2.6.2 Level 1b product definition

This section only contains a high level description of the level 1B products. For more information refer to the “Envisat-1 Products Specifications - Volume 11 – MERIS Products Specifications” document ( R-8).

2.6.2.1 Level 1b High-Level Organisation of Products

The MERIS Level 1b product is composed of: the Main Product Header (MPH), the Specific Product Header (SPH), one Global Annotation Data Sets (GADS), two Annotation Data Sets and sixteen Measurement Data Sets. The MPH allows to identify the product and some of its main characteristics.

The SPH contains references to external data files and Data Sets descriptors, as well as general information applicable to the product such as sensor characteristics, PCD and metrics summary. The GADS contains all the data scaling factors and general information like reference extraterrestrial solar flux and some instrument settings which may be useful to analyse results.

The first ADS (LADS for location ADS) contains information on geolocation, measurement viewing and illumination geometry and auxiliary environment parameters for the tie points, a subset of the product pixels.

The second ADS (SQADS for summary quality ADS) contains quality information, aggregated at the level of a group of granules.

The first fifteen MDS are dedicated to top of atmosphere radiance measured in the 15 MERIS spectral bands and the last one to the associated flags: classification and measurement quality indicators.

Information coming either from input Level 0 product, from external data sources, or generated by any processing step are gathered, organised, scaled and coded according to specifications to build the Level 1b product file.

2.6.2.1.1 Reduced Resolution Geolocated and Calibration TOA Radiance

The high-level structure of the product is shown in the table below.

Table 2.6 - Reduced resolution product structure.

2.6.2.1.2 Full Resolution Geolocated and Calibration TOA Radiance

The high-level structure of the product is shown below in the table 2.20

Table 2.7 - Full resolution product structure.

2.6.2.1.3 Main Product Header

Main product header is formatted as described in the “Envisat-1 Products Specifications - Volume 5 – Product structures” document ( R-7). Only time of first and last frames of the product are input from the processing to the MPH formatting.

The Main Product Header (MPH) identifies the product and its main characteristics. It is an ASCII structure containing information needed for all ENVISAT sensors. It is of fixed length and format for all products. The MPH contains the following major types of information:

· Product Identification Information

· Information Regarding Data Acquisition and Processing

· Information on Time of Data

· Information on ENVISAT Orbit and Position

· SBT to UTC Conversion Information

· Product Confidence Data

· Product Size Information

2.6.2.1.4 Specific Product Header

Specific product header is formatted as described in the “Envisat-1 Products Specifications - Volume 11 – MERIS Products Specifications” document ( R-8). The PCDs, issued by the previous steps 1.1 to 1.7 (see Figure 2.6 - ), as well as the geolocation of first and last tie frames, from step 1.5.2, are inputs to the SPH (note that the transmission errors and the format errors counters are transformed into flags set if the mean numbers of errors per packet exceed given thresholds). In the case of the FR Scene Product, for which there is an even number of tie points, linear interpolation between the closest tie points is considered sufficiently accurate to compute geolocation of the mid sample of first and last frames.

2.6.2.1.5 Global Annotation Data Set

Global Annotation Data Set is formatted as described the “Envisat-1 Products Specifications - Volume 11 – MERIS Products Specifications” document ( R-8). Inputs come either from algorithm step 1.6 (solar flux corrected according to day of year) or from auxiliary data bases (gain settings, scaling factors).

2.6.2.1.6 Annotation Data Set "Tie Points Location and corresponding Auxiliary Data"

The annotation data set is composed of one Annotation Data Set Record (ADSR) for every 16 (Reduced Resolution) or 64 (Full Resolution) product frame (time sample), plus one at the last product frame. This leads to 925 ADSR per orbital product in Reduced Resolution (RR) and 36 ADSR per scene product in Full Resolution (FR), or 19 per FR imagette.

Each ADSR is composed of:

· MJD, modified Julian Day of time sample

· attachment flag: set when the MDSR corresponding to the ADSR are present in the product

· one annotation set for every tie point: 71 in RR, 36 in FR scene, 19 in FR imagette.

An annotation set includes:

1. tie point longitude

2. tie point latitude

3. tie point altitude

4. tie point surface roughness parameter

5. tie point longitude correction due to altitude

6. tie point latitude correction due to altitude

7. tie point Sun zenith angle

8. tie point Sun azimuth angle

9. tie point viewing zenith angle

10. tie point viewing azimuth angle

all the above quantities from Geolocation Processing 2.6.1.5.

11. ECMWF zonal wind components

12. ECMWF meridional wind components

13. ECMWF pressure

14. ECMWF total ozone

15. ECMWF relative humidity

all the above quantities from External Data Assimilation 2.6.1.7.

Note: for all tie points with a negative altitude, fields 4 to 6 are forced to zero.

2.6.2.1.7 Annotation Data Set "Product Quality"

The annotation data set is composed of one Annotation Data Set Records (ADSR) for every 128 (reduced resolution) or 512 (full resolution) product line, i.e. every 8 tie frames. This leads to 114 ADSR per orbital product in Reduced Resolution (RR) and 5 ADSR per scene product in Full Resolution (FR).

Each ADSR is composed of:

· MJD, modified Julian Day of time sample

· attachment flag

· one "out of range" flag register for the image pixels

· one "out of range" flag register for the blank pixels

An "out of range" flag register is composed of one flag per band and per MERIS module. A given flag is set if the number of "out of range»Image or blank band samples for the given module in the region between this Quality Annotation Frame and the next one (or the product end) is above a given threshold (in %). Specific thresholds are used for image pixels and blank pixels.

Note: both "out of range" PCDs are actually linked with MERIS frames instead of Level 1b product's ones. The alignment of the Quality Annotations with the latter is equivalent to a zero along-track depointing assumption.

2.6.2.1.8 Measurement Data Sets

There are 16 MDS, 15 for the radiances of the 15 MERIS bands and 1 for the associated flags, with the same record structure : an MDS is composed of one Measurement Data Set Record (MDSR) by product time sample.

The radiance MDSR contains:

· MJD, modified Julian Day of time sample

· quality flag: set to 0 when all data in the MDSR are invalid.

· one (scaled) radiance value per pixel (1121 in RR, 2241 in FR, 1153 in FR imagette).

Radiances are expressed in counts using the scaling factor stored in the SPH. Each value is stored in a two bytes unsigned integer.

The flag MDSR contains:

· MJD, modified Julian Day of time sample

· quality flag: set to 0 when all data in the MDSR are invalid.

· one flag set (one byte) per pixel (1121 in RR, 2241 in FR, 1153 in FR imagette).

The flag set contains 8 binary values meaning:

Table 2.8 - Flag set.

Each value is coded on 1 bit of the same byte, from least significant bit for flag 1 to most significant bit for flag 8.

The "land/ocean", "bright" and "coastline" flags are direct inputs from Pixel Classification; the "duplicate" flag is a direct input of the Radiance Resampling; the "glint" flag is a direct input from Geolocation; they are stored without further processing and do not need new definitions.

The "invalid" flag is a direct input, logically recombined with other flags, in order to gather all pixels satisfying any one of the following conditions:

· samples of all bands are saturated;

· out-of-swath product pixels;

· pixels added at the end of the product to reach the last tie frame;

· pixels added to fill a transmission gap of more than sixteen packets.

The "cosmetic" flag coming from the processing chain is a per band flag; the "suspect" flag is a new flag gathering pixels with diverse internal flags configurations; they are defined below:

· are considered "cosmetic" those pixels for which at least one radiance sample has been replaced by interpolation from neighbours;

· are considered "suspect" those pixels satisfying one of the following conditions:

- for any pixel, if it is flagged "stray light risk";

- for a "clear sky" and "ocean" pixel, at least one of the radiance samples is "saturated" or "dubious";

- for a "clear sky" and "land" pixel, at least one of the radiance samples of the bands dedicated to "land»Is "saturated" or "dubious" (list of "land" bands a processing parameter).

2.6.2.2 Level 1b Accuracies

Accuracies for Level 1b products are given in the table below.

Table 2.9 - Accuracies for Level 1b products.

2.6.2.3 Level 1b Engineering Quantities

The MERIS Level 1b product is given in radiance units [LU] after scaling.

The in-band reference irradiances for the 15 MERIS bands are computed by averaging the in-band solar irradiance of each pixel. The in-band solar irradiance of each pixel is computed by integrating the reference solar spectrum with the band-pass of each pixel.

The 15 Sun spectral flux values provided in the "GADS Scaling" record of the Level 1b products are the in-band reference irradiances adjusted for the Earth-Sun distance at the time of measurement.

The band-pass of each pixel is derived from on-ground and in-flight characterisation via an instrument model.

The values "Band wavelength" and "Bandwidth" provided in the SPH of the Level 1b products are the averaged band-pass of each pixel over the instrument field of view. The centre wavelength of each pixel as characterised on ground is shown in figure 2.43 below for a sample CCD row pertaining to band 11. Band 11 is dedicated to the observation of oxygen absorption and an accurate knowledge of its central wavelength is crucial for the retrieval of atmospheric pressure.

The detailed information can be found in document R-10. This workbook contains 9 data sheets:

1. the central wavelengths of each FR pixel for each band (pixel 1 is East, pixel 3700 is west)
2. corresponding in-band irradiances (derived using above wavelengths, computed instrument response functions and reference irradiances from sheet 4, scaled to Sun-Earth distance for MERIS reference day of year 95)
3. FWHM (full width at half modulation) of the instrument response functions for each band of each camera (each camera has 740 FR pixels – there is no FWHM variation for a given band within a given camera FOV -; camera 1 is east, camera 5 is west)
4. sheet 4 contains the reference Irradiance at 1 AU adopted for ENVISAT (Thuillier, G., M. Hersé, P. C. Simon, D. Labs, H. Mandel, D. Gillotay, and T. Foujols, The solar spectral irradiance from 200 to 2400 nm as measured by the SOLSPEC spectrometer from the ATLAS 1-2-3 and EURECA missions, Solar Phys., submitted, 2002.)
5. sheets 5 to 9 hold the instrument response functions for cameras 1 to 5 (hence 5 sheets), for each band, computed at centre of FOV (between pixels 370 & 371). Each function is scaled to max=1.

All wavelengths and widths are expressed in nm, irradiances in mW.m^-2.nm^-1.

Figure 2.7 - Sample CCD row for band 11.

Corrections

The instrument response can be computed by including, or not, a non-linearity and stray light corrections.

The non-linearity correction consists of transforming a 16-bit telemetry word into a floating point value before any offset and smear corrections (see CCD output amplifier. With the correction switched on, both the calibration and measurement signals are corrected before the instrument gains are computed or applied, respectively.

The stray light correction consists of the de-convolution of the signal with a stray light kernel. The kernels are shift invariant across-track and include for each spectral band a contribution from different spectral regions of the CCD. (See Stray Light Correction Algorithm 2.6.1.2.4. .)

With the correction switched on, both the calibration and measurement signals are corrected after the instrument gains are computed/applied. During calibration processing, an iterative process is used to return the instrument gains, corrected for stray light.

2.6.2.4 Level 1b Essential Product Confidence Data

Flags are provided on a pixel-by-pixel basis and are listed below:

· Bright clouds - based on a radiometric test in the blue part of the spectrum.

· Coastline - based on an a priori atlas.

· Land/sea - based on an a priori atlas.

· Sun glint risk - determined by the solar elevation angle at tie points.

· Invalid.

· Suspect - due to transmission errors.

· Cosmetic - filling of defective detection elements.

· Duplicated - due to resampling using the nearest-neighbour method.

2.6.2.5 Browse Products

Quicklooks are enhanced RGB images created from MERIS level 1b swath products. The enhancement is obtained using a local histogram equalisation algorithm:

for each pixel of the swath, an histogram equalisation is done considering only surrounding pixels (a square window centred on the current pixel). The continuity of the output image is obtained from the overlapping of the window considered for each pixel.

To obtain an enhance RGB both on water and lands (using the level 1b flag masks), different wavelength bands are respectively used:

- water pixels are handle in a way main oceanic structure are usually visible (ratio of bands 560/442) and displayed as blue tint,

- for lands and clouds is applied a combination of bands to create a pseudo RGB image.

The output image is the merge of the two previous complementary images. The result provide a very good enhancement for each image. The drawback are the discontinuities which are created by the two algorithms, especially for invalid cloud flags or when clouds over water become transparent. Note that due to the definition of the MERIS flags, there is no overlap between the water and land/cloud maps.