Minimize Notations

Collocation grid: this is a correspondence grid established using only the Level-1B orthorectified geolocation information.

SLSTR along track view is also called oblique view.

Solar channels: channels with central wavelength lower than 3.0 µm (SLSTR S1 to S6 and all OLCI channels).

Thermal channels: channels with central wavelength larger than 3.0 µm (SLSTR S7 to S9 and F1, F2).

Visible radiation: electromagnetic radiation detectable by the human eye with a wavelength between approximately 400 nm and 700 nm (OLCI Oa1 to Oa11 and SLSTR S1 and S2 channels).

Infra-red (IR) radiation: electromagnetic radiation of wavelengths between approximately 750 nm and 1 mm. This is broken down into five wavelength regions:

  • Near-IR - 0.75-1.4 µm (OLCI Oa12 to Oa21, SLSTR S3 and S4 channels)
  • Short-Wave IR - 1.4-3 µm (SLSTR S5 and S6 channels)
  • Medium-Wave IR - 3-8 µm (SLSTR S7 and F1 channels)
  • Long-Wave IR - 8–15 µm (SLSTR S8, S9 and F2 channels).

SWIR sub-bands: for any SLSTR SWIR channel (S4 to S6) the two-column detectors acquire simultaneously two images called the "A" and "B" sub-bands (or stripes). Furthermore the SLSTR Level-1B processing computes a third image from these two images which is called the "TDI" (or "averaged") sub-band. In this document the 1 km channels are sometimes referred to as 1 km sub-bands for simplification.

Ground Control Point (GCP) or correlation point: landmark, visible and located on two images, where local residual mis-registration between these images is estimated by a matching process.

Earth surface: the Earth surface is modelled as a Digital Elevation Model (DEM) (provided as CFI) on top of the WGS84 ellipsoid model.

(Direct) geolocation function: function that maps a point (k,j) (possibly non-integers) in an image to a point (x,y,z) on the ellipsoid surface. It is subtended by a model of the line of sight coming from point (k,j).

Inverse geolocation function: the inverse function of the direct geolocation function.

(Direct) ortho-rectified geolocation function: function that maps a point (k,j) (possibly non-integers) in an image to a point (x,y) on the Earth surface, by taking into account a DEM z = DEM(x,y). Theoretically (x,y,z) is the intersection of the line of sight coming from (k,j) with the Earth surface modelled as a DEM on top of a reference ellipsoid. The terrain point location is corrected from the relief effect, compared to the one computed with the direct geolocation function.

Inverse ortho-rectified geolocation function: the inverse function of the direct orthorectified geolocation function.

Restituted value: value retrieved when all known corrections have been applied.

(Mis-)knowledge error: residual error when all known corrections have been applied. The true value is given by adding the (unknown) (mis-)knowledge error to the restituted value.

Inter-channel spatial co-registration, simply referred to here as co-registration or mis-registration: The definition given in [AD03] is: maximum equivalent ground distance between the positions of all pairs of spatial samples acquired in two spectral channels and related to the same target on Earth.

Inter-instrument spatial mis-registration: mis-registration between one reference OLCI channel and one reference SLSTR channel.

Intra-instrument spatial mis-registration: mis-registration between all the channels within the same instrument.

Image sample/pixel: pixel stands for Picture Element. Each pixel is a measure of radiance generally gridded, with coordinates (k,j) in an image (k indexes the rows, j indexes the columns).

Instrument sample / instrument pixel / acquired pixel: (all equivalent terms). Pixels really acquired by an instrument, before any geometric transformation. An instrument sample may or may be not calibrated. When not calibrated, this is a synonym of a digital count.

Frame: the set of measurements acquired by the OLCI instrument at a given time

Coastal zone: sea surface extending from the coast up to 300 km offshore (from [AD-1])

Ancillary data: a classical definition is "all on-board data, other than observation and HKTM data, necessary for the products processing". This would include in particular not only various parameters and settings but also satellite data such as OBT and time correlations if needed, navigation data, etc.

Auxiliary data: We limit our understanding of auxiliary data to all complementary data provided to the Ground Segment (PDGS) by external providers in order to process the Level-1 and above products.

Product Data: Any data produced by the ground segment processing.

Search window: small window (grid) centred on a tie-point in OLCI geometry. It is a set of coordinates. It is used to extract a search imagette of SLSTR channel for correlation with the context window during the inter-instrument mis-registration estimation.

Context window: small window (grid) moved around the search window (along shift vectors) in OLCI channel geometry. It is a set of coordinates. It is used to extract a context imagette of OLCI channel for correlation with the search chip during the inter-instrument mis-registration estimation.

Context imagette: the radiometric counterpart of the context window, obtained by extracting the OLCI channel radiometry corresponding to the context window. If C is a context imagette W(C) represents the corresponding context window

Search imagette: the radiometric counterpart of the search window, obtained by resampling the SLSTR channel radiometry to the search window. If S is a search imagette W(S) represents the corresponding search window

Orbital Revolution Number: this number identifies the Sentinel-3 orbit within the orbital cycle. There are 385 orbits per cycle, making the Orbital Revolution Number between 0 and 384.

Orphan pixels (for SLSTR) or removed pixels (for OLCI): These are pixels acquired by the instruments but not retained in the Level-1B gridded image, due to the Level-1B (nearest neighbour) projection on the product grid. For OLCI, those pixels mainly come from overlapping areas between adjacent camera modules. For SLSTR they may come from a possible oversampled acquisition at nadir of the nadir-view, with respect to the Level-1B image gridding. In oblique view there are many orphans due to scan-to-scan along track overlap. To answer the Level-1 requirements, all those pixels are retained in Level-1B products but not gridded.

Scan: a scan is defined as a complete rotation of the SLSTR scan mirrors.

Instrument scan or scan trace: It is the trace of a single SLSTR detector element on the ground. For example in the thermal channels, each detector has two elements, and so a single scan will give two scan traces, displaced by 1 km in the along track direction. Adjacent scan traces represent adjacent 'rows' of the instrument grid.

Image scan: a line of pixels in the SLST Level-1B product. Note that in the Level-1 product an image scan and an instrument scan should refer to the same thing.

Deformation model: in the Level-1 processing this terms refers to the interpolation model applied on the (potentially) irregular grid of tie-points and representing the deformation field between OLCI and SLSTR in the OLCI geometry.

Correspondence grids: the main output of the Level-1 processing, stored in the mis-registration data files in the Level-1 product. These are grids that link any OLCI pixel in the reference band to the corresponding sub-pixel location in the other OLCI and SLSTR bands such that if a detector were placed at the sub-pixel location it would have seen the same target on Earth as the reference pixel.