide with the length in the FEP annotations (i.e. error during transmission of data from the satellite).

ADVICE: If this flag is set to 1, the source packet is not processed at L1b (the corresponding L2 record will be set to default). Therefore, only source packets with this flag set to 0 should be used for any processing.

b) OBDH validity flag (bit 1 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 (bad) if the source packet counter difference of consecutive packets is different from 1 (data gap), or if the OBDH datation difference of consecutive packets (with counter difference = 1) is different from the expected value (584056) +/- a tolerance (i.e. instrument timing error).

ADVICE: If it happens only once, the data can be used for further processing. If the flag is set very often for consecutive packets, there could be an error in the instrument (i.e. corrupted OBDH values or data associated to a wrong datation by the on-board s/w). In this case the data should be discarded.

c) USO validity flag (bit 2 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 (bad) if the difference in the source packet counter of consecutive packets is different from 1 (data gap), or if the USO datation difference of consecutive packets (with counter difference = 1) is different from the expected value (111400) +/- a tolerance (i.e. instrument timing error).

ADVICE: If it happens only once, the data can be used for further processing. If the flag is set very often, there could be an error in the instrument (i.e. corrupted USO values or data associated to a wrong USO datation). In this case the data should be discarded.

d) Fault identifier (bit 3 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 if at least one of the 20 elementary measurements has the on-board width of the discriminator set to 0, the CoG discriminator out of range, the leading edge position out of range, the sum of the samples of the on board averaged waveform out of range, the AGC predicted rate out of range, the AGC corrected value out of range, the AGC X0 out of range, the time delay predicted rate out of range, the time delay corrected value out of range, the time delay X0 out of range, the SNR out of range, N' out of range or the waveform samples not available.

ADVICE: Most of the above errors situations are not likely to appear anymore since the corresponding on-board parameters were tuned during Comm. Phase to avoid them to happen. In the case of the first one, W set to 0 (i.e. very noisy waveform), the instrument automatically reacts by reducing the resolution.

Therefore, the data with this flag set to 1 can be used for further processing.

e) AGC Fault identifier (bit 4 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 if any of the 20 elementary AGC_Ku input values is out of bounds.

ADVICE: The min/max thresholds currently used to perform this check are wide enough to include all possible values of the L1b Ku AGC values. Therefore, no flagged records are expected.

Anyway, this flag can be eventually considered as editing criterion in case something wrong happened during the L1b processing.

f) Rx delay identifier (bit 5 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 if any of the 20 elementary Ku on-board Rx delay input values is out of bounds.

ADVICE: The min/max thresholds currently used to perform this check are wide enough to include all possible values of the L1b Ku Rx delay values.

Therefore, no flagged records are expected.

g) Waveform samples fault identifier (bit 6 in L2 RA2 MCD)

DEFINITION: This flag is set to 1 as soon as any of the 20 input elementary measurements has all Ku/S samples set to 0.

ADVICE: 1 second of data should not be thrown away only because of this flag. The 18 Hz retracking quality flags (fields 132 to 138 in FD/I/GDRs) indicate specifically which elementary measurement had all Ku/S samples set to 0 (corresponding bit set to 1).

h) MWR validity (bits 8-10 in L2 RA2 MCD)

DEFINITION: Bit 10 is set to 1 if a MWR data gap is present (failed check on consecutive MWR source packets OBDH datations). Bit 9 is set if there is something wrong with MWR thermal control (e.g. instrument temperatures out of bounds). Bit 8 is set if the blanking pulse is present.

ADVICE: Bit 10 just indicates if there was a gap between consecutive MWR SPs, not that the data inside the SPs is wrong. But 9 is a warning on the thermal control of the instrument that could imply a lower quality of the output Tbs (no need of throwing away the data in this case either). Bit 8 is always set to 0 by macrocommand.

i) Brightness temperatures range check (bits 11-12 in RA2 L2 MCD)

DEFINITION: These bits are set to 1 if the brightness temperatures values are out of bounds.

ADVICE: If these flags are set to 1, the corresponding Tbs should not be used for the processing.

j) Ku/S ocean/ice2/ice1 retracking and Ku sea-ice retracking quality flags (bits 16 to 22 in RA2 L2 MCD)

DEFINITION: These flags are set to 1 if the retracking failed for at least one of the 20 elementary measurements. That is; if input waveform samples are set to 0, input data not in Tracking, Preset Tracking or Preset Loop Output modes, too low average power wrt noise power, or leading edge position out of bounds.

ADVICE: There is no need of throwing away the whole source packet only for this flag. The 18 Hz retracking quality flags (fields 132 to 138 in FD/I/GDRs) indicate specifically which elementary measurement was wrong (corresponding bit set to 1).

k) Absence of processing errors (arithmetic faults) (bit 24 in RA2 L2 MCD)

DEFINITION: This flag is set to 1 if mathematical exceptions (e.g. division by 0, log of a negative number, ....) were encountered while processing any of the 20 elementary measurements.

ADVICE: There is no need of throwing away the whole source packet only for this flag. The 18 Hz retracking quality flags (fields 132 to 138 in FD/I/GDRs) indicate specifically which elementary measurement was wrong (corresponding bit set to 1).

l) Meteo data state (bits 25-26 in RA2 L2 MCD)

DEFINITION: These bits indicate if two valid Meteo files, two degraded Meteo files (i.e. too large time difference between the RA2 record and the Meteo time), only one or none, were used for the processing. The 2 bits flag is set to 0, 1, 2 or 3, respectively.

ADVICE: This flag should not be used for the editing. The user has only to know that in case this flag is set to 1 or 2, the quality of the meteo corrections will be lower than in case flag is 0. If the flag is set to 3, all meteo corrections will be set to default values.

m) Orbit status flag (bits 28-31 in RA2 L2 MCD)

DEFINITION: In NRT this flag can be set to 0 (if a fatal error happened while calling the orbit routines), to 1 ( orbit routines output ok for all 20 elementary measurements) or 2 (some non-fatal error occurred for at least one of the elementary measurements).

Currently these 4 bits are set to 0 in OFL products.

ADVICE: If the flag value is 2, the routines output is still valid and the data can be used for further processing. If the flag value is 0 the processing gets blocked and no more data is processed.

Question 4.9 : What's the use of the RA-2 peakiness?

The computation of the peakiness is part of the preprocessing Level 2 'ice' algorithms, aimed at determining if the received echo has a shape sufficiently recognisable so that an elevation may be determined from it.

The peakiness algorithm consists essentially in calculating the ratio of the maximum filter (bin) value to the mean filter value of the filters to the 'right' of the tracking point.

This processing is performed on 128 bins for Ku band, since the Ku waveforms are composed of 128 samples, and on 64 bins for S band.

Note that even for cases when the echo is regarded as non valid (i.e. if the echo waveform is contaminated by the surface return or if the leading edge does not lie within the range window) the peakiness is always calculated since this is an independent waveform quality assessment parameter.

Furthermore, the peakiness has proved to be very efficient at removing most of sea-ice data (around 95% of it, depending on the season), and in a lesser extent rain. In fact, ocean data could be retained if peakiness values are in the range 1.5-1.8.

Question 4.10 : What's the RA2 reference point for the derivation of the range?

The way in which the altimeter derives the range is by measuring the time since a radar pulse is transmitted and then received, after having bounced on the surface. The counting of the time starts when the pulse is transmitted by the Tx, and stops when the received pulse is mixed with a second pulse from Rx in the on-board deramp mixer.

However, the altimeter range is assumed to be referred to the antenna phase center. Therefore, the physical delays inside the altimeter hardware have to be measured and compensated for. These values have been characterised during testing on ground and are stored in an external auxiliary file.

They are applied to the Level 1b window time delay to obtain the final altimeter range measurement (at L2) with respect to the antenna phase center.

Moreover, since the satellite height is referred to the Centre of Mass (CoM), and the altimeter range to the antenna phase center, a new correction term has to be accounted for in order to obtain a coherent surface height (satellite height - altimeter range) with respect to the ellipsoid.

This is done by adding the distance between the CoM and the antenna phase centre, before calculating the final Level 2 altimeter range.

Question 4.11 : What is the set of criteria that can be used to exclude "non-ocean" data?

Most users are interested in editing non-ocean data. The following is a listing of the main fields/flags ESA makes available in the L2 products for this scope:

  • RA2 land/sea flag: values of this flag set to 0 or 1 indicate open oceans and semi enclosed seas, or enclosed seas and lakes, respectively. Values set to 2 or 3 indicate continental ice and land, respectively. This is the first criterion to consider when editing non ocean data.
  • MCD flags: see paragraph "Which are the quality flags to be used for editing RA2 and MWR data in this section
  • SWH: In principle, all values of SWH (apart from the default one, stored in the output product in case the number of valid elementary SWH values is less than a minimum value, set currently to 6, or in case the number of valid elementary SWH values with a scatter about the mean smaller than an upper bound is less than a threshold, set currently to 6) could be used for further analysis. It's up to the user to set more stringent limits for the accepted SWH values, according to the applications.
  • standard deviation of SWH: all values, apart from 0 and the default one, could be used for further processing. Infact, it has been seen by some users that high values of the standard deviation could be due in most cases to radar returns affected by sea ice, but not always. Therefore discarding records with standard deviation greater than a fixed value, might not be appropriate.
  • number of valid points used for the computation of SWH: This value (in the range 1 to 19) may be used in different ways for keeping valid records (e.g Nval> 15), according to users needs.
  • sigma0: In principle, all values of sigma0 (apart from the default one, stored in the output product in case the number of valid elementary sigma0 values is less than a minimul value, set currently to 6, or in case the number of valid elementary sigma0 values with a scatter about the mean smaller than an upper bound is less than a threshold, set currently to 6) could be used for further analysis. Actually, it has been seen by several users that negative values of sigma0 are almost always located at high latitudes, therefore, negative sigma0 values could be discarded for ocean applications.
  • standard deviation of sigma0: all values, apart from the default one could be used for further processing. However, for general analysis of wind speed, it has been recommended by several users that records with standard deviation greater than 0.3 be flagged as dubious, since the affected records do not seem to be only at high latitudes.
  • number of valid points used for the computation of sigma0: As in the case of SWH, the number of valid points may be used differently by the users according to the applications.
  • wind speed: In principle, all values (apart from the default one) could be used for further analysis. Yet, note that due to some limitations of the current wind algorithm, there is an accumulation of wind speeds at 20.15 m/s (from input sigma0 values less than 7 dB) and close to 1m/s (from all data with sigma0 > 15 dB). The wind algorithm is to be upgraded in a near future to cope with these problems.
  • range:In principle, all values of range (apart from the default one, stored in the output product in case the number of valid elementary range values is less than a minimum value, set currently to 6, or in case the number of valid elementary range values with a scatter about the mean smaller than an upper bound is less than a threshold, set currently to 6) could be used for further analysis.
  • standard deviation of range: all values, apart from the default one could be used for further processing. It's up to the user to set an upper limit according to the needs (e.g. some users have seen that most of the high range standard deviation values were at high latitudes, recommending a preliminary test on this parameter such that records with standard deviation greater than 0.2 or 0.7, for Ku and S bands respectively, or records with negative or zero value, could be discarded).
  • number of valid points used for the computation of the range: As in the case of SWH, the number of valid points may be used differently by the users according to the applications.
  • peakiness: this parameter has proved to be quite effective in removing sea-ice data (and in a lesser extent rain), apart from seasonal variations. Up to 95% of sea-ice data may be edited if keeping peakiness values in the range 1.5-1.8


Keywords: ESA European Space Agency - Agence spatiale europeenne, observation de la terre, earth observation, satellite remote sensing, teledetection, geophysique, altimetrie, radar, chimique atmospherique, geophysics, altimetry, radar, atmospheric chemistry