6. L2b Products Format Specification

6.1 Main Product header

The Main Product Header (MPH) is an XML file present in all the BIOMASS L2 products.

BIOMASS L2 products are folder based and the MPH is top level leaf of the product tree.

The MPH main scope is to provide all the information related to the:

• Genesis of the product;

• Identification of the product for example for cataloguing purposes;

• Product consistency (filesize).

For the detailed description of the MPH content and format refer to [AD3]. It is very similar to the one described in 4.1, without following fields, which are not present in L2b:

• eop:swathIdentifier Swath identifier

• eop:orbitDirection Orbit direction

• eop:wrsLongitudeGrid Track number

• eop:wrsLatitudeGrid Frame number

• sar:antennaLookDirection SAR antenna look direction

Instead, it has a new field “agbTileStatus” (see the “l2b only” comment reported also in 4.1), the field is under eop:metaDataProperty/bio:EarthObservationMetaData as:

eop:metaDataProperty Observation metadata

bio:EarthObservationMetaData

[…] (see 4.1)

bio:agbTileStatus Flag raised when a new iteration is needed (l2b only)

bio:agbTileIteration iteration index (integer value, starting form 1) (l2b only)

The field is a string, with following possible values ‘Complete’, “NotComplete’, ‘Partial’ [AD10]

6.2 Measurement data sets (MDS)

The MDS of the L2b products is constituted by following quantities:

• the dry weight of above ground biomass image binary raster data layer;

• a layer of quality indication for each of the above ground biomass image pixel;

• BPS_FNF: this is the annotation of Forest mask which defines, for each geographical pixel, the presence or absence of forest and which is used to discard invalid forest regions during aggregation of L2a in L2b products.

If the L2b FD product is provided as input, this MDS is the CFM pre-computed contained in L2b FD product, otherwise it is the FNF from L2a annotation LUT.

• a Heat Map annotation of integer numbers (ubyte) which stores, pixel wise, how many L1c images have been used to compute each pixel (this information is completed with the list of all L1c images used by the processor, see input information DSR 6.3.1.3).

• Acquisition ID Image: a multilayer image (COG format) where each contributing L2a input incidence map is stored (as a binary map); the number of layers is equal to the number of L2a inputs and their order is the same with respect to the L2a inputs ().

For all the images, following considerations are the same.

All the L2b image products are geocoded to specific BIOMASS tiling latitude-longitude geographic DGG, described in [RD9], ensuring a ground sampling always better than the resolution.

In particular, we have that L2b images are aggregations of the L2a products insisting on the same location on ground (a single tile of the above DGG); a different number of L2a products may be available for the MDS geographic grid tile, based on tile dimensions and swath coverage.

The format used to store the L2b images is again the Cloud Optimized GeoTIFF (COG) [RD1].

Details about COG format and its usage to store images are already explained in sec. 4.2 for L2a products and apply as well to L2b products;

Metadata

Metadata is the same of the L2a one, already described in sec. 4.2 and Tab.2. the only difference is the absence of “MATRIX_REPRESENTATION”, “PolarisationsSequence”, “Swath”, and “POLARIMETRIC_INTERP” metadata in L2b.

Overviews

Same considerations already explained in 4.2

6.3 Annotation data sets (ADS)

6.3.1 Main ADS

This section describes in detail the content of main ADS for L2b products. Differences from L2b products are in the description, where needed.

An overview is provided by the two figures below and by the table which describes all the elements, both for L2b.

Fig. 19 — main L2b ADS

Table 41 — Main L2b ADS

Name	Description	Data Type	Cardinality
product	Product L2b DSR. This DSR contains the L2b product information.	productL2bType	1
rasterImage	Raster image DSR. This DSR contains all the necessary information to exploit the L2b raster images.	rasterImageType	1
InputInformation	Input Information L2b DSR. This DSR contains the necessary information to identify the L2a products in input to L2b processor and also a list of all the L1c acquisitions used to generate those L2a products.	InputInformationL2bL3ListType	1
processingParameters	Processing parameters DSR. This DSR contains the description of L2b processing parameters.	processingParametersL2bType	1

6.3.1.1 Product L2b DSR

This DSR contains the parameters to describe the L2b product.

An overview is provided by figures and table below.

Fig. 20 — product L2b DSR

Table 42 — Product L2b DSR (productL2Type)

Name	Description	Data Type	Cardinality
mission	Mission (BIOMASS).	missionType	1
tileID	DGG geographic cell (tile) identifier (one element list).	xsd:string	1
basinID	List of DGG basin identifiers.	stringListType	1
productType	Product type (AGB_L2B).	productType	1
startTime	Start time of the image [UTC].	timeType	1
stopTime	Stop time of the image [UTC].	timeType	1
radarCarrierFrequency	Radar carrier frequency [Hz].	doubleWithUnit	1
missionPhaseID	Mission phase identifier.	missionPhaseIDType	1
sensorMode	Sensor mode (always Measurement).	sensorModeType	1
globalCoverageID	Global coverage identifier.	xsd:unsignedShort	1
contributingTiles	List of all the DGG tiles in the input neighbourhood, contributing for AGB parameter estimation (5x5 neighbourhood). Mandatory only in L2B AGB products.	stringListType	0,1
calABcoveragePerTile	CAL_AB coverage per tile [% and absolute numbers], for each of the tiles in the input neighbourhood for AGB parameter estimation (5x5 neighbourhood). Mandatory only in L2B AGB products.	calABcoverageTilesListType	0,1
GNcoveragePerTile	Input GN products coverage per tile [% and absolute numbers], for each of the tiles in the input neighbourhood for AGB parameter estimation (5x5 neighbourhood). Mandatory only in L2B AGB products.	GNcoverageTilesListType	0,1

Table 43 — Product L2b DSR (calABcoverageTilesListType)

Name	Description	Data Type	Cardinality
ID	Tile ID	xsd:string	1
tile	CAL_AB coverage per tile [% and absolute numbers], for each of the tiles in the input neighbourhood for AGB parameter estimation	calABcoverageType	0…25

Table 44 — Product L2b DSR (calABcoverageType)

Name	Description	Data Type	Cardinality
originalCoverage	CAL_AB original coverage, percentage and number of pixel values	percentPixelsType	1
filteredCoverage	CAL_AB filtered coverage, percentage and number of pixel values after each filtering. [RD] BPS_AGB_ATBD v3.1.2, equation 4.6	calABfilteredCoverageType	1

Table 45 — Product L2b DSR (calABfilteredCoverageType)

Name	Description	Data Type	Cardinality
afterGlobalFiltering	CAL_AB filtered coverage, percentage and number of pixel values after all filtering concatenated.[RD] BPS_AGB_ATBD v3.1.2, equation 4.6	percentPixelsType	1
afterAGBvalueFiltering	CAL_AB filtered coverage, disaggregated percentage and number of pixel values after filtering only for mean AGB values limits. [RD] BPS_AGB_ATBD v3.1.2, equation 4.6	percentPixelsType	1
afterAGBstdFiltering	CAL_AB filtered coverage, disaggregated percentage and number of pixel values after filtering only for AGB STD limits. [RD] BPS_AGB_ATBD v3.1.2, equation 4.6	percentPixelsType	1
afterAGBrelativeStdFiltering	CAL_AB filtered coverage, disaggregated percentage and number of pixel values after filtering only for AGB relative STD limits. [RD] BPS_AGB_ATBD v3.1.2, equation 4.6	percentPixelsType	1
afterLCMClassFiltering	CAL_AB filtered coverage, disaggregated percentage and number of pixel values after filtering only for LCM class. [RD] BPS_AGB_ATBD v3.1.2, equation 4.6,	percentPixelsType	1

Table 46 — Product L2b DSR (GNcoverageTilesListType)

Name	Description	Data Type	Cardinality
ID	Tile ID	xsd:string	1
tile	Input GN products coverage, for each tile	GNcoverageListType	0…25

Table 47 — Product L2b DSR (GNcoverageListType)

Name	Description	Data Type	Cardinality
acquisitionID	GN product acquisition ID	xsd:string	1
GN	Input GN product coverage, for each input GN (unbounded) and current tile	GNcoverageType	1 unbounded

Table 48 — Product L2b DSR (GNcoverageType)

Name	Description	Data Type	Cardinality
originalCoverage	Input GN product original coverage (current GN and current tile) percentage and number of pixel values	percentPixelsType	1
filteredCoverage	Input GN product filtered (current GN and current tile) percentage and number of pixel values after each filtering.[RD] BPS_AGB_ATBD v3.1.2, equation 4.5	GNfilteredCoverageType	1

Table 49 — Product L2b DSR (GNfilteredCoverageType)

Name	Description	Data Type	Cardinality
afterGlobalFiltering	Input GN product filtered coverage, percentage and number of pixel values after all filtering concatenate.[RD] BPS_AGB_ATBD v3.1.2, equation 4.5	percentPixelsType	1
afterSigmaFiltering	Input GN product filtered coverage, disaggregated percentage and number of pixel values after filtering only for GN Sigma (backscatter) values limits. [RD] BPS_AGB_ATBD v3.1.2, equation 4.5, σ_p^0	percentPixelsType	1
afterAngleFiltering	Input GN product filtered coverage, disaggregated percentage and number of pixel values after filtering only for GN Incidence Angle values limits. [RD] BPS_AGB_ATBD v3.1.2, equation 4.5, ϑ	percentPixelsType	1

Table 50 — Product L2b DSR (percentPixelType)

Name	Description	Data Type	Cardinality
percentage	Coverage in percentage of the whole Tile	xsd:float	1
pixels	Coverage in number of pixels of the whole Tile	xsd:unsignedInt	1

6.3.1.2 Raster Image L2b DSR

This DSR contains all the necessary information to exploit the raster images of all the L2b MDS, it is already described in Fig. 4, Tab.10, Tab.11, Tab.12, Tab.13, the differences are:

• the projection which here is the specific “DGG” described in [RD9].

• Tab.11 is replaced by Tab.51 for the L2b AGB choice

Table 51 — pixelRepresentationChoiceType (only L2b AGB choice described here)

Name	Description	Data Type	Cardinality
AGB	Representation of the image pixels within the image MDS: Above Ground Biomass [t/ha]	pixelRepresentationType	1
AGB_STANDARD_DEVIATION	Representation of the image pixels within the image MDS: Above Ground Biomass quality [t/ha]	pixelRepresentationType	1
AGB_heatMap	Representation of the image pixels within the image MDS: Heat Map	pixelRepresentationType	1
BPS_FNF	Representation of the image pixels within the image MDS: BPS_FNFN	pixelRepresentationType	1
Acquisition_id_image	Representation of the image pixels within the image MDS: Acquisition ID image	pixelRepresentationType	1

6.3.1.3 Input Information L2b DSR

This DSR contains, for each of the L2a Product in input to L2b processor:

• the folder name of L2a product;

• the L1c acquisitions used to generate that L2a product; it is an InputInformationL2aType, see description in 4.3.1.3 and Tab.14, Tab.15.

An overview is provided by figure below and in the table below for the L2inputs.

Fig. 21 — Input information L2b DSR (InputInformationL2bL3ListType)

Table 52 — Input information L2b DSR (InputInformationL2bL3ListType )

Name	Description	Data Type	Cardinality
count	XML attribute.	xsd:unsignedInt	required
L2aInputs	One entry for each L2a product in input to L2b. Each entry contains two elements: L2aProductFolderName and L1inputs.	L2aInputsType	1…N

Table 53 — Input information L2b DSR (L2aInputsType)

Name	Description	Data Type	Cardinality
L2aProductFolderName	Folder name of each L2a product, as defined in sec. 3.2 (One string for each “L2aInputs” entry).	xsd:string	1
L2aProductDate	Product date for each L2a product	xsd:string	1
L1inputs	L1c acquisitions used to generate the L2a product. See 4.3.1.3 and Tab.14, Tab.15, Tab.20. (One element for each “L2aInputs” entry).	InputInformationL2aType	1

6.3.1.4 Processing parameters L2b DSR

This DSR contains the exhaustive list of L2b processing parameters applied. An overview is provided by figures and table below.

Fig. 22 — processing parameters L2b DSR

Table 54 — Processing parameters L2b DSR

Name	Description	Data Type	Cardinality
processorVersion	Version of the processor used to generate the product.	xsd:string	1
productGenerationTime	Product generation time [UTC].	timeType	1
forestMaskingFlag	True if Forest/Non-Forest map masking has been performed during L2a products merging, false otherwise.	bool	1
minimumL2aCoverage	The minimum percentage of tile coverage used to trigger the computation.	xsd:float	1
rejectedLandcoverClasses	Set of landcover class indices ignored	intArray	1
backscatterLimits	Lower and upper limits on backscatter in linear units. Default values are 0.0001,100 for all polarisations)	BackscatterLimitsType	1
angleLimits	Used lower and upper limits on the local incidence angle [rad].	minMaxTypeWithUnit	1
meanAGBLimits	Used lower and upper limits on the AGB mean in [t/ha].	minMaxTypeWithUnit	1
stdAGBLimits	Used lower and upper limits on the AGB std in [t/ha]	minMaxTypeWithUnit	1
relativeAGBLimits	Used lower and upper limits on AGB standard deviation relative AGB mean (coefficient of variability), in linear units.	minMaxType	1
referenceSelection	Selection of reference data (second iteration only). String with following possible values: ‘refOnly’, ‘firstIterationOnly’, ‘weightedMean’	referenceSelectionType	1
indexingL	Used “L” parameter variability with polarization ’p’, date ’j’, forest class ’k’	agbIndexingType	1
indexingA	Used “a” parameter variability with polarization ’p’, date ’j’, forest class ’k’.	agbIndexingType	1
indexingN	Used “n” parameter variability with polarization ’p’, date ’j’, forest class ’k’.	agbIndexingType	1
useConstantN	If True, indexingN=p has been forced (for all polarisations).	bool	1
valuesConstantN	Values used if useConstantN=True.	floatArray	1
regressionSolver	Computational complexity used for the regression: “double” “float”	xsd:string	1
regressionMatrixSubsamplingFactor	Subsampling factor used during AGB regression, computed live basing on available RAM remaining (starting from job order Amount_of_RAM)	xsd:unsignedByte	1
minimumPercentageOfFillableVoids	The minimum percentage of invalid pixels which triggered a new iteration.	xsd:float	1
estimatedParameters	L2B AGB algorithm estimated inversion parameters “N”, “A”, “L”: the parameters can have variability along polarisations, LCM classes and temporal dates.	estimatedParametersL2bAGBType	1
compressionOptions	Configurable compression options for all the L2b MDS COG.	compressionOptionsL2bType	1

Table 55 — BackscatterLimitsType L2b DSR

Name	Description	Data Type	Cardinality
BackscatterLimitsType	Used lower and upper limits on backscatter in linear units: three enumerates elements “HH”, “VH”, “VV”, each of MinMaxType	MinMaxType	3…3

Fig. 23 — Processinc parameters L2b DSR (estimatedParametersL2bAGBType)

Table 56 — Processing parameters L2b DSR (estimatedParametersL2bAGBType)

Name	Description	Data Type	Cardinality
rho	AGB estimation Rho value: logarithmic bias correction as the ratio of the average reference AGB to the average estimated AGB	xsd:float	1
polarisations	Attribute of beow described “Polarisation” element; possible values: “HH”, “VH”, “VV”, “HH VH VV”	xsd:string	1
Polarisation	Parameters estimated during AGB processing	estimatedParametersPolarisationType	1 or 3

Note on estimatedParametersPolarisationType

It is a unique complex type (see Fig. 23), containing an xsd:sequence of “LCM” elements, which contains an xsd:sequence of “date” elements, which contains xsd:all “N”, “A”, and “L” elements, each of them each composed by the final xsd:sequence with “mean” and “std” elementary “xsd:float” elements)

Table 57 — Processing parameters L2b DSR (estimatedParametersPolarisationType )

Name	Description	Data Type	Cardinality
classes	String with integers separated by spaces, indicating group of LCM classes, i.e. “1 2 3”	xsd:string	required
LCM	Sequence of “LCM” elements. Each LCM in the sequence collects, for each date, the “N”, “A”, “L” estimated parameters variability for the LCM classes specified in the attribute.	xsd:complexType (LCM)	1…unbounded

Table 58 — Processing parameters L2b DSR (xsd:complexType (LCM))

Name	Description	Data Type	Cardinality
dates	String containing two dates (in the timeType format), separated by a space: start and stop date to describe an interval.	xsd:string	required
date	Sequence of “date” elements. Each “date” collects the “N”, “A”, “L” estimated parameters variability inside the dates interval specified in the attribute.	xsd:complexType (date)	1…unbounded

Table 59 — Processing parameters L2b DSR ((sd:complexType (date))

Name	Description	Data Type	Cardinality
N	AGB inversion algorithm estimated parameter “N” for a specific polarisation group, LCM classes group and specified dates interval.	xsd:complexType (N, A, L)	1
A	AGB inversion algorithm estimated parameter “A” for a specific polarisation group, LCM classes group and specified dates interval.	xsd:complexType (N, A, L)	1
L	AGB inversion algorithm estimated parameter “L” for a specific polarisation group, LCM classes group and specified dates interval.	xsd:complexType (N, A, L)	1

Table 60 — Processing parameters L2b DSR (xsd:complexType (N, A, L))

Name	Description	Data Type	Cardinality
mean	Mean value for the estimated regression parameter	xsd:float	1
std	Standard deviation value for the estimated regression parameter	xsd:float	1

Fig. 24 — Processing parameters L2b DSR (compressionOptionsL2bType)

Table 61 — Processing parameters L2b DSR (compressionOptionsL2bType)

Name	Description	Data Type	Cardinality
MDS à AGB à compressionFactor MDS à AGBstandardDeviation à compressionFactor MDS àBPS_FNF à compressionFactor MDS à HeatMap à compressionFactor MDS à acquisitionIdImage à compressionFactor	ZSTD algorithm compression factors used for all the images MDS. From 0 to 9, where zero means no compression.	xsd:unsignedByte	1
MDS à AGB à MAX_Z_ERROR MDS à AGBstandardDeviation à MAX_Z_ERROR MDS à HeatMap à MAX_Z_ERROR	Used absolute maximum error admitted (which defines exactly how lossy has been the LERC compression algorithm). Zero means loss-less compression.	xsd:float	1
MDS_blockSize	MDS COG blocking algorithm size. Same value is used for both data array dimensions.	xsd:unsignedByte	1

6.3.1.5 Annotation LUT DSR

This has no difference respect to the one already described in 4.3.2.

6.3.1 Quick-look ADS

The Quick-look ADS is composed by lower resolution versions of all the MDS images.

The details about Quick-look ADS already explained in sec. 4.3.1 for L2a products apply as well to L2b products.

6.3.2 Overlay ADS

The Overlay ADS is a Keyhole Markup Language (KML) file that describes the product coverage area and is suitable for viewing the Quick-look ADS in any application that supports KML format (like, e.g., Google Earth or GIS applications).

The details about Quick-look ADS already explained in sec. 4.3.2 for L2a products apply as well to L2b products.