Minimize SCIAMACHY Product Handbook

SCIAMACHY Operational Data Products and Algorithms

Table of Contents

 


CHAPTER 6

6: SCIAMACHY Operational Data Products and Algorithms


The operational data processing follows the specific guidelines and rules of the ENVISAT Payload Data Segment (PDS) architecture and organisation. Scientific product generation is based on science groups’ algorithms which have not been developed in the framework of the PDS.

The baseline for each higher level product is the level 1b product containing orbit wise geo-referenced measurements including their calibration and instrument monitoring data. Level 2 products are specifically dedicated to geophysical parameters including column densities, stratospheric profiles of atmospheric constituents and information about clouds and aerosols. Each day of SCIAMACHY operations generates 14 or 15 level 1b and level 2 products. With level 2, ENVISAT’s responsibility for data generation, dissemination and archiving ends. This does not however, as outlined above, exclude experienced science users from processing their own scientific products of the same levels. It is up to the individual user how to specify and process value added (VA) products. In the VA environment, geophysical parameters are often gridded on a global scale, i.e. these products are well suited to provide the interested public with SCIAMACHY’s view of the Earth atmosphere.

Basic processing related information is provided via the file names of the products, e.g.
SCI_NL__1PRDPA20061029_012353_000060442052_00275_24377_0673.N1
File names include the sequence


image

product ID: SCI_NL__0P, SCI_NL__1P or SCI_OL__2P for level 0, 1b or 2 data (SCI_NL__1P)

image

processing status flag: N for NRT (Near-realtime) products (this flag is no longer used, since we are in Phase F), letters between N and W for consolidated products (R, currently used Y)

image

originator ID: PDK, PDE, LRA, D-P for PDHS-K, PDHS-E, LRAC or D-PAC (DPA)

image

start date: year, month and day of measurement start (20061029)

image

start time: hours, minutes and seconds of measurement start in UTC (012353)

image

duration of product coverage in seconds (00006044)

image

ENVISAT mission phase (2)

image

ENVISAT cycle number (052)

image

relative orbit number within cycle (00275)

image

absolute orbit number at product start (24377)

image

processing file counter (0673)

image

file extension: N1 corresponds to ENVISAT


That permits an unambiguous allocation of the file to a particular measurement. All operational data products are subject to quality monitoring. The Payload Data Control Center (PDCC) does a processing control, while the main quality control is performed within ESA SPPA-IDEAS. Its goal is to screen all products at the time of generation in order to identify anomalies or deviations from expected results. Quality monitoring includes content and consistency checks, e.g. formal correctness of the product or parameter limits. In case of detected anomalies, data shall be flagged to initiate further actions. The SPPA executes quality monitoring activities on various timescales ranging from daily to multi-monthly.

6.1 Level 0 Products

A level 0 product contains all measurement data for one complete orbit in the case of consolidated products. The difference to the raw data as sent down to the ground station is that level 0 data are time ordered and already have data headers which describe the data. Usually users do not need level 0 products, since level 1b data contain the same data with the addition of the geolocation and calibration data. Level 0 data are not distributed operationally.

6.2 Operational Level 1b Products and Algorithms

6.2.1 Level 1b Processor


SCIAMACHY level 1 data products comprise geolocated and calibrated radiances of the scientific measurements, as well as measurements for calibration and instrument monitoring.


The algorithms used in operational level 0 to 1 processing are primarily driven by the scientific needs to convert measured signals into calibrated radiances (see following section). However, constraints imposed by the instrument operation, and in particular constraints imposed by the operational data processing environment, may force one to take different strategies as one would employ for a ground-based instrument. The wish to obtain a level 1 data product which is not excessively large and complicated imposes an additional constraint. The principle processing cycle starting with instrument level 0 data and ending with the level 2 product is outlined in figure 6-1. A major constraint from the ENVISAT PDS architecture rules is that there may be only one output product per processing chain, in this case the level 1b product. As a consequence, the level 1b product must not only hold processed science data, but also calibration measurements and instrument monitoring data, as well as newly calculated calibration parameters. The latter are collected in the Annotation Data Set (ADS) in contrast to the measurement data which can be found in the Measurement Data Sets (MDS). In order to keep the size of the product as small as possible, the individual dark signal calibration measurements are discarded after calculation of the dark signal calibration parameters. For each dark state only the average signal of all measurements together with the deviation derived from the averaging is kept. All other calibration and monitoring measurements are retained, albeit in unprocessed form.

 

click to enlarge

fig 6-1:

Overview of processing flow.

 


In the operational processing from level 0 to 1b, all necessary calibration constants for each science measurement are processed from the input calibration data, ground-based and in-flight as well. The level 1b data product contains the raw detector signals of the science measurements plus these calibration constants, mainly coded in 1-byte integers. In addition to measurement-specific calibration constants, e.g. stray-light or atmospheric polarization determination for each measurement, lookup-tables are generated for globally applicable calibration constants, e.g. instrument polarization sensitivity as function of scan angle. The calibration data files themselves are generated by the Level 0-1 processor using Level 0 data and key data as an input. The calibration data are saved into a database and written to the Level 1b product during processing. A calibration application tool (SciaL1c) is provided to the users supplementing the operational level 1b product such that the users can inflate level 1b products to level 1c containing fully calibrated data. For the user’s convenience certain calibrations can be optionally omitted at extraction or a sub-set of data can be filtered out. However, operational level 1c products are not generated by the PDS, they need to be generated by the users.
Fig. 6-2 provides an overview of the processing steps for level 0 to 1b and level 1b to 1c calibration of scientific measurements (nadir, limb and solar occultation), as well as the corresponding output on the level 1b and level 1c data products in a flow diagram. The individual calibration related steps are reflecting the calibration characteristics as described in the following sectionand specified in (Slijkhuis 2004). (fig. 6-2)

 

Table of Contents

 

0 Attachments
2629 Views
Average (0 Votes)
The average rating is 0.0 stars out of 5.
Comments