4. GOME-2 Products Overview
Table of Contents

4. GOME-2 Products Overview

4.1 The GOME-2 instrument

4.1.1 Instrument hardware

GOME-2 is a medium-resolution double UV-VIS spectrometer, fed by a scan mirror which enables across-track scanning in nadir, as well as sideways viewing for polar coverage and instrument characterisation measurements using the moon. The scan mirror directs light into a telescope, designed to match the field of view of the instrument to the dimensions of the entrance slit. This scan mirror can also be directed towards internal calibration sources or towards a diffuser plate for calibration measurements using the sun (see Figure 4.1).

GOME-2 comprises four main optical channels which focus the spectrum onto linear silicon photodiode detector arrays of 1024 pixels each, and two Polarisation Measurement Devices (PMDs) containing the same type of arrays for measurement of linearly polarised intensity in two perpendicular directions.

The four main channel detectors are actively cooled in a closed loop configuration to -38°C to maximise sensitivity and minimise noise. In contrast the two PMD detectors are cooled in an open loop configuration to an operating temperature of around 0°.

The PMDs are required because GOME-2 is a polarisation sensitive instrument and therefore the intensity calibration must take account of the polarisation state of the incoming light. This is achieved using information from the PMDs.

    Figure 4.1: Artists impression of the GOME-2 optical layout (courtesy of ESA)

    1 - Disperser 10 - Beam splitter 19 - Channel # 2
    2 - Calibration Slit 11 - Channel # 3 20 - Grating # 1
    3 - Detector 12 - Channel # 4 21 - Grating # 2
    4 - Double Brewster Prism 13 - 590 -790 nm 22 - Calibration lamp
    5 - Telescope mirror 14 - 401 - 600 nm 23 - Calibration Unit
    6 - Predisperser Prism 15 - 240 - 315 nm 24 - Sun diffuser
    7 - Channel Separator 16 - 311 - 403 nm 25 - Telescope mirrors
    8 - Grating # 3 17 - Electronics box 26 - Scan mirror
    9 - Grating # 4 18 - Channel # 1  

Light is collimated by an off-axis parabolic mirror, behind the entrance slit, onto the double Brewster and pre-disperser prisms which generate the s- and p- polarised beams. These beams are subsequently dispersed onto detectors contained within the Polarisation Unit (PU).

    Figure 4.2: The GOME-2 Polarisation Unit (PU) detailed optics (courtesy of ESA)

Light passing through the pre-disperser prism is also directed onto the main spectrometer. A second off-axis parabolic mirror focuses the dispersed beam onto the channel separator prism. This is a quartz prism, the first surface of which is partially coated with a transmission coating for channel 1 and a reflective coating for channel 2. The light for channels 3 and 4 passes the prism edge and a dichroic filter separates it into the two channels.

The four main channels provide continuous spectral coverage of the wavelengths between 240 and 790 nm with a spectral resolution (FWHM) between 0.25 nm and 0.5 nm. Compared to the main channels, the PMD measurements are performed at lower spectral resolution, but at higher spatial resolution. The two PMD channels are designed such that maximum similarity in their optical properties is ensured. The wavelength-dependent dispersion of the prisms causes a much higher spectral resolution in the ultraviolet than in the red part of the spectrum.

In order to calculate the transmission of the atmosphere, which contains the relevant information on trace gas concentration, the solar radiation incident on the atmosphere must be known. For this measurement a solar viewing port is located on the flight-direction side of the instrument. When this port is opened, sunlight is directed via a ~40° incidence mirror to a diffuser plate. Light scattered from this plate, or in general, light from other calibration sources such as the Spectral Light Source (SLS or HCL) for wavelength calibration, and the White Light Source (WLS) for etalon (and, optionally, pixel-to-pixel gain) calibration are directed to the scan mirror using auxiliary optics. Diffuser reflectivity can be monitored internally using light from the SLS. All internal calibration sources with their optics are assembled in a subsystem called the 'Calibration Unit' (CU). The only exception are light emitting diodes (LEDs) which are located in front of the detectors to monitor the pixel-to-pixel gain. For more information on the GOME-2 instrument see [SCD1].

    Figure 4.3: The GOME-2 calibration unit (courtesy of ESA)

4.1.2 Data packet structure and basic instrument operation

Every 375 milliseconds, GOME-2 generates one science data packet. A data packet comprises 9369 2-byte words, leading to an average data rate of 8x2x9369/0.375 bit/s = 400 kbit/s. A detailed description of the science data packet format is provided in [RD9]. Briefly, a GOME-2 data packet consists of three basic parts (apart from header information): instrument housekeeping (HK) data (e.g., temperatures, scan mirror angles, lamp currents and voltages), PMD data, and main channel Focal Plane Assembly (FPA) data. The maximum temporal resolution differs between main channel FPA and PMD data. One data packet contains up to two main FPA readouts, corresponding to a 187.5 ms temporal resolution, and up to 16 PMD readouts, corresponding to a 23.4 ms temporal resolution. A detailed description of the options for PMD readout and data transfer is given in Appendix B of [RD7] GOME-2 Level 1 Product Generation Specification.

A basic concept in the operation of the GOME-2 instrument is that of the 'scan'. A scan is defined as a time interval of 6 seconds, consisting of 16 'subsets' of 375 ms each, equivalent to one data packet. The subsets are numbered from 0 to 15. In the Earth scanning mode, a scan consists of one scan cycle: 4.5 s forward scan (subsets 0 to 11) and 1.5 s flyback (subsets 12 to 15). In the static and calibration modes the scan mirror does not move, but the data packet structure is identical to the scanning mode.

In the default measuring mode, the nadir scan, the scan mirror sweeps in 4.5 seconds (12 subsets) from negative to positive viewing angles, followed by a flyback of 1.5 seconds (the last 4 subsets) back to negative viewing angles as shown in Figure 4.4.

     

    Figure 4.4: The GOME-2 scan pattern in the default measuring mode

    Solid line: forward scan; dashed line: flyback.
    Each subset pixel (0-15) corresponds to 375 milliseconds. In one of the four subsets of the flyback (subset 14 is shown as an example) the 'unused' parts of the PMD detectors (i.e. Block A - see Appendix B of [RD7]) are read out.
    Note that the figure is not drawn to scale.

The default swath width of the scan is 1920 km which enables global coverage of the Earth's surface within 1.5 days (note that other swath widths are also commandable). The scan mirror speed can be adjusted such that, despite the projection effect, the ground is scanned at constant speed. The along-track dimension of the instantaneous field-of-view (IFOV) is ~40 km which is matched with the spacecraft velocity, such that each scan closely follows the ground coverage of the previous one. The IFOV across-track dimension is ~4 km. For the 1920 km swath, the maximum temporal resolution of 187.5 ms for the main channels (23.4 ms for the PMD channels) corresponds to a maximum ground pixel resolution (across track x along track) of 80 km x 4 0 km (10 km x 40 km for the PMDs) in the forward scan.

The actual integration time used (and thus the ground pixel size) will depend on the light intensity. The integration time can be separately set for each channel; in channels 1 and 2 it is even possible to subdivide each channel into two parts (called 'band 1a', 'band 1b' and 'band 2a', 'band 2b' respectively) having separate integration times. It is anticipated that a default integration time of 187.5 ms (yielding two spectrum readouts per data packet) will be used in all channels but with two exceptions where longer integration times are needed because of low light intensity:
(i) Band 1a has a default integration time of 1.5 seconds (yielding three spectra per scan and one from the fly-back with the possibility of co-adding spectra to improve signal to noise characteristics).
(ii) The integration time for all channels will be increased for low solar elevations (high solar zenith angles).

4.1.3 Observation modes

This section gives a classification of the GOME-2 observation modes. The observation modes can be assigned to three categories: Earth observation modes, calibration modes, and other modes.

The observation mode is derived in the data processing chain by combining fields from the data packet, such as scan mirror position, subsystem status flags, etc. There is no dedicated field in the data packet indicating the observation mode. Any GOME-2 data packet which does not fit into one of the modes below will be classified as "invalid" by the Level 0 to 1 data processor.

4.1.3.1 Earth observation modes

Earth observation (or "earthshine") modes are those modes where the Earth is in the field of view of GOME-2. They are usually employed on the dayside of the Earth (sunlit part of the orbit). The scan mirror can be at a fixed position (static modes), or scanning around a certain position (scanning modes). All internal light sources are switched off and the solar port of the calibration unit is closed.

Nadir scanning

This is the mode in which GOME-2 will be operated most of the time. The scan mirror performs a nadir swath as described above. The swath width is commandable, its default value is 1920 km. Scanning can be performed either with constant ground speed, resulting in equally sized ground pixels (this is the default), or with constant angular speed ("GOME-1 mode"), resulting in larger ground pixels for the extreme swath positions as compared to the swath centre.

North polar scanning

The scan mirror performs a swath around the viewing angle +46.696º (default value) in order to cover the North Pole which would not be observable with the normal nadir scanning mode. This mode will typically be used during northern hemisphere spring.

South polar scanning

The scan mirror performs a swath around the viewing angle -46.172º (default value) in order to cover the South Pole which would not be observable with the normal nadir scanning mode. This mode will typically be used during southern hemisphere spring.

Other scanning

The scan mirror performs a swath around another off-nadir position.

Nadir static

The scan mirror is pointing towards nadir. This mode will typically be used during the monthly calibration. It is valuable for validation and long-loop sensor performance monitoring purposes.

Other static

The scan mirror is pointing towards an off-nadir position.

4.1.3.2 Calibration modes

In-orbit instrument calibration and characterisation data are acquired in the various calibration modes. They are usually employed during eclipse with the exception of the solar calibration which is performed at sunrise. Both internal (WLS, SLS, LED) and external (sun, moon) light sources can be employed. The various sources are selected by the scan mirror position.

Dark

The scan mirror points towards the GOME-2 telescope. All internal light sources are switched off and the solar port is closed. Dark signals are typically measured every orbit during eclipse.

Sun (over diffuser)

The scan mirror points towards the diffuser. All internal light sources are switched off and the solar port is open. Solar spectra are typically acquired once per day at the terminator in the northern hemisphere. The Sun Mean Reference spectrum will be derived from this mode.

White light source (direct)

The scan mirror points towards the WLS output mirror. The WLS is switched on and the solar port is closed. The WLS can be operated at four different currents (360, 380, 400, 420 mA). Etalon (and optionally Pixel-to-Pixel Gain (PPG) calibration) data will be derived from this mode.

Spectral light source (direct)

The scan mirror points towards the SLS output mirror. The SLS is switched on and the solar port is closed. Wavelength calibration coefficients will be derived from this mode.

Spectral light source (over diffuser)

The scan mirror points towards the diffuser. The SLS is switched on and the solar port is closed. Light from the SLS reaches the scan mirror via the diffuser. This mode is employed for in-orbit monitoring of the sun diffuser reflectivity.

LED

The scan mirror points towards the GOME-2 telescope. The LEDs are switched on and the solar port is closed. PPG calibration data will be derived from this mode.

Moon

The scan mirror points towards the moon (typical viewing angles are +70º to +85º). As the spacecraft moves along the orbit, the moon passes the GOME-2 slit within a few minutes. This mode can be employed only if geometrical conditions (lunar azimuth, elevation and pass angle) allow it which will typically occur a few times per year.

4.1.3.3 Other modes

These modes are either transitory (idle mode) or used in instrument maintenance (dump and test modes). In these modes, data packets are generated; however, they do not contain any useful scientific data. A typical example is during an in- or out-of-plane manoeuvre of the satellite.

Idle

This mode is reached during instrument switch-on or switch-off.

Dump

In place of PMD and main channel data, memory contents are downlinked. This mode is used for diagnostic purposes.

Test

In place of PMD and main channel data, a fixed test pattern is downlinked. This mode is used for diagnostic purposes.

4.1.4 GOME-2 timelines and timeline tables

The GOME-2 instrument may be operated using timelines (GTL) and timeline tables (GTT). Timelines are used primarily to reduce the load on the satellite uplink and additionally to provide on-board autonomy. One GTL is pre-loaded as a series of up to 33 individual instrument commands that are executed without the intervention of the Instrument Control Unit (ICU) or the Payload Module Controller (PMC). GOME-2 can store up to 16 timelines. Twelve default timelines will be loaded prior to launch and represent a library immediately available for use at the start of instrument operations.

GOME-2 operations and science data acquisition are strongly linked to the viewing geometry and the solar zenith angle (SZA), which determine the expected intensity of light received and the corresponding integration times required. There is in principle no restriction on the duration of a timeline or when it can be activated during an orbit. However, in order to simplify the sequencing and generation of timelines all default timelines start with an SZA equal to 90° minus an offset of 580 seconds and will have a duration of one orbit. The sequence of commands as well as their duration will remain constant over the year.

A GTT can be loaded, started and stopped by use of a macro command. The GTT allows 28 timelines and their execution times to be pre-loaded without the intervention of the ground segment or the PMC. There will be no default GTT stored on board and it is currently not planned to use the GTT capability for instrument operations.

For a full description of instrument operation and the default timelines see the [RD4] MetOp GOME-2 Instrument Operation Manual.

4.1.5 Timeline planning

GOME-2 timelines are scheduled according to a fixed timeline pattern per orbit repeat cycle. In the schematic shown below (Table 4.1) the day numbers are indicative only. This scheme has the following advantages:

The scheme is generally not affected by instrument switch-offs (i.e., there will be only gaps, no shifts) and the start orbit for the first cycle can be arbitrarily selected.

GOME-2 timeline planning per 412/29 repeat cycle. Version 3.0, 19 Jul 2007
Day
Orbit offset
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1
0
X
X
X
M1
M2
X
D
X
X
S
S
R
X
X
X
2
15
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
3
29
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
4
43
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
5
57
X
X
X
X
X
X
D
X
X
X
X
X
X
X
X
6
72
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
7
86
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
8
100
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
9
114
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
10
128
X
X
X
X
X
X
D
X
X
X
X
X
X
X
X
11
143
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
12
157
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
13
171
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
14
185
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
15
199
N
N
N
N
N
N
D
N
N
N
N
N
N
N
N
16
214
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
17
228
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
18
242
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
19
256
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
20
270
X
X
X
X
X
X
D
X
X
X
X
X
X
X
X
21
285
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
22
299
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
23
313
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
24
327
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
25
341
X
X
X
X
X
X
D
X
X
X
X
X
X
X
X
26
356
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
27
370
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
28
384
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
29
398
X
X
X
X
X
X
D
X
X
X
X
X
X
X
 
 
  D CAL0 Daily calibration (incl. Sun mode)
  M1 CAL4 Monthly calibration, part 1 (LED, WLS, SLS modes)
  M2 CAL5 Monthly calibration, part 2 (SLS over diffuser mode)
  N NS Narrow swath (320 km)
  S NADIR Nadir static
  R PMDRAW PMD monitoring (nominal readout/raw transfer mode)
  X NOT1920 Nominal swath (1920 km)

Table 4.1: GOME-2 timeline planning per 412/29 repeat cycle

4.1.6 On-ground calibration and characterisation

The GOME-2 instrument is built by an industrial team led by Galileo Avionica (I) with support from Laben (I), TNO-TPD (NL), Arcom Space (DK), Innoware (DK) and Finavitec (FIN). TNO-TPD are responsible for the calibration and characterisation of the instrument. The requirements against which the instrument is built are detailed in [RD10]. Requirements for calibration and characterisation measurements, derived from the expected accuracy of individual calibration parameters and the experimental set-up for on-ground calibration and characterisation activities, are listed below (see also [RD11] GOME-2 Calibration Plan and [RD12] GOME-2 Calibration Error Budget).

Requirements for measurements made in thermal vacuum

  • Measurement of absolute spectral radiance with an accuracy of 2.4% at 250 nm and 1.6% at 655 nm
  • Measurement of absolute spectral irradiance with an accuracy of 2.3% at 250 nm and 1.5% at 655 nm
  • Characterisation of the BSDF of the calibration unit to better than 1.2%
  • Characterisation of all polarisation responses to better than 1.2%
  • Characterisation of the instrument slit function

Requirements for measurements made under ambient conditions

  • Measurement of the scan angle dependence on polarisation and radiometric responses to 1.5% accuracy
  • Characterisation of the instrument stray light response
  • Characterisation of the instrument field of view.

Calibration and characterisation measurements needed in order to meet these requirements are taken during an extensive on-ground campaign. The detailed characterisation measurements are fully described in [RD11] GOME-2 Calibration Plan and [RD12] GOME-2 Calibration Error Budget. Characterisation measurements are post-processed to provide Calibration Key Data files which are documented both in terms of content and format in [RD13], [RD18] and [RD19]. Verification of the Calibration Key Data is addressed during dedicated Calibration Results Reviews for each flight model - see the Review Board reports [RD20], [RD21] and [RD22]. Verification of the on-ground instrument performance against instrument requirements is also carried out at this time. A sub-set of the Calibration Key Data are a required input to the GOME-2 Level 0 to 1 processor e.g. the radiance, irradiance and polarisation response of the instrument (see Figures 4.5, 4.6 and 4.7 for examples from the GOME-2 FM-203 Calibration Key data). For a full list of those Key Data used by the GOME-2 Level 0 to 1b processing chain see [RD7]. Other Key Data describe aspects of the on-ground behaviour of the instrument which will also be measured in-orbit using on-board calibration targets e.g. dark signal performance, pixel-to-pixel gain, spectral calibration and etalon. For these aspects the Calibration Key data form the starting point for instrument monitoring activities which are further discussed in Product Quality Evaluation.

Additionally, the GOME-2 Error Assessment Study (see [SCD2] and [SCD3]) has shown for the first time that O3 profile retrieval is very sensitive to knowledge of the shape of the slit function in the wavelength interval of the ozone Huggins bands (320-340 nm), which is used for retrieval at low altitudes. It was therefore concluded that, for height-resolved O3 data products from GOME-2 to meet specified user requirements (see [RD1] EPS End User Requirements Document and [O3M1] Ozone SAF User Requirements Document), the slit function shape must be characterised at sub-pixel resolution pre-flight, as this cannot be determined adequately from information available in-flight. As a result additional slit function characterisation data are being acquired during the on-ground calibration and characterisation campaign (see [RD17] GOME-2 FM3 Slit Function Test Report). These data have been further analysed by TNO-TPD and the Rutherford Appleton Laboratory. The result of this study is provided with additional calibration key data that describe the slit function shape at sub-pixel resolution, for use in Level 1 to 2 processing - see [RD26] and [RD27]. GOME-2 slit function key data is available together with related documentation on the EUMETSAT website under the EXTRANET section, under EPS CalVal > GOME CalVal > Calibration Data Sets. (Note that the EXTRANET pages are under development, and during the current CalVal phase external users will require a logon name and password to gain access. Please contact the EUMETSAT Helpdesk Service.)

Trace gas absorption spectra measurements have been carried out with each flight model after completion of the other on-ground calibration and characterisation activities. This is a dedicated activity carried out by the University of Bremen using the Calibration Apparatus for Trace Gas Absorption Spectroscopy (CATGAS) and with the support of TNO-TPD. In particular, absorption spectra of O3, NO2 and O2 have been measured in the wavelength region 230-800nm for a range of temperatures. These data have also been made available for use in Level 1 to 2 processing. CATGAS absorption spectra data are available on the EUMETSAT EXTRANET under EPS CalVal > GOME CalVal > Calibration Data Sets.

     

    Figure 4.5: Absolute Radiance Response of GOME-2 FM-203 for the main channel FPAs

     

    Figure 4.6: Absolute Irradiance Response of GOME-2 FM-203 for the main channel FPAs

     

    Figure 4.7: Polarisation Response Eta (intensity ratio of s to p polarised light for the exact nadir direction) of GOME-2 FM-203

4.1.7 In-flight characterisation and calibration

A range of in-flight characterisation and calibration activities are carried out routinely during GOME-2 operations. These activities provide input to Level 0 to 1b processing, in addition to the calibration key data measured on-ground, to ensure the generation of high quality spectrally and radiometrically calibrated radiance and irradiance data and continuous monitoring of instrument performance.

The frequency of on-board calibration activities is scheduled taking into account:

Calibration activities interleaved with nominal observations comprise dark signal measurements performed every orbit in eclipse, and sun calibration measurements performed once per day at sunrise in the northern hemisphere. The sun calibration uses one of the twelve on-board stored timelines which includes, in addition to the sun measurements themselves, both a wavelength calibration and a radiometric calibration. The timeline must be triggered such that the instrument is commanded into SUN observation mode prior to the sun appearing in the instrument field of view. For the remainder of the orbit the timeline consists of nadir scanning observations.

In addition, regular monthly calibration activities are carried out. The frequency is determined by the expected change in the mean optical bench temperature, resulting from seasonal variations in the external heat load from the sun, and long-term degradation of thermo-optical surfaces. Although it is primarily the wavelength calibration that is expected to vary, it is logical to perform diffuser characterisation and other calibration and monitoring activities during the monthly calibration activities.

A brief summary of the in-flight calibration and characterisation and monitoring activities is given in Table 4.2. Further details may be found in [RD5].

Activity
Mode
Frequency
Duration
Usage
Radiometric Calibration: Dark Signal Dark Every eclipse ~30 mins per orbit Calculate Dark Signal Correction
Radiometric Calibration: Uniformity of Pixel Response LED Monthly ~10 mins Calculate Pixel-to-Pixel Gain Correction
Wavelength Calibration Spectral Light Source (Direct) Monthly

Daily before Sun calibration		 ~4 min x 3 per orbit

~2 mins		 Calculate Spectral Calibration Parameters for Main Channels

Calculate Spectral Calibration Parameters for PMD Channels
Radiometric Calibration: Relative White Light Source (Direct) Daily after Sun calibration ~4 mins Calculate Etalon Correction
Radiometric Calibration: Sun Calibration Sun over
Diffuser		 Daily at the terminator in the Northern hemisphere ~30 seconds - full operational mode lasts ~2 mins Calculate Solar Mean Reference
Monitoring:
Diffuser Monitoring		 Spectral Light Source (over
Diffuser)		 Monthly ~15 mins during eclipse Instrument Monitoring: stability of the Sun diffuser
Monitoring:
Moon Measurements		 Moon Whenever the Moon is in the Field of View and the pass angle between the Moon pass and the along-slit direction is sufficient

Typically 3 to 6 times per year		 Variable during eclipse Instrument Monitoring: wavelength degradation of the scan mirror reflectivity
Monitoring:
Static View Measurements		 Nadir Static After monthly calibration Two full orbits Instrument Monitoring: potential degradation in the polarisation measuring devices

Table 4.2: Summary of the in-flight calibration and characterisation and monitoring activities

4.2 GOME-2 data processing

4.2.1 Level 0 to 1b data processing

The central processing facility, located at EUMETSAT headquarters in Darmstadt, is responsible for the processing of all GOME-2 data up to Level 1b, and delivers Level 0, Level 1a and Level 1b products to the user community. This Level 0 to 1b processing is carried out within the Core Ground Segment (CGS) by the GOME-2 Product Processing Facility (PPF) which converts raw instrument data (Level 0 data stream) into time-stamped, geolocated, and fully spectrally and radiometrically calibrated radiances or irradiances (Level 1b data stream). Level 0, 1a and Level 1b data products, and product quality and monitoring information are also generated by the CGS. The first level of functional decomposition of the GOME-2 processor is shown in Figure 4.8:

Figure 4.8: Functional decomposition of the GOME-2 processor

 

Receive and validate Level 0 and auxiliary data

The receive and validate function, in addition to the generic checks identified in [RD2], performs the instrument-specific acceptance and checking of the input data. Its purpose is to accept the Level 0 data and to perform all checks required for validation of the input data before passing them to the algorithmic functions. This functionality correlates Level 0 data with auxiliary data and also produces reporting statistics.

Level 0 to 1a processing

The Level 0 to 1a processing comprises both the determination of geolocation information on a fixed time grid, and the determination of applicable calibration parameters. From measurements of the various calibration sources encountered during each run of the processor, new calibration constants are calculated and written into an in-flight calibration data storage location. They are also retained in memory for use in processing those data acquired after the satellite comes out of the dark side of the orbit and before the next dump. Calibration parameter usage will be updated at the terminator. Calibration parameters are stored for the lifetime of the mission. The calibration constant determination comprises:

  • dark current correction
  • pixel-to-pixel gain correction
  • determination of spectral calibration parameters
  • etalon correction
  • determination of stray light correction factors for the sun and polarisation measurements
  • determination of the solar mean reference spectrum and atmospheric polarisation state

The geolocation of the measurements is calculated from the appropriate orbit and attitude information, and time correlation information in the Level 0 data stream. Note, any application of calibration parameters in the Level 0 to 1a processing should be regarded as interim, to facilitate the generation of new calibration parameters and correction factors. There is no application of calibration parameters to FPA Earth observation measurements. Schematics of the Level 0 to 1a processing flow are provided below (Figure 4.10 etc.). The output of the Level 0 to 1a processor is to be formatted into the Level 0 and 1a products as described in GOME-2 product formats and dissemination and specified in [RD3] and [RD8].

Level 1a to 1b processing

The Level 1a to 1b processing comprises the calculation of geolocation parameters for the actual integration time of each measurement, determination of stray light correction factors for the earthshine measurements, and the conversion of the raw binary readouts on the Level 1a data stream to calibrated radiance and irradiance data. Effective cloud fraction and cloud top pressure are also determined. Furthermore, calibrated measurements from the on-board calibration sources and from the sun and moon are available in the Level 1b product. Schematics of the Level 1a to 1b processing flow are provided below (Figure 4.14 etc.). Level 1b data are formatted as described in GOME-2 product formats and dissemination and specified in [RD3] and [RD8].

Sensor performance assessment (SPA)

The Sensor Performance Assessment (SPA) function allows instrument performance to be monitored for the lifetime of the mission. Performance is monitored both from an engineering point of view, using selected housekeeping data, and from a scientific point of view using spectral data, in particular in-flight calibration data, along with their respective time tags and geolocation. The default sampling interval is one instrument science data packet. Housekeeping data and selected earthshine data are extracted from the Level 1 product files. In-flight calibration data are extracted from the in-flight calibration data storage location. Housekeeping data are converted from binary units to physical units. Spectral data are preprocessed according to the instrument mode. The SPA function handles the extraction and preprocessing of these monitoring parameters. The extracted and preprocessed monitoring parameters are stored in the SPA data storage location and are made available on request for further analysis. Degradation correction factors will be calculated from these data where appropriate, following detailed scientific analysis.

Product quality evaluation

The Product Quality Evaluation (PQE) function provides information about the quality of the generated Level 1 data products. A number of checks are performed during Level 0 to 1a processing and Level 1a to 1b processing and the results are stored in Product Confidence Data records (PCDs) in the Level 1a and Level 1b data products. PCDs are provided both at the product and scan level. PCDs containing information about the quality of applied calibration parameters are also included. A preprocessing function extracts the PCDs directly after processing of the Level 1a and 1b data and updates the PQE storage location with the extracted data. A second function further condenses the extracted data to provide daily, weekly, monthly and yearly Product Quality Summaries and "Quick-Look" information. The data generated by the PQE function are made available for further analysis and visualisation. The highest level of detail, on measurement pixel level, is not covered by automatic PQE functionality.

Product and sensor performance real-time monitoring

According to the specified requirements for sensor performance assessment and product quality monitoring, the instrument and the whole processing chain is monitored on a near-real-time basis and results are made available online to the users under gome.eumetsat.int. For each processed GOME-2 orbit a summary of instrument performance parameters is provided with the GOME-2 Level 0 reports available online as a PDF on the server, together with an assessment of the quality of the calibrated Level 1b data which is included in the GOME-2 daily report - also available on the GOME-2 monitoring database server (see snapshot below, Figure 4.9). Individual orbit files with their full EPS filenames as available in the EUMETSAT Data Centre are linked to the appropriate timeline (see Section 4.1.4) carried out during that orbit. Anomalies of instrument or processing are summarised as out-of-limits events in the first page of the daily report (top of orbit table page) and indicated in the orbit table by colour for each orbit (red for instrument anomaly and yellow for processing anomaly, otherwise green for no anomaly detected during the orbit). This is illustrated in the following orbit table example:

Figure 4.9: Orbit table of the online GOME-2 monitoring database on gome.eumetsat.int

Auxiliary data required by the GOME-2 PPF

Initialisation data

This data set contains all parameter settings for the PPF, such as threshold values, switches between algorithm options, and instrument parameters not contained in the pre-flight calibration key data.

Orbit data

For near-real-time processing a predicted orbit state vector is required as input for the geolocation calculations. During reprocessing restituted orbit data are expected to be available.

Time correlation information

This information is required for the conversion of On-Board Time (OBT) to Coordinated Universal Time (UTC).

Static auxiliary data

The static auxiliary data comprise the static data sets that are required for use in the Level 1a to 1b processor. They are required in particular for the effective cloud fraction and cloud top pressure determination.

Key data

The calibration key data comprise the complete set of pre-flight calibration data which is provided by the instrument provider. A sub-set of these data are required as input to the Level 0 to 1 processor.

Correction factor data

Instrument characteristics such as radiance and irradiance sensitivity will change during the GOME-2 lifetime due to in-orbit degradation of the instrument. A subset of Level 0, 1a and 1b data necessary for the calculation of correction factors will be generated by the SPA function using in-flight measurements and will be made available for the derivation of correction factors. The derived correction factors will subsequently be used in the PPF for Level 0 to 1b processing. These data will not be available at the beginning of the in-orbit life of GOME-2.

In-flight calibration data

The Level 0 to 1a processing includes the determination of in-flight calibration parameters. From measurements of the various calibration sources encountered during each run of the processor, new calibration constants are calculated and written into an in-flight calibration data storage location. They are also retained in memory for use in processing those data acquired after the satellite comes out of the dark side of the orbit and before the next dump. As noted previously, calibration parameter usage will be updated at the terminator. Calibration parameters will be stored for the lifetime of the mission.

Figure 4.10: A2 functional decomposition: Level 0 to 1a processor (1)

 

Figure 4.11: A2 functional decomposition: Level 0 to 1a processor (2)

 

Figure 4.12: A2 functional decomposition: Level 0 to 1a processor (3)

 

Figure 4.13: A2 functional decomposition: Level 0 to 1a processor (4)

 

Figure 4.14: A3 functional decomposition: Level 1a to 1b processor (1)

 

Figure 4.15: A3 functional decomposition: Level 1a to 1b processor (2)

 

Figure 4.16: A3 functional decomposition: Level 1a to 1b processor (3)

 

Figure 4.17: A3 functional decomposition: Level 1a to 1b processor (4)

 

4.2.1.1 Level 1b product summary and estimated accuracies

A summary of the expected relative errors on the GOME-2 Level 1b products is given in the table below. The analysed errors in the absolutely calibrated radiance and irradiance spectra, the sun-normalised radiance spectra and the wavelength calibration parameters have been estimated by TPD-TNO (responsible for the instrument calibration and characterisation) on the basis of the accuracy of on-ground calibration and characterisation measurements. The product accuracies listed are generated using a Monte Carlo simulation method (see [RD12]). Using this method errors are generated for each contributing measurement according to a specified distribution function: Gaussian for random errors and Uniform for systematic errors. The parameters of the distribution depend on the measurement in question. The input measurements are perturbed according to these generated errors and a distribution of possible values in the final data product calculated. The 1σ estimated error in the final data product is equal to the standard deviation of this distribution.

Estimates of the absolute error in the absolutely calibrated radiance and irradiance spectra, and the sun-normalised radiance spectra, are also provided in the GOME-2 Level 1b product (see [RD8] GOME-2 Level 1 Product Format Specification). These error estimates are calculated from the root sum square of variances associated with all contributing error sources.

 
Analysed Error
Design Goal
Source
Sun-normalised nadir radiance error (1σ) 2.1% (UV)

1.4% (visible)		 1.5% (UV)

1.35% (visible)		 See:

[RD12] GOME-2 Calibration Error Budget

 and

[RD11] GOME-2 Calibration Plan
Absolute nadir radiance error (1σ) 1.9% (UV)

1.2% (visible)		 1.6% (UV)

1.4% (visible)
Absolute sun irradiance error (1σ) 1.7% (UV)

1.0% (visible)		 1.15% (UV)

0.8% (visible)
FPA wavelength calibration accuracy <0.04 pixel 0.04 pixel
PMD wavelength calibration accuracy <0.07 pixel 0.07 pixel
Effective cloud fraction Global monthly average difference:
0.0 in January
0.1 in July

The standard deviation of the differences in absolute values is 0.1 for both months		 N/A See:

[SCD22] R.B.A, Koelemeijer, P. Stammes, J.W. Hovenier and J.F. de Haan, "Global distributions of effective cloud fraction and cloud top pressure derived from oxygen A-band spectra measured by the Global Ozone Monitoring Instrument: Comparison to ISCCP data", J. Geophys. Res.,Vol. 107 D12, 2002
Cloud top pressure Global monthly average difference:
36 hPa in January
27 hPa in July

The standard deviation of the difference:
104 hPa in January
110 hPa in July		 N/A

Table 4.3: Expected relative errors for the GOME-2 Level 1b data products, except cloud parameters which are expressed in absolute values. Note that the wavelength calibration accuracies quoted are applicable to on-ground measurements and may not be representative of in-orbit accuracies.

Estimates of the variances associated with these contributing error sources are obtained as follows.

TPD-TNO provide an estimate of the relative error for:

These are used in the calculation of the relative error in the Müller Matrix elements in the Level 0 to 1 processing (see [RD7]). Estimates of shot and readout noise are derived from the in-orbit dark signal measurements [RD7], and estimates of the accuracy of the PPG, etalon and stray light corrections are derived from on-ground performance measurements ([RD15] and [RD16]). Estimates of the accuracy of the polarisation correction will be based on the results of the GOME-1 polarisation measurement validation ([SCD4]) and are currently being refined in orbit on the basis of in-orbit validation of measured GOME-2 Stokes fractions.

4.2.2 Level 1b to 2 data processing

4.2.2.1 The Ozone Monitoring Satellite Application Facility

The responsibility for extraction of meteorological or geophysical (Level 2) products from GOME-2 lies with the Satellite Application Facility on Ozone & Atmospheric Chemistry Monitoring (O3M SAF). The development of the O3M SAF was started in 1997 and is coordinated by the Finnish Meteorological Institute (FMI) in Helsinki. The O3M SAF consortium comprises:

  • Finnish Meteorological Institute or Ilmatieteen Laitos (FMI) - host institute, Finland
  • Koninklijk Nederlands Meteorologisch Instituut (KNMI), Netherlands
  • Deutsches Fernerkundungsdatenzentrum (DLR), Germany
  • Deutscher Wetterdienst (DWD), Germany
  • Aristotle University of Thessaloniki, Greece
  • Hellenic National Meteorological Service (HNMS), Greece
  • Danmarks Meteorologiske Institut (DMI), Denmark
  • Météo-France, France
  • Koninklijk Meteorologisch Instituut van België / Institut Royal Météorologique de Belgique (KMI - IRM), Belgium

As part of the distributed element of the EUMETSAT Applications Ground Segment, the O3M SAF will provide operational services to end-users, e.g. real-time or off-line product services, data management and related user services, including coordination of and support to relevant research and development. The SAF Visiting Scientist Programme allows involvement of scientific experts external to the SAF Consortium.

Information on O3M SAF activities can be found on the O3M SAF website. Access to GOME-2 Level 2 data produced by the O3M SAF is provided via the same website or directly from the EUMETSAT Data Centre.

4.2.2.2 Operational Level 2 products from GOME-2

The O3M SAF produces three classes of Level 2 products from GOME-2. These are:
- near-real-time (NRT) products, which are made available and distributed to users within 3 hours of sensing,
- off-line products which are available no later than 15 days after sensing from the O3M SAF archive,
- experimental products whose dissemination and coverage is yet to be decided.

Reprocessing the geophysical products for climate applications (see [O3M1] Ozone SAF User Requirements Document, [O3M2] Ozone SAF Science Plan and [O3M3] Ozone SAF Scientific Prototyping Report) is also anticipated. For a discussion of individual products see below. The O3M SAF also provides online validation services on an operational basis (for a summary please see Validation of Ozone Monitoring SAF Products).

4.2.2.3 Ozone profile and aerosol products

The operational ozone profile and aerosol products from GOME-2 are produced under the responsibility of KNMI. The ozone profile products, both NRT and off-line, will be made available at a frequency determined by the Metop orbit repeat cycle. The aerosol products comprise an Absorbing Aerosol Index (AAI) and an Aerosol Optical Depth (AOD) and will be produced daily off-line. The ozone profile and AAI products will be available at a horizontal resolution equal to the GOME-2 instrument ground pixel size (i.e. 80 km across track x 40 km along track, assuming a default integration time of 0.1875 s and a swath width of 1920 km). The Aerosol Optical Depth product will be available at a horizontal resolution equal to that of the GOME-2 PMD measurements (i.e. 10 km across track x 40 km along track, assuming a default integration time of 23.4375 ms and a swath width of 1920 km).

The ozone profile products are produced using the VERA (VErsatile Retrieval Algorithm) retrieval algorithm. VERA (and its predecessor algorithm Opera) has a strong heritage of operational use for retrievals from GOME-1 on ERS-2, SCIAMACHY on Envisat and OMI on EOS-Aura (see for example [SCD5]).

The VERA retrieval algorithm uses an optimal estimation formalism, the details of which are fully described in [O3M2] Ozone SAF Science Plan and [O3M3] Ozone SAF Scientific Prototyping Report. The forward radiative transfer model used is a scalar version of the LIDORT model ([SCD6] and [SCD7]). For the retrieval of ozone profile information, wavelengths in the range 270 - 330 nm are used. The forward model calculations are carried out on 40 atmospheric layers and the ozone profile information is retrieved on the same grid. The ozone profile information will however only be reported as a layer amount for five tropospheric layers and seven stratospheric layers. A vertical resolution of 15 km in the troposphere and 7 km in the stratosphere (below 2 hPa) can be expected. In order to minimise residual errors in the retrieved ozone profile resulting from the use of a scalar forward model, look-up tables depending on a number of parameters including view angle, solar zenith angle, relative azimuth angle, surface albedo, surface pressure and ozone profile information, will be used to correct the modelled radiances for errors due to the neglect of polarisation. The error remaining after correction is expected to be negligible. Additionally, in order to correct for neglect of the Ring effect in the modelled radiances, a pre-calculated Ring spectrum derived from a high resolution solar reference spectrum is applied. The amplitude of the ring absorption is included as an auxiliary parameter in the state vector and retrieved. This does not however correct for the filling in of ozone absorption lines which will be the subject of further study. An example of the GOME/ERS-2 Fast Delivery Ozone Profile Product is shown in Figure 4.18.

    Figure 4.18: Example of the GOME/ERS-2 Fast Delivery Ozone Profile Product (courtesy of KNMI)

    Figure 4.19: Dust storm event over the Atlantic Ocean as seen by GOME-2 Absorbing Aerosol Index 22-24 June 2007 (provided courtesy of O3MSAF/KNMI - O. Tuinder)

The Aerosol Optical Depth will be retrieved using measurements from the GOME-2 PMDs. The retrieval algorithm will incorporate a linearized vector radiative transfer model and the inversion will be performed using the Levenberg-Marquardt method. The algorithm is currently under development and is the subject of an O3M SAF Visiting Scientist activity. The expected accuracy of the Aerosol Optical Depth product remains to be demonstrated.

In addition to the AOD, an Absorbing Aerosol Index (AAI) will also be produced on an operational basis. The AAI is an index based on the ratio of the back-scattered radiance at 340 nm and to that at 380 nm. As the ozone absorption is small in this spectral region, and the spectral signature of non-absorbing aerosols is flat, it is assumed that any deviation from the signal associated with pure Rayleigh scattering indicates the presence of absorbing aerosols. An AAI product has been produced on a regular basis from TOMS data for a number of years (see [SCD8]) and its value as a tracer of dust and smoke aerosols, particularly from biomass burning, has been clearly demonstrated (e.g. see [SCD9] and Figure 4.19). As the AAI is sensitive to wavelength-dependent degradation of the instrument response, it may also be used for diagnostic purposes in assessing the quality of the Level 1b data product.

Both the ozone profile and aerosol products will be disseminated to users in HDF5 format. See [O3M4] Ozone SAF Output Product Format Document for OOP and ARS for a full description of the product structure.

4.2.2.4 Total column ozone and trace gases

The total column ozone and trace gas products from GOME-2 are produced by DLR. This set of products comprises NRT total column O3 and NO2, off-line total column O3 and NO2, and experimental total column BrO, SO2 and HCHO. These products will also be produced at a frequency determined by the Metop orbit repeat cycle and will be produced at a horizontal resolution equivalent to the GOME-2 ground-pixel size (i.e. 80 km across track x 40 km along track assuming a default integration time of 0.1875 s and a swath width of 1920 km). Global coverage of the experimental products is not however assured.

The algorithms used to generate the total column and trace gas products are based on the Differential Optical Absorption Spectroscopy (DOAS) technique. The operational GOME/ERS-2 Data Processor V3.0 (GDP V3.0) provides the basis for the further developments and improvements planned for GOME-2 on Metop. A full description of the GDP V3.0 algorithm and related validation activities may be found in [SCD10]. Further planned enhancements to the algorithm include an improved treatment of the molecular ring effect, calculation of Air Mass Factor s (for conversion from a slant to a vertical column) at 325.5 nm for high SZAs >80°, and application of an I0 corrected ozone reference spectrum. A discussion of these improvements can be found in [O3M3] Ozone SAF Scientific Prototyping Report. Under the responsibility of ESA, the updated processor GDP4.0 will be used to reprocess the entire GOME/ERS-2 data set. Continuity between GOME/ERS-2 and GOME-2 on Metop is therefore assured.

Examples of the total column ozone and trace gas products that have been produced both by DLR and the wider scientific community from GOME-2 data are shown in the figures below. Both the total column ozone and trace gas products will be disseminated to users in HDF5 format. See [O3M5] for a full description of the product structure. 

    Figure 4.20: Example of a GOME-2 Total Column Ozone Field (provided courtesy of O3MSAF/DLR)

 

    Figure 4.21: Example of a GOME-2 Total Column NO2 Field (provided courtesy of O3MSAF/DLR)

 

    Figure 4.22: Example of a GOME-2 Total Column BrO Field (provided courtesy of A. Richter, University of Bremen)

 

    Figure 4.23: Example of a GOME-2 Total Column SO2 Field and comparison to OMI (provided courtesy of A. Richter, University of Bremen, and N. Krotkov, NASA)

4.2.2.5 Near-real-time UV products

The near-real-time UV product produced by DMI provides daily clear-sky UV fields expressed as a UV index (see [SCD11]). The product consists of 23 contour maps, for pre-specified regions, two ASCII files containing general information and estimated accuracies for the clear-sky UV fields, and an html file for easy access to the maps and information contained within the product. For a full description of the product contents see [O3M6].

The NRT UV product processor employs the widely used UVSPEC radiative transfer model (see [SCD12]) based on the discrete ordinate method (DISORT) [SCD13] and is based on look-up tables. In the look-up tables pre-calculated UV index values have been tabulated for wide ranges of solar zenith angles (SZAs), ozone amounts, surface albedo and five different ozone profiles. Additionally, tables have been produced for model atmospheres representative of different latitudes and seasons. The UV index values are determined by interpolating in total column ozone amount, SZA and albedo for the most appropriate latitude and season. Corrections for sun-Earth distance, altitude and aerosol are also applied. The only dynamic input parameter used in the calculations is total column ozone. Climatological parameters are used for all other input parameters including surface albedo. The total column ozone value used is an assimilated GOME-2 total column ozone field provided as an internal O3M SAF product by KNMI. In the case that the assimilated total column ozone data are not available, monthly average total column ozone data from TOMS will be used and the product flagged accordingly. An alternative approach is to use total column ozone forecast fields provided to the National Meteorological Services by ECMWF. For a full description of the algorithm see [O3M2] Ozone SAF Science Plan and [O3M3] Ozone SAF Scientific Prototyping Report.

The NRT UV product is produced daily with a maximum delay of 60 minutes after receipt of the assimilated total column ozone, which is not later than 2:45 UTC. The horizontal resolution of the NRT UV index contour maps is 0.25 x 0.25 degrees. Accuracy is expected to be within the required 1 index limit. A prototype NRT UV product is provided by DMI at http://www.dmi.dk/dmi/index/verden/uv_idag.htm.

4.2.2.6 Off-line UV fields including clouds and surface albedo

The off-line UV product, including clouds and surface albedo, is produced by FMI. The product will contain the daily UV dose in J/m2 weighted with four different action spectra: erythema induction (CIE87) [SCD14], generalised plant damage [SCD15], DNA damage [SCD16] and skin cancer induction [SCD17]. The product is currently planned to be made available off-line at a frequency determined by the Metop orbit repeat cycle, with a horizontal resolution equal to the GOME-2 instrument ground pixel size (i.e. 80 km across track x 40 km along track assuming a default integration time of 0.1875 s and a swath width of 1920 km). Note that a daily gridded product is being considered as an alternative.

The two most critical inputs to the off-line UV product are the estimated diurnal cloud cover and surface albedo. For the estimation of diurnal cloud cover, two approaches are being considered, the first being the current baseline. In the first approach cloud optical thickness is estimated from the AVHRR Global Area Coverage (GAC) data, using an AVHRR preprocessor module which employs pre-calculated look-up tables parameterised in terms of AVHRR channel one radiance, solar zenith angle, satellite zenith angle, relative azimuth angle, terrain pressure, surface albedo and total column ozone. Cloud optical thickness estimated from GOME-2 data will be used as a back-up in the event of AVHRR data being unavailable. The second approach is to use cloud optical thickness from other projects, in particular from the cloud products of the Climate Monitoring SAF (CM SAF). At present the CM SAF cloud products are produced on a regional basis and are therefore unsuitable for use by the off-line UV data processor; however during the operational phase of the CM SAF global cloud products are planned. At an appropriate time the possibility of using these products as input to the off-line UV data processor of the O3M SAF will be reconsidered.

In the case of the estimation of surface albedo, if an ECMWF snow analysis field is available, and indicates the presence of snow, an empirical conversion between snow depth and surface albedo, following the method of Arola et al. 2003 (see [SCD18]), is used. Note that further than 250 km from an observation station, the current ECMWF snow analysis produces a climatological value only. If no snow is detected or the ECMWF snow field is unavailable, the surface albedo is selected from the Minimum Lambert Equivalent Reflectivity (MLER) climatology compiled from fourteen and a half years of Nimbus-7/TOMS data by Herman and Celarier (see [SCD19]).

The off-line UV product is planned to be provided in HDF5 format. For specific details on the structure of the products please see [O3M7] Ozone SAF Output Product Format Document for OUV. Note that for a daily gridded product, GRIB format would be considered.

4.2.2.7 Level 2 product summary and expected accuracies & precision

Product
Characteristics
Estimates Uncertainties
Source
Total column ozone Total vertical column amount in Dobson Units

NRT and Off-line 		SZA Accuracy

(1σ)		 Precision

(1σ)		 See:

[O3M3] Ozone SAF Scientific Prototyping Report
< 70° -2%... +4% < 2%
< 90° 5%... +8% < 3%
Trace Gases Total vertical column amount in mol.cm-2

Off-line		 SZA Accuracy

(1σ)		 Precision

(1σ)
NO2 Operational < 70° 5%...20% 5%...20%
NO2 Operational < 90° 5%...20% 5%...10%
BrO tropics Operational N/A 20%...50% 50%...100%
BrO tropical enhancement Operational N/A > 100% 10%...50%
BrO other Operational N/A 20%...50% 10%...50%
SO2 Experimental < 65° 50%...100% 50%...100%
SO2 Experimental > 65° > 100% > 100%
SO2 volcanic Experimental < 65° 50%...100% 5%...30%
SO2 volcanic Experimental > 65° > 100% > 30%
HCHO Experimental N/A > 100% >100%
HCHO biomass burning Experimental N/A 50%...100% 20%...50%
OClO ozone hole Experimental > 75° 50%...100% 20...50%
OClO Experimental other > 100% > 100%
Ozone Profiles Eleven layers of ozone mixing ratio (ppm)

NRT and Off-line		 < 10% in the stratosphere and < 30% in the troposphere for six to eight independent pieces of information. See:

[O3M2] Ozone SAF Science Plan
Aerosol Absorbing Aerosol Indicator (AAI), Aerosol Optical Depth (AOD) and Aerosol Type (desert dust, smoke and volcanic ash)

NRT		 AAI - No accuracy estimates are given.

AOD - Accuracy < 20%

The AOD will only be computed if AAI > 1
Clear Sky UV Fields Surface level spectral UV irradiance

NRT

Product available as a clear sky UV index		 Accuracy 1 UV index
UV fields with Clouds and Albedo Spectral UV doses weighted with an appropriate action spectra for clear sky and cloudy conditions

Off-line		 Accuracy < 20% for a 100 x 100 km grid.

Table 4.4: Expected product accuracies for operational GOME-2 Level 2 products produced by the O3M SAF