product user manual: level 1b and 2 products - rom saf · the rom saf leading entity is the danish...

Ref: SAF/ROM/DMI/UG/PUM/001

Product User Manual: Level 1B and 2 products

Version 3.7

14 May 2020

ROM SAF Consortium Danish Meteorological Institute (DMI)

European Centre for Medium-Range Weather Forecasts (ECMWF) Institut d’Estudis Espacials de Catalunya (IEEC)

Met Office (UKMO)

Ref: SAF/ROM/DMI/UG/PUM/001 Version: 3.7 Date: 14 May 2020


2 of 50

DOCUMENT AUTHOR TABLE

Author(s) Function Date Prepared by: Johannes K. Nielsen ROM SAF Scientist 14/5 2020

Reviewed by (internal):

Sean Healy ROM SAF Science Manager 20/1 2019

Approved by: Kent B. Lauritsen ROM SAF Project Manager 14/5 2020

DOCUMENT CHANGE RECORD

Version Date By Description 0.1 03/05/07 FRR 1st draft 1.0 16/05/08 FRR ORR-A version 1.1 02/09/08 FRR ORR-A Close-out version based on RIDs # 1 →

16, 31, 33, 35 → 38, 60, 72, and 73. Parts of Chapter 4 moved to Chapter 3 for better text structure. Textual updates to chapter 1.2, 2.4, 3.1, 3.2.10, 3.3.1, 3.4, 4.3.2, 4.3.3, and 5, and to Figure 6-1.

1.2 27/03/09 FRR Update related to declaring the NRT Refractivity Product pre-operational

1.2.1 31/03/09 FRR Minor corrections to version 1.2 1.3 15/05/09 FRR ORR-B1 version. Minor updates to products

tables in Chapter 3, to align this document with the PRD. New chapter 5.2 included.

1.4 01/02/10 FRR Version associated with GRM-01 being upgraded to version 1.3 and declared operational (chapter 5.1 updated). New colour codes in Table 1-1, new logo and updates from ORR-B1 RIDs # 1 → 21, 56 → 60, 62 → 77, 105, 106, 119, 121. Chapter 5.2 temporarily removed pending GRM-02 → 05 being declared operational.

1.5 08/02/11 FRR ORR-B Close-Out version. Updates from RID # 120 (plots in chapter 3.2). Updated appendices.

1.6 09/06/11 FRR Updates according to Action 2 from ORR-B1 Close-Out; Chapter 5.2 reinserted



3 of 50

Version Date By Description 2.0 28/3/2014 KBL Version for ORR_GRM_40_part2;

Chap. 5 updated with description of both GRM-01 and GRM-40 data quality; editorial changes (GRAS SAF -> ROM SAF, etc.)

2.1 9/4 2014 KBL Version updated with ORR_GRM_40_part2 RIDs: AvE01; OR6 action 12 implemented (Sect. 1.4);

2.2 12/5 2014 JKN Version for ORR4 & ORR-B backlog; section 5.2-5.5 added; Table 1.1 updated; minor updates.

2.3 20/5 2014 JKN Version submitted for re-scoped ORR4 & ORR-B-backlog review with PRD version 2.2draft as baseline. Section 5.4-5.5 updated.

2.4 26/6 2014 JKN Updated version for ORR4 & ORR-B-backlog review, implemented RIDs: 16, 37

2.5 12/10 2016 JKN Updated version submitted as input to the ORR8 review related to Wave Optics BA input data. Only Sections 5.1 - 5.5 are updated

2.6 24/10 2016 KBL Updated after the ORR8 review: Implemented RIDs 1 and 2.

3.0 13/6 2018 JKN Version for DRR-RE1 and ORRs review. List of changes:

- adaptations from CDOP-2 to CDOP-3 - Ch. 4, updated all figures and figure

captions - Ch. 4, added section about dry

temperature - Ch. 4, added section about tropopause

height - Ch. 1, 2, 4, mention reprocessing

3.1 4/9 2018 JKN Updated version implementing the following RIDS for DRR-RE1 & ORRs review: - RIDs: 002, 004, 007, 012, 013, 331, 342, 347, 348, 352, 353, 354, 355, 356, 357, 358, 359, 361, 362, 363, 367, 368, 369, 370, 371, 372, 373, 374, 383, 384, 386, 388,389, 390, 399, 409, 414, 418, 420: Editorial changes implemented. - RID 008: Sec. 1.1 updated with reference to ATBD -RID 009 Sec. 3.3 Section title changed. -RID 010 Sec. 3.4 last sentence changed wording and referenced to ATBD - RID 299 ERA5 homogeneity issues mentioned in Quality and Limitations



4 of 50

Version Date By Description In addition the following changes were made: Ch. 2 and Ch. 4 titles changed.

3.2 22/11 2018 HGL Updated version based on ICDR concept discussions at ROM SAF SG22: : Update of the ROM SAF introduction. Sections 1.1, 1.4, and 2: Revised descriptions of the offline, CDR, and ICDR data types. Table 2-1 separated into four tables. Introduced a new Section 3.3 listing missions and time coverages (including Table 3-1). Section 4.1: slightly updated text. Section 4.1.2: updated descriptions of the offline, CDR, and ICDR data types. Chapter 7: simplified, removed sub-sections.

3.3 13/3 2019 JKN Updated version based on internal ROM SAF review: -section 4.1: 2nd paragraph; changed wording. -section 4.2.5: Range: [180;350] -> [150;350] -section 4.3.2: height -> undulation -p.39: models -> forecast and analysis etc. Updated with NRT Metop-C products: - section 1.2.2: Metop-C VRs in [RD.25/26] - tables 2.2, 2.3, 3.1: Metop-C inserted, dates adjusted

3.4 20/4 2019 JKN Added TPH (GRM-24) to table 2.2 3.5 24/5 2019 HGL Updated version with more detailed descriptions of

offline and ICDR data: Section 3.3: Update of text and Table 3-1.

3.6 16/3 2020 JKN Version prepared for the ORR12 review: - Sec1.2: References to Offline v1.1 Validation Reports - Sec 1.3: ERA5 in acronyms - Chapter 2: Include Metop-C products - Sec 3.3: Mentioning Metop-C - Sec 4.3.3: Mentioning ERA5

3.7

6/4 2020 JKN Updated version prepared for ORR12 Closeout implementing the following: - Sec. 3.5 updated with description of “kappa” method [RID 097] - Sec. 4.2.12: Changed PCD-flag description [RID 011] - Chapter 7. Removed (DVD) [RID 135] - Sec. 3.2: Inserted text (fundamental) [RID 004] - Sec. 3.3: Clarified versions [RID 005] - Sec. 4.2: Accuracy changed [RID 137] - Sec. 4.4: Mentioned balloons [RID 012]



5 of 50

ROM SAF The Radio Occultation Meteorology Satellite Application Facility (ROM SAF) is a decentralised processing centre under EUMETSAT which is responsible for operational processing of radio occultation (RO) data from the Metop and Metop-SG satellites and radio occultation data from other missions. The ROM SAF delivers bending angle, refractivity, temperature, pressure, humidity, and other geophysical variables in near real-time for NWP users, as well as reprocessed Climate Data Records (CDRs) and Interim Climate Data Records (ICDRs) for users requiring a higher degree of homogeneity of the RO data sets. The CDRs and ICDRs are further processed into globally gridded monthly-mean data for use in climate monitoring and climate science applications. The ROM SAF also maintains the Radio Occultation Processing Package (ROPP) which contains software modules that aid users wishing to process, quality-control and assimilate radio occultation data from any radio occultation mission into NWP and other models. The ROM SAF Leading Entity is the Danish Meteorological Institute (DMI), with Cooperating Entities: i) European Centre for Medium-Range Weather Forecasts (ECMWF) in Reading, United Kingdom, ii) Institut D'Estudis Espacials de Catalunya (IEEC) in Barcelona, Spain, and iii) Met Office in Exeter, United Kingdom. To get access to our products or to read more about the ROM SAF please go to: http://www.romsaf.org Intellectual Property Rights All intellectual property rights of the ROM SAF products belong to EUMETSAT. The use of these products is granted to every interested user, free of charge. If you wish to use these products, EUMETSAT's copyright credit must be shown by displaying the words “copyright (year) EUMETSAT” on each of the products used.

http://www.romsaf.org/



6 of 50

List of Contents 1. INTRODUCTION ............................................................................................................................. 8

1.1 PURPOSE OF THE DOCUMENT ......................................................................................................... 8 1.2 APPLICABLE AND REFERENCE DOCUMENTS ................................................................................... 8

1.2.1 Applicable documents ............................................................................................................. 8 1.2.2 Reference documents ............................................................................................................. 8

1.3 ACRONYMS AND ABBREVIATIONS .................................................................................................. 11 1.4 DEFINITIONS .................................................................................................................................. 12 1.5 OVERVIEW OF THIS DOCUMENT .................................................................................................... 13

2. LIST OF LEVEL 1B AND 2 DATA PRODUCTS ........................................................................ 14

3. THE RADIO OCCULTATION METHOD ..................................................................................... 18

3.1 OVERVIEW ..................................................................................................................................... 18 3.2 BENEFITS OF RADIO OCCULTATION PROFILING ............................................................................. 19 3.3 SATELLITE MISSIONS AND TIME COVERAGE .................................................................................. 19 3.4 CHARACTERISTICS OF GRAS AND OTHER RO INSTRUMENTS .................................................... 20 3.5 PROCESSING TO LEVEL 2 PRODUCTS ........................................................................................... 20

4. DATA PRODUCT DESCRIPTION ............................................................................................... 24

4.1 OVERVIEW OF ROM SAF LEVEL 1B AND 2 DATA PRODUCTS ..................................................... 24 4.1.1 NRT products ......................................................................................................................... 24 4.1.2 Offline data, CDRs, and ICDRs ........................................................................................... 25 4.1.3 Supporting data ...................................................................................................................... 25 4.1.4 Error-covariance matrices .................................................................................................... 26

4.2 LEVEL 1B AND 2 ATMOSPHERIC PROFILE DATA PRODUCTS ......................................................... 27 4.2.1 General .................................................................................................................................... 27 4.2.2 Bending angle ........................................................................................................................ 28 4.2.3 Refractivity profile .................................................................................................................. 29 4.2.4 Dry temperature profile ......................................................................................................... 30 4.2.5 Temperature profile ............................................................................................................... 31 4.2.6 Humidity profile ...................................................................................................................... 32 4.2.7 Pressure profile ...................................................................................................................... 33 4.2.8 Surface pressure.................................................................................................................... 34 4.2.9 Tropopause heights ............................................................................................................... 34 4.2.10 Heights................................................................................................................................ 34 4.2.11 Location .............................................................................................................................. 35 4.2.12 Quality information ............................................................................................................ 35

4.3 SUPPORTING DATA ........................................................................................................................ 36 4.3.1 Satellite orbits (POD) ............................................................................................................ 36 4.3.2 Local radius of curvature, azimuth, and geoid undulation ............................................... 37 4.3.3 Background profiles from ECMWF ...................................................................................... 38 4.3.4 Other data ............................................................................................................................... 38

4.4 VALIDATION DATA .......................................................................................................................... 38 4.5 ERROR COVARIANCE MATRICES ................................................................................................... 40

5. FORMAT DESCRIPTIONS .......................................................................................................... 41

5.1 INTRODUCTION .............................................................................................................................. 41 5.2 BUFR FORMAT ............................................................................................................................. 41 5.3 NETCDF FORMAT .......................................................................................................................... 41

5.3.1 Structure .................................................................................................................................. 41



7 of 50

5.3.2 File format ............................................................................................................................... 41 5.3.3 File names .............................................................................................................................. 42

6. DATA QUALITY AND LIMITATIONS ......................................................................................... 44

7. DISSEMINATION OF DATA PRODUCTS .................................................................................. 45

8. SUMMARY .................................................................................................................................... 46

ANNEX A NETCDF HEADER FORMAT ......................................................................................... 47

ANNEX B LEVEL 1 DATA NETCDF FORMATS ........................................................................... 49

ANNEX C LEVEL 2 DATA NETCDF FORMATS ........................................................................... 50



8 of 50

1. Introduction 1.1 Purpose of the document

This document provides a user guide to the Level 1B (bending angle) and Level 2 (geophysical variables) atmospheric profile data available from the ROM SAF. The data products covered by this user guide are listed in Section 2. They include NRT data products, available for time-critical applications, offline data products that are generated on a regular basis for non-time-critical applications, Climate Data Records (CDRs) that were generated in a dedicated reprocessing activity using a single version of processing system and software, and Interim Climate Data Records (ICDRs) that are generated on a regular basis with the same algorithms as the CDRs but using currently available data. This document only describes the Level 1B and Level 2 data products (profile data). The Level 3 gridded data products have a separate PUM [RD.12], and the ROPP software package has its own User Guide [RD.8]. The algorithms used and the processing methods adopted are discussed in the ATBDs [RD.6], [RD.16] and [RD.17].

1.2 Applicable and reference documents

1.2.1 Applicable documents The following list contains documents with a direct bearing on the contents of this document:

[AD.1] CDOP-3 Proposal: Proposal for the Third Continuous Development and Operations Phase (CDOP-3); Ref: SAF/ROM/DMI/MGT/CDOP3/001 Version 1.2 of 31 March 2016, Ref: EUM/C/85/16/DOC/15, approved by the EUMETSAT Council at its 85th meeting on 28-29 June 2016

[AD.2] CDOP-3 Cooperation Agreement: Agreement between EUMETSAT and DMI on the Third Continuous Development and Operations Phase (CDOP-3) of the Radio Occultation Meteorology Satellite Applications Facility (ROM SAF), Ref. EUM/C/85/16/DOC/19, approved by the EUMETSAT Council and signed at its 86th meeting on 7 December 2016

[AD.3] ROM SAF Product Requirements Document, Ref: SAF/ROM/DMI/MGT/PRD/001

[AD.4] ROM SAF Service Specifications, Ref: SAF/ROM/DMI/RQ/SESP/001

1.2.2 Reference documents The following documents provide supplementary or background information, and could be helpful in conjunction with this document:

[RD.1] EPS End-User Requirements Document (EURD), Ref. EPS/MIS/REQ/93001



9 of 50

(also Annex I to EUM/C/36/97/DOC/54)

[RD.2] User Requirement Document, Ref: SAF/GRAS/METOFFICE/RQ/URD/001

[RD.3] GRAS Level 1 Product Format Specification. Ref. EPS/MIS/SPE/97234

[RD.4] GRAS Level 1 Product Generation Specification. Ref. EPS/SYS/SPE/990010

[RD.5] ROM SAF WMO FM94 (BUFR) Specification for radio occultation data. Ref: SAF/ROM/METO/FMT/BUFR/001

[RD.6] ROM SAF, Algorithm Theoretical Baseline Document: Level 2A Refractivity, SAF/ROM/DMI/ALG/REF/001

[RD.7] ROM SAF CT2 Processing Code: Operational Processing of CHAMP and COSMIC data: Mathematical Methods, Data Filtering and Quality Control, version 1.1. Ref: SAF/GRAS/DMI/ALG/CT2/002

[RD.8] ROM SAF, Validation report: Reprocessed level 3 gridded CDR v1.0 products, Ref: SAF/ROM/DMI/REP/CLM/001

[RD.9] ROM SAF WMO FM94 (BUFR) Specification For Radio Occultation Data, Ref: SAF/ROM/METO/FMT/BUFR/001

[RD.10] M. E. Gorbunov: Ionospheric correction and statistical optimization of radio occultation data. Radio Science, vol. 37, no. 5, 1084, doi:10.1029/2000RS002370, 2002

[RD.11] ROM SAF Product Output Format. Ref: SAF/ROM/DMI/FMT/POF/001

[RD.12] ROM SAF Product User Manual; Offline Level 3 Gridded Data, Ref: SAF/ROM/DMI/UG/GRD/001

[RD.13] The Radio Occultation Processing Package (ROPP) Pre-processor Module User Guide Version 9.0, Ref: SAF/ROM/METO/UG/ROPP/004

[RD.14] The Radio Occultation Processing Package (ROPP) Applications Module User Guide Version 9.0, Ref: SAF/ROM/METO/UG/ROPP/005

[RD.16] ROM SAF, Algorithm Theoretical Baseline Document: Level 2B and 2C 1D-Var products, SAF/ROM/DMI/ALG/1DV/002

[RD.17] ROM SAF Algorithm Theoretical Baseline Document: Level 2C tropopause height, Ref: SAF/ROM/DMI/ALG/TPH/001



10 of 50

[RD.18] ROM SAF Validation Report: Near Real-Time Level 2A Refractivity Profiles: Metop-A (GRM-01, NRPMEA) and Metop-B (GRM-40, NRPMEB), Ref: SAF/ROM/DMI/RQ/REP/001

[RD.19] ROM SAF Validation Report: NRT profiles Level 2B and 2C 1D-Var products: Metop-A (GRM-02,03,04,05) and Metop-B (GRM-41,42,43,44), Ref: SAF/ROM/DMI/RQ/REP/002

[RD.20] ROM SAF Validation report: Reprocessed level 1B bending angle, level 2A refractivity, level 2A dry temperature CDR v1.0 products, Ref: AF/ROM/DMI/REP/PRF/001

[RD.21]

ROM SAF Validation Report: Reprocessed Level 2B and 2C 1D-Var CDR v1.0 products, Ref: SAF/ROM/DMI/REP/1DVAR/001

[RD.22] ROM SAF Validation Report: Reprocessed Level 2C tropopause height CDR v1.0 products (Short title: Validation Report: Reprocessed TPH products) , Ref: SAF/ROM/DMI/REP/TPH/001

[RD.23] ROMSAF ATBD: Level 2A dry temperature profiles, Ref: SAF/ROM/DMI/ALG/TDRY/001

[RD.24] ROM SAF ATBD: Level 1B bending angles, Ref: SAF/ROM/DMI/ALG/BA/001

[RD.25] ROM SAF Validation Report: NRT Level 2A Refractivity Profiles: Metop-C (GRM-60), Ref: SAF/ROM/DMI/RQ/REP/003

[RD.26]

ROM SAF Validation Report: NRT Level 2B and 2C 1D-Var products: Metop-C (GRM-61,62,63,64), Ref: SAF/ROM/DMI/RQ/REP/004

[RD.27]

ROM SAF Validation Report: Offline Level 1B bending angle, Level 2A refractivity, Level 2A dry temperature products version 1.1, Ref: SAF/ROM/DMI/REP/ATM/002

[RD.28]

ROM SAF Validation Report: Offline Level 2B and 2C 1D-Var products version 1.1, Ref: SAF/ROM/DMI/REP/1DVAR/002

[RD.29]

ROMSAF Validation Report: Offline Level 2C tropopause height products version 1.1, Ref: SAF/ROM/DMI/REP/TPH/002

[RD.30] ROM SAF Validation Report: Offline Level 3 gridded products version 1.1, Ref: SAF/ROM/DMI/REP/GRD/002

[RD.31] Healy SB, Culverwell, ID (2015) A modification to the standard ionospheric correction method used in GPS radio occultation. Atmos. Meas. Tech. 8:3385-3393, doi:10.5194/amt-8-3385-2015



11 of 50

1.3 Acronyms and abbreviations

ATBD Algorithm Theoretical Baseline Document BUFR Binary Universal Form of Representation CDOP Continuous Development and Operations Phase (EUMETSAT) CDR Climate Data Record CGS Core Ground Segment (EPS) CHAMP CHAllenging Minisatellite Payload (Germany) COSMIC Constellation Observing System for Meteorology, Ionosphere, and Climate CT2 Canonical Transform type 2 DMI Danish Meteorological Institute ECF Earth-Centered, earth-Fixed ECI Earth-Centered Inertial ECMWF European Centre for Medium-range Weather Forecast EGM96 Earth Geopotential Model 1996. Standard model for geoide undulations and

gravity field, referenced to the WGS-84 ellipsoid ERA-I ERA-Interim (ECMWF global atmospheric reanalysis from 1979) ERA5 ECMWF Reanalysis 5th Generation EPS EUMETSAT Polar satellite System EUMETSAT EUropean organisation for the exploitation of METeorological SATellites FSI Full Spectrum Inversion GNSS Global Navigation Satellite System (generic name for GPS, GLONASS,

and similar future systems) GPS Global Positioning System (US) GRAS GNSS Receiver for Atmospheric Sounding (METOP instrument) GRM EUMETSAT acronym for the ROM SAF product identifier GSN Ground Support Network GTS Global Telecommunication System IEEC Institut d’Estudis Espacials de Catalunya (Spain) ICDR Interim Climate Data Record L1 GPS carrier frequency, 1575.42 MHz L2 GPS carrier frequency, 1227.6 MHz LC L phase Corrected, linear combination of L1 and L2 LEO Low Earth Orbit Met Office United Kingdom Meteorological Office METOP METeorological Operational Polar satellite (EPS/EUMETSAT) MSL Mean Sea Level (The geoid) netCDF Network Common Data Form NCO Numerically Controlled Oscillator NRT Near-Real Time NWP Numerical Weather Prediction PARF ROM SAF Product Archive and Retrieval Facility PCD Product Confidence Data POD Precise Orbit Determination PPF Product Processing Facility (EPS/EUMETSAT) PRD Product Requirements Document (ROM SAF) QC Quality Control



12 of 50

RD Reference Document (ROM SAF) RMDCN Regional Meteorological Data Communication Network (GTS in WMO

Region 6) RO Radio Occultation ROM SAF Radio Occultation Meteorology SAF (former GRAS SAF) ROPP Radio Occultation Processing Package (ROM SAF) SAF Satellite Application Facility (EUMETSAT) SNR Signal-to-Noise Ratio TPH Tropopause Height UMARF Unified Meteorological Archive and Retrieval Facility (EUMETSAT) WGS84 World Geodetic System 1984; standard Earth model ellipsoid WMO World Meteorological Organization 1D-Var 1 Dimensional Variational Analysis

1.4 Definitions

RO data products from the Metop and Metop-SG satellites and RO data from other missions are grouped in data levels (level 0, 1, 2, or 3) and product types (NRT, offline, CDR, or ICDR). The data levels and product types are defined below1. The lists of variables should not be considered as the complete contents of a given data level, and not all data may be contained in a given data level.

Data levels: Level 0: Raw sounding, tracking and ancillary data, and other GNSS data before clock correction and reconstruction;

Level 1A: Reconstructed full resolution excess phases, total phases, pseudo ranges, SNRs, orbit information, I, Q values, NCO (carrier) phases, navigation bits, and quality information;

Level 1B: Bending angles and impact parameters, tangent point location, and quality information;

Level 2: Refractivity, geopotential height, “dry” temperature profiles (Level 2A), pressure, temperature, specific humidity profiles (Level 2B), surface pressure, tropopause height, planetary boundary layer height (Level 2C), ECMWF model level coefficients (Level 2D), quality information;

Level 3: Gridded or resampled data that are processed from Level 1 or 2 data, and that are provided as, e.g., daily, monthly, or seasonal means on a spatiotemporal grid, including metadata, uncertainties and quality information.

Product types: NRT product: Data product delivered less than: (i) 3 hours after measurement (ROM SAF Level 2 for EPS); (ii) 150 min after measurement (ROM SAF Level 2 for EPS-

1Note that the level definitions differ partly from the WMO definitions: http://www.wmo.int/pages/prog/sat/dataandproducts_en.php

http://www.wmo.int/pages/prog/sat/dataandproducts_en.php



13 of 50

SG Global Mission); (iii) 125 min after measurement (ROM SAF Level 2 for EPS-SG Regional Mission);

Offline product: Data product delivered from less than 5 days to up to 6 months after measurement, depending on the requirements. The evolution of this type of product is driven by new scientific developments and subsequent product upgrades;

CDR: Climate Data Record generated from a dedicated reprocessing activity using a fixed set of processing software1. The data record covers an extended time period of several years (with a fixed end point) and constitutes a homogeneous data record appropriate for climate usage;

ICDR: An Interim Climate Data Record (ICDR) regularly extends in time a (Fundamental or Thematic) CDR using a system having optimum consistency with and lower latency than the system used to generate the CDR2.

1.5 Overview of this document

This document is organized as follows:

Chapter 1: Introduction.

Chapter 2: Contains a list and overview description the Level 1B and 2 data products.

Chapter 3: The radio occultation method, processing principles and GRAS and other RO instruments.

Chapter 4: Product description, describing all NRT, Reprocessed and Offline profile products with specifications.

Chapter 5: Format descriptions, giving details on the NetCDF format.

Chapter 6: Data quality, describing the current results of validating the products.

Chapter 7: Dissemination channels, describing the different means of receiving the products.

Chapter 8: Brief summary of the document.

Annexes A, B, C: Formats for NetCDF Header, Level 1, and Level 2.

2 (i) GCOS 2016 Implementation Plan; (ii) http://climatemonitoring.info/home/terminology/ 3 http://climatemonitoring.info/home/terminology/ (the ICDR definition was endorsed at the 9th session of the joint CEOS/CGMS Working Group Climate Meeting on 29 March 2018)

http://climatemonitoring.info/home/terminology/

http://climatemonitoring.info/home/terminology/

http://ceos.org/meetings/wgclimate-9/

http://ceos.org/meetings/wgclimate-9/



14 of 50

2. List of Level 1B and 2 data products The ROM SAF Level 1B and Level 2 data products are listed in Table 2-1 to 2-4. They include NRT data for time-critical applications, such as assimilation of data in NWP systems. They also include offline data that are generated on a regular basis for non-time-critical applications, Climate Data Records (CDRs) that were generated in a dedicated reprocessing activity using the same software versions throughout the length of the time series, as well as Interim Climate Data Records (ICDRs) that are generated on a regular basis with the same algorithms as the CDRs but using currently available input data. Note that the algorithms used to generate the NRT and offline data are allowed to evolve with time, to reflect the latest scientific development. The NRT data products (Table 2-1) and the offline data products (Table 2-2) are based on data from the Metop mission. There are currently four CDRs available in the ROM SAF archive, based on the four RO satellite missions Metop, COSMIC, GRACE, and CHAMP (Table 2-3). In addition, there is a single ICDR available, based on data from the Metop mission (Table 2-4).

Beside these ROM SAF products, the EUMETSAT Secretariat is disseminating full-resolution GRAS bending angles on EUMETCast and thinned GRAS bending angles in BUFR format on EUMETCast and GTS, cf. [RD.3] and [RD.4]. Thinned versions of the full-resolution EUMETSAT bending angles are also included in the ROM SAF Level 2 products for convenience.

Table 2-1 Level 1B and Level 2 ROM SAF near-real time data products

Product ID Product name Product acronym

Mission Dissemination

Format

GRM-01

NRT Refractivity Profile

NRPMEA

Metop-A

GTS EUMETCast

BUFR BUFR/netCDF

GRM-100

NRT Dry Temperature Profile

NDPMEA

Metop-A

GTS EUMETCast

BUFR BUFR/netCDF

GRM-02

NRT Temperature Profile

NTPMEA

Metop-A

GTS EUMETCast

BUFR BUFR/netCDF

GRM-03

NRT Specific Humidity Profile

NHPMEA

Metop-A

GTS EUMETCast

BUFR BUFR/netCDF

GRM-04

NRT Pressure Profile

NPPMEA

Metop-A

GTS EUMETCast

BUFR BUFR/netCDF

GRM-05

NRT Surface Pressure

NSPMEA

Metop-A

GTS EUMETCast

BUFR BUFR/netCDF

GRM-40


NTPMEB

Metop-B

GTS EUMETCast

BUFR BUFR/netCDF

GRM-102


NDPMEB

Metop-B

GTS EUMETCast

BUFR BUFR/netCDF

GRM-41


NTPMEB

Metop-B

GTS EUMETCast

BUFR BUFR/netCDF

GRM-42


NHPMEB

Metop-B

GTS EUMETCast

BUFR BUFR/netCDF

GRM-43


NPPMEB

Metop-B

GTS EUMETCast

BUFR BUFR/netCDF



15 of 50



Format

GRM-44


NSPMEB

Metop-B

GTS EUMETCast

BUFR BUFR/netCDF

GRM-60


NTPMEC

Metop-C

GTS EUMETCast

BUFR BUFR/netCDF

GRM-104


NDPMEC

Metop-C

GTS EUMETCast

BUFR BUFR/netCDF

GRM-61


NTPMEC

Metop-C

GTS EUMETCast

BUFR BUFR/netCDF

GRM-62


NHPMEC

Metop-C

GTS EUMETCast

BUFR BUFR/netCDF

GRM-63


NPPMEC

Metop-C

GTS EUMETCast

BUFR BUFR/netCDF

GRM-64


NSPMEC

Metop-C

GTS EUMETCast

BUFR BUFR/netCDF

Table 2-2 Level 1B and Level 2 ROM SAF offline data products



Format

GRM-08 Offline Bending Angle OBAMEA Metop-A Webpage BUFR/netCDF GRM-09 Offline Refractivity Profile ORPMEA Metop-A Webpage BUFR/netCDF GRM-101 Offline Dry Temperature Profile ODPMEA Metop-A Webpage BUFR/netCDF

GRM-10 Offline Temperature Profile OTPMEA Metop-A Webpage BUFR/netCDF GRM-11 Offline Specific Humidity Profile OHPMEA Metop-A Webpage BUFR/netCDF GRM-12 Offline Pressure Profile OPPMEA Metop-A Webpage BUFR/netCDF GRM-13 Offline Surface Pressure OSPMEA Metop-A Webpage BUFR/netCDF GRM-46 Offline Bending Angle OBAMEB Metop-B Webpage BUFR/netCDF GRM-47 Offline Refractivity Profile ORPMEB Metop-B Webpage BUFR/netCDF GRM-103 Offline Dry Temperature Profile ODPMEB Metop-B Webpage BUFR/netCDF

GRM-48 Offline Temperature Profile OTPMEB Metop-B Webpage BUFR/netCDF GRM-49 Offline Specific Humidity Profile OHPMEB Metop-B Webpage BUFR/netCDF GRM-50 Offline Pressure Profile OPPMEB Metop-B Webpage BUFR/netCDF GRM-51 Offline Surface Pressure OSPMEB Metop-B Webpage BUFR/netCDF GRM-66 Offline Bending Angle OBAMEC Metop-C Webpage BUFR/netCDF GRM-67 Offline Refractivity Profile ORPMEC Metop-C Webpage BUFR/netCDF GRM-105 Offline Dry Temperature Profile ODPMEC Metop-C Webpage BUFR/netCDF

GRM-68 Offline Temperature Profile OTPMEC Metop-C Webpage BUFR/netCDF GRM-69 Offline Specific Humidity Profile OHPMEC Metop-C Webpage BUFR/netCDF GRM-70 Offline Pressure Profile OPPMEC Metop-C Webpage BUFR/netCDF GRM-71 Offline Surface Pressure OSPMEC Metop-C Webpage BUFR/netCDF GRM-24 Tropopause height TPH Metop Webpage BUFR/netCDF



16 of 50

Table 2-3 Level 1B and Level 2 ROM SAF Climate Data Records



Format

Metop, 20061027 – 20161231

GRM-29-L1-B-R1 Reprocessed bending angle RBAMET Metop Webpage BUFR/netCDF GRM-29-L2-R-R1 Reprocessed refractivity profile RRPMET Metop Webpage BUFR/netCDF

GRM-29-L2-D-R1 Reprocessed dry temperature profile RDPMET Metop Webpage BUFR/netCDF

GRM-29-L2-T-R1 Reprocessed temperature profile RTPMET Metop Webpage BUFR/netCDF

GRM-29-L2-H-R1 Reprocessed specific humidity profile RHPMET Metop Webpage BUFR/netCDF

GRM-29-L2-P-R1 Reprocessed pressure profile RPPMET Metop Webpage BUFR/netCDF

GRM-29-L2-S-R1 Reprocessed surface pressure RSPMET Metop Webpage BUFR/netCDF

GRM-29-L2-C-R1 Reprocessed tropopause height RCHMET Metop Webpage BUFR/netCDF

COSMIC, 20060422 – 20161231

GRM-30-L1-B-R1 Reprocessed bending angle RBACO1 COSMIC Webpage BUFR/netCDF GRM-30-L2-R-R1 Reprocessed refractivity profile RRPCO1 COSMIC Webpage BUFR/netCDF

GRM-30-L2-D-R1 Reprocessed dry temperature profile RDPCO1 COSMIC Webpage BUFR/netCDF

GRM-30-L2-T-R1 Reprocessed temperature profile RTPCO1 COSMIC Webpage BUFR/netCDF

GRM-30-L2-H-R1 Reprocessed specific humidity profile RHPCO1 COSMIC Webpage BUFR/netCDF

GRM-30-L2-P-R1 Reprocessed pressure profile RPPCO1 COSMIC Webpage BUFR/netCDF

GRM-30-L2-S-R1 Reprocessed surface pressure RSPCO1 COSMIC Webpage BUFR/netCDF

GRM-30-L2-C-R1 Reprocessed tropopause height RCHCO1 COSMIC Webpage BUFR/netCDF

CHAMP, 20010901 – 20081004

GRM-32-L1-B-R1 Reprocessed bending angle RBACHA CHAMP Webpage BUFR/netCDF GRM-32-L2-R-R1 Reprocessed refractivity profile RRPCHA CHAMP Webpage BUFR/netCDF

GRM-32-L2-D-R1 Reprocessed dry temperature profile RDPCHA CHAMP Webpage BUFR/netCDF

GRM-32-L2-T-R1 Reprocessed temperature profile RTPCHA CHAMP Webpage BUFR/netCDF

GRM-32-L2-H-R1 Reprocessed specific humidity profile RHPCHA CHAMP Webpage BUFR/netCDF

GRM-32-L2-P-R1 Reprocessed pressure profile RPPCHA CHAMP Webpage BUFR/netCDF

GRM-32-L2-S-R1 Reprocessed surface pressure RSPCHA CHAMP Webpage BUFR/netCDF

GRM-32-L2-C-R1 Reprocessed tropopause height RCHCHA CHAMP Webpage BUFR/netCDF

GRACE, 20070301 – 20161231

GRM-33-L1-B-R1 Reprocessed bending angle RBAGRA GRACE Webpage BUFR/netCDF GRM-33-L2-R-R1 Reprocessed refractivity profile RRPGRA GRACE Webpage BUFR/netCDF

GRM-33-L2-D-R1 Reprocessed dry temperature profile RDPGRA GRACE Webpage BUFR/netCDF

GRM-33-L2-T-R1 Reprocessed temperature profile RTPGRA GRACE Webpage BUFR/netCDF

GRM-33-L2-H-R1 Reprocessed specific humidity profile RHPGRA GRACE Webpage BUFR/netCDF

GRM-33-L2-P-R1 Reprocessed pressure profile RPPGRA GRACE Webpage BUFR/netCDF

GRM-33-L2-S-R1 Reprocessed surface pressure RSPGRA GRACE Webpage BUFR/netCDF

GRM-33-L2-C-R1 Reprocessed tropopause height RCHGRA GRACE Webpage BUFR/netCDF



17 of 50

Table 2-4 Level 1B and Level 2 ROM SAF Interim Climate Data Records



Format

Metop, 20170101 –

GRM-29-L1-B-I1 ICDR bending angle IBAMET Metop Webpage BUFR/netCDF GRM-29-L2-R-I1 ICDR refractivity profile IRPMET Metop Webpage BUFR/netCDF

GRM-29-L2-D-I1 ICDR dry temperature profile IDPMET Metop Webpage BUFR/netCDF

GRM-29-L2-T-I1 ICDR temperature profile ITPMET Metop Webpage BUFR/netCDF

GRM-29-L2-H-I1 ICDR specific humidity profile IHPMET Metop Webpage BUFR/netCDF

GRM-29-L2-P-I1 ICDR pressure profile IPPMET Metop Webpage BUFR/netCDF

GRM-29-L2-S-I1 ICDR surface pressure ISPMET Metop Webpage BUFR/netCDF

GRM-29-L2-C-I1 ICDR tropopause height ICHMET Metop Webpage BUFR/netCDF



18 of 50

3. The Radio Occultation method 3.1 Overview

Products obtained from the radio occultation (RO) measurements consist of bending angles and vertical profiles of refractivity, dry temperature, temperature, pressure, and humidity as functions of height, plus tropopause height. The various profiles (e.g. temperature) are obtained from the excess phases of radio signals travelling through the atmosphere along horizontal paths (see Figure 3-1). The signals are emitted from the GNSS (Global Navigation Satellite System) satellites orbiting some 20,000 km above the Earth surface and received by an instrument (e.g. the GRAS instrument) on board a LEO (Low-Earth Orbit) satellite. The GNSS radio signals scan the atmosphere vertically until they are occulted by the Earth (setting occultation) or from the moment they appear behind the Earth (rising occultation). The first step is to compute the bending angle of the signal as an integrated measure along the entire signal path. The refractivity at a given so-called tangent point is then derived through an inversion of the bending angle. The fact that parts of the signal paths travel through the same levels of the atmosphere causes the profile errors to be vertically correlated. This statistical correlation will be separately specified in an error covariance matrix. It should be noted that in general the profile will not be given along a straight, vertical line but rather along a slightly curved, slanted line such that the deviation of the topmost point relative to the point closest to the Earth (the so-called tangent point drift) can be considerably larger than 100 km at high latitudes. The profiles will cover the Earth evenly but the locations of the individual profiles vary from profile to profile and locations are not repeatable. Also, note that the profiles are generated at random times (i.e. not at synoptic times). For more details see [AD.4], [RD.3], [RD.4].

Figure 3-1. Schematic representation of the data (red line) observed by the GRAS instrument on Metop during an occultation.



19 of 50

3.2 Benefits of radio occultation profiling

Despite the relatively poor horizontal sampling (mean spacing) of RO data from a single instrument – but still better than the global coverage for radiosondes – and the inherent horizontal line-of-sight resolution of about 250 km, the system has several very significant benefits:

- High stability – both in time for one instrument and inter-instrument, leading to very stable long-term data for climate applications

- High accuracy – better than 1K over much of the middle atmosphere (5-30 km) - High vertical resolution – of order 200 m or better in the lower troposphere –

comparable to radiosondes and significantly superior to current vertical passive sounders

- All weather capability – GNSS signals are virtually unaffected by clouds and precipitation

- Global coverage

The fundamental measurement upon which all the GNSS-RO products are based is a time delay and a measurement that is fundamentally SI traceable and stable. The general characteristics of the RO method make it a quite complementary observing system within the WMO's WWW programme. The potential of the RO technique has been amply demonstrated with the CHAMP, COSMIC and Metop missions. E.g. the single GRAS instrument on Metop generates approximately 500–650 profiles per day, depending on the actual number of GNSS satellites. RO data provides bias free information to Numerical Weather Prediction (NWP), which also helps to anchor biased (satellite) data.

3.3 Satellite missions and time coverage

The RO missions, their respective satellites, and the time periods included in the generation of the bulk products are given in Table 3-1. The ROM SAF CDR v1.0 spans the period September 1 2001 to December 31 2016. The ICDR-v1.0 and Offline-v1.0 data, from January 1 2017, and ICDR-v1.1 and Offline-v1.1 data, from August 1 2019 are based on Metop data only. It should be noted that not all COSMIC satellites were operational during the whole time period, especially towards the end of the period. For GRACE, both RO instruments are never in operation at the same time. Metop-A was launched in October 2006 and NRT RO operation started in April 2008. Metop-B was launched in September 2012, and NRT RO operation started in October 2012. Metop-C was launched in November 2018 and NRT RO operations started in March 2019. The very early data from CHAMP (until September 2001) has not been included.



20 of 50

Table 3-1 Missions, satellites, and time periods included in the generation of ROM SAF Level 1B and Level 2 CDRs, and in the generation of the corresponding ICDR and offline data.

Mission Metop COSMIC CHAMP GRACE

Satellites

Metop-A Metop-B Metop-C

FM1, FM2, FM3, FM4, FM5, FM6

CHAMP

GRACE-A GRACE-B

Included in CDRs Oct’06 – Dec’16 Apr’06 – Dec’16 Sep’01 – Sep’08 Mar’07 – Dec’16 Included in ICDR and offline data Jan’17 – N/A N/A N/A

3.4 Characteristics of GRAS and other RO instruments

The main objective of the GRAS instrument (and other RO instruments) is the measurement of the excess phase of signals from GNSS satellites as they are refracted by the atmosphere. Excess phase, measured as the phase change in the signal carrier phase, depends on the refractive index of the atmosphere, which is a function of electron density, temperature, pressure, and humidity. An RO instrument provides carrier phase measurements for the occultation mission and also for the navigation mission (top-side antenna). EUMETSAT’s EPS-CGS facility is responsible for GRAS Level 0 to Level 1B NRT processing (the products are subsequently disseminated to the ROM SAF), whereas the ROM SAF itself is responsible for the offline processing from GRAS Level 1A to higher levels. The ROM SAF is responsible for processing the NRT Level 1B data into NRT Level 2 products. The sampling rate of the carrier phase, pseudo-range, signal amplitude, occultation and navigation measurement is configurable. The occultation measurement is nominally sampled at 50 Hz in closed-loop mode and 1000 Hz in GRAS raw sampling mode. Bending angles are provided for heights above the Earth's surface ranging from 80 km down to, or close to, the surface (for both setting and rising occultations). The bending angle accuracy requirement is to be better than 1 µrad or 0.4% (whatever is larger). The impact parameter localisation in Earth co-ordinates is required to be better than 0.01° in longitude and latitude, and better than 6 metres in altitude. With the nominal GPS constellation the GRAS instrument generates some 500 occultations per day, globally distributed, although the current 31 GPS satellites yield as many as 650 nominal daily occultations. Typically other instruments produce a similar amount or less occultations per day.

3.5 Processing to Level 2 products

Figure 3.2 shows schematically the data flow for ROM SAF NRT Level 2 products. The processing up to NRT bending angle (Level 1B) is performed at the EUMETSAT Secretariat for GRAS/Metop data. Figure 3.3 shows schematically the data flow for ROM



21 of 50

SAF offline Level 1B and 2 products. The processing is based on Level 1A input data from EUMETSAT Secretariat and other data providers. UCAR delivers Level 1A data for the COSMIC, CHAMP and GRACE missions. The Level 3 processing is shown for reference. The processing steps involved in generation of reprocessed CDR data products are identical to the steps in the offline processing. The ATBDs [RD.6, RD.17, RD.17, RD.23, RD.24] specify the details of the algorithms which are used to process the occultation data. The two GPS radio frequencies (L1 and L2) received by GRAS at the Metop satellite are characterised by their amplitude and phase values. The bending angle profiles are obtained using the positions and velocities of the GPS and Metop satellites. The bending angle profiles are subject to a correction in order to eliminate the effect of the ionosphere on the signals (LC). This step is done using the standard linear combination of bending angles, and may include a higher order residual ionospheric correction (also known as the “kappa” method) [RD.31]. In the case of single ray propagation the phase contains all the necessary information in order to derive the bending angle whereas in the case of multiple ray propagation (multipath), caused by strong vertical gradients in the atmosphere, both the amplitude and the phase are needed to obtain a bending angle profile free of multipath artefacts. The processing using both the phase and the amplitude data in order to solve for the multipath propagation is based on canonical transform (or wave-optics) methods such as CT2 and FSI (see e.g. [RD.7, 13]). The index of refraction (from which the refractivity is derived), is obtained from a statistically optimized bending angle profile through the use of an Abel transform inversion method, cf. [RD.7] and [RD.10]. In order to get an estimate for the temperature, pressure and humidity (Level 2B), some ancillary data are needed. For the ROM SAF products we use as ancillary data profiles of temperature, humidity and surface pressure from ECMWF forecasts (for NRT processing) or forecast from reanalysis datasets (for offline processing and CDR or ICDR generation), appropriate to the time and location of the occultation (interpolated bi-linearly in the horizontal on model levels, see [RD.8] about model levels). This set of ancillary data ('background' or 'first-guess') in combination with the refractivity is then used in a 1D-Var algorithm in order to simultaneously estimate the temperature, humidity and pressure profiles, together with surface pressure. The solution is constrained by the assumption that the atmosphere is in hydrostatic equilibrium. Note that unique humidity profiles cannot be obtained from radio occultation measurements without using some source of ancillary information on temperature. This problem is referred to as the “water vapour ambiguity”. The 1D-Var procedure overcomes the water vapour ambiguity and takes full account of the observation and model uncertainties in an optimal way. The 1D-Var method simultaneously produces temperature, specific humidity and surface pressure. See [RD.13] for more detailed information. It is possible to calculate temperature and pressure directly from the refractivity (i.e. without the use of background data) using the ideal gas law under the assumption of no water vapour in the atmosphere. These “dry” products (dry temperature and dry pressure) are good estimates of the measurands in the stratosphere, but not in the troposphere [RD.23]. The dry products are less precise than the wet products, but the wet products may



22 of 50

be suspected to contain biases or in-accuracies inherited from the background model. For that reason the dry temperature is considered as a better reference for some users.

Figure 3.2. Schematic drawing showing the processing steps for the ROM SAF NRT Level 2 products. The processing is based on Level 1B input data from EUMETSAT Secretariat.



23 of 50

Figure 3.3. Schematic drawing showing the processing steps for the ROM SAF offline Level 1B and 2 products. The processing is based on Level 1A input data from EUMETSAT Secretariat and other data providers (e.g. UCAR). The Level 3 processing is shown for reference. The processing steps involved in generation of CDRs and ICDRs are identical to the steps in the offline processing.



24 of 50

4. Data product description 4.1 Overview of ROM SAF Level 1B and 2 data products

In this and the following sections, we describe all the ROM SAF Level 1B and Level 2 data products. The products of the ROM SAF operational system are targeting different types of user groups. NRT products are targeting National Meteorological Centres, and comparable regional or independent centres. These users will receive the products with a timeliness of 3 hr. The offline data products are targeting a diverse set of users, including meteorologists and atmospheric scientists, while the CDRs and the ICDRs are primarily targeting climate and atmospheric scientists and anyone interested in climate monitoring. The data file formats do not depend on which users they are intended for. Instead different types of products (see Section 1.4) and means of delivery are designed for different user groups. The NRT data are delivered through guaranteed performance channels and under operational timeliness restrictions, while the offline data, CDRs, and ICDRs are made available through a variety of channels including the ROM SAF web page http://www.romsaf.org. Each product type is divided into levels, as listed in this table:

Data product Level 1 Level 1A SNR, excess phases and POD data as function of time

Level 1B Bending angle as function of impact parameter Level 2 Level 2A Refractivity as function of height

Level 2A Dry temperature as function of height Level 2B Temperature, humidity, and pressure as a function of height Level 2C Surface pressure Level 2C Tropopause Height Level 2D Additional data describing the vertical model level structure

Table 4-1 Data product level descriptions. Product files on each level contain further additional data, appropriate to the specific type of product. Level 2A contains geometric and geopotential height, Level 2B contains only geopotential height and level 2D is described in romsaf_pum_grd_v20.pdf. See Appendices for details. Further descriptions are available in [RD.13-15]. The operational ROM SAF system consists of the data retrieval and processing system and the archival system, both located at the DMI. The input data for NRT processing is level 1B data received from EPS-CGS through EUMETCast. Auxiliary data sources are forecasts and analyses received from ECMWF.

4.1.1 NRT products The ROM SAF’s primary data product is the Level 2 products processed in near-real time (NRT) within 3 hours of observation. Since this time constraint may mean that processing



25 of 50

is simplified, and some ancillary data may not be available in time, NRT products may not represent the optimum possible quality, although it will still meet user requirements for NRT data. The main parameters are:

• Refractivity profiles • Temperature profiles • Pressure profiles • Specific humidity profiles • Surface pressure

In addition, a thinned-out version of the EUMETSAT Level 1B (ionosphere corrected) bending angle is included in the Level 1B NRT data products. Various supporting data and selected parameters (with some post-processing applied) are included with the Level 2 NRT Sounding Products (see below).

4.1.2 Offline data, CDRs, and ICDRs The parameters in the offline, CDR, and ICDR data products are identical to those contained in the NRT products. The CDRs have been processed to different specifications than the NRT and offline products, the major differences being the use of reprocessed Level 1A data. Reprocessed Level 1A data is produced with the optimum algorithms and post-processed GPS and Metop precise orbital determination (POD) information and the inclusion of other auxiliary data, which may not have been available on the timescale of the NRT Products. CDRs (and also ICDRs) are disseminated as data records covering longer time periods, possibly including other missions than those provided by EUMETSAT. Offline products on the other hand contain only EUMETSAT missions, and are intended to incorporate new scientific developments.

• Bending angles • Refractivity profiles • Dry Temperature profiles • Temperature profiles • Pressure profiles • Specific humidity profiles • Surface pressure • Tropopause Height

4.1.3 Supporting data ROM SAF NRT and offline data products also include (but are not limited to) the following supporting parameters:

• LEO and occulting GPS satellite identifiers • Horizontal location (latitude, longitude) • Vertical location (height above geoid/MSL, geopotential height, pressure

level)



26 of 50

• Date and time • Quality information (estimated uncertainties, QC flags) • POD information • Radius of curvature information • ROM SAF software version

Note that the ROM SAF also includes ‘raw’ parameters such as signal-to-noise ratio, excess phase, and uncorrected bending angles in its archived products.

4.1.4 Error-covariance matrices The error covariance matrix is a product that specifies the correlations in the observation and background errors between all possible pairs of vertical observation data levels. It is given as 2-dimensional arrays, of size NxN, where N is the number of vertical levels in the sounding product.



27 of 50

4.2 Level 1B and 2 atmospheric profile data products

4.2.1 General This section and following sections contain a detailed description of all parameters in the ROM SAF NRT, offline, CDR, and ICDR (Level 2) sounding products. For format descriptions, please refer to Annex B and Annex C. Note that all accuracies given within Section 4.2 are target accuracies from the PRD [AD.3], i.e. not reflecting the actual status of the products. Please also note that “target accuracy” is understood as standard deviation of the difference between the retrieval and an NWP model in this context. For each parameter, the description includes the output quantities (e.g. units and ranges of values). Data in the form of profiles are provided as a function of height (height above MSL/geoid and geopotential height) and pressure, or as a function of time, consistent with the user requirements in [RD.2]. All product profiles are given in ascending order, regardless of whether the occultation was a setting or rising measurement. The product domain is global, and ranges from the surface to a maximum of 80 km. The height range of individual profiles produced by the SAF critically depends on the output of the RO instrument and processing up to Level 1A/1B within the CGS. The geographical and temporal coverage of the ROM SAF products are limited only by the characteristics of the radio occultation instrument and not by the processing algorithms. The following specifications are common to all NRT and offline data parameters: Delay from observation to start of delivery to users:

>95% within 3 hours (NRT) >98% within 30 days (Offline)

PRD-1-06, [AD.3]

Horizontal domain: Global See Annex A of the PRD [AD.3]

Horizontal sampling: All available occultations See Annex A of the PRD [AD.3]

The algorithms used to process the CGS Level 1A/1B products to ROM SAF Level 2 Products can be found in [RD.6], [RD.6], [RD.17], [RD.23] and [RD.24]



28 of 50

4.2.2 Bending angle The bending angle is a function of impact parameter (and part of the level 1B data set). Level 1B data is sampled at 50-100 m, depending on altitude. This is an offline, CDR, and ICDR data product but a thinned version of the EUMETSAT Central Facilities bending angle is included in the NRT Level 2 products. See [RD.24,19,21]; Quantity Values Remarks Units radians (rad) Range -0.0001 to 0.05 rad Vertical sampling 50-100 m Depending on altitude

Target accuracy 0.4% or 1 μrad

whichever is greater. Level 1b requirement. Here “accuracy” is understood as the STDV of the difference between the retrieval and an NWP reference.

Figure 4-1 Examples of GRAS bending angles as function of impact height, from three different Metop occultations, 15/12 2016. Left: Observations (full curves) and background model (dashed curves). Legend shows positions in degrees. Right: Observation minus background for the same three occultations, with corresponding colours.



29 of 50

4.2.3 Refractivity profile This parameter is a profile and contains the neutral refractivity as a function of height level (above geoid/MSL and geopotential, see Section 4.2.10) above a given location on the Earth. See [RD.6,19,21]. Quantity Values Remarks Units Refractivity units (N) Neutral atmosphere Domain 0-50 km Surface to ~1 hPa Range 0-450 N-units Vertical resolution 150-250 m Target accuracy

0-5 km 5-30 km

30-50 km

0.6%-2% 0.6% 0.03 N-units

Here “accuracy” is understood as the STDV of the difference between the retrieval and an NWP reference.

Figure 4-2. Examples of GRAS refractivity profiles, from three different Metop occultations, 15/12 2016. Left: Observations (full curves) and background model (dashed curves). Legend shows positions in degrees. Right: Observation minus background for the same three occultations, with corresponding colours.



30 of 50

4.2.4 Dry temperature profile This parameter is a profile that contains the dry temperature as a function of height level (above geoid/MSL and geopotential, see Section 4.2.10) above a given location on the Earth. See [RD.23,21]. Quantity Values Remarks Units Kelvin (K) Domain 0-50 km Surface to ~1 hPa Range 0-300 K Vertical resolution 150-250 m Target accuracy

0-5 km 5-20 km

20-40 km

2-1 K 1 K 1-10 K


Figure 4-3 Examples of GRAS dry temperature profiles and background physical temperature, from three different Metop occultations, 15/12 2016. Observations (full curves) and background model (dashed curves). Legend shows positions in degrees.



31 of 50

4.2.5 Temperature profile This parameter is a profile and contains the atmospheric temperature as a function of height above a given location on the Earth. The heights are given on fixed model levels, see [RD.17,15,20,22 ] for details. Quantity Values Remarks Units Kelvin (K) Domain 0–50 km Surface to ~1 hPa Range 150–350 K Vertical resolution 250-500 m depends on background

pressure levels Target accuracy

0–30 km 30–50 km

Offline 1 K 1–2 K


Figure 4-4. Examples of GRAS temperature profiles, from three different Metop occultations, 15/12 2016. Left: Observations (full curves) and background model (dashed curves). Legend shows positions in degrees. Right: Observation minus background for the same three occultations, with corresponding colours.



32 of 50

4.2.6 Humidity profile This parameter is a profile and contains the atmospheric water vapour content, as specific humidity, as a function of height (on model levels) above a given location on the Earth. See [RD.17,20,22]. Quantity Values Remarks Units g/kg Specific humidity Domain 0–12 km Range -10–50 g/kg Vertical resolution 250-500 m depends on background

pressure levels

Target accuracy

Offline 35%

Here “accuracy” is understood as the STDV of the difference between the retrieval and an NWP reference

Figure 4-5. Examples of GRAS specific humidity profiles, from three different Metop occultations, 15/12 2016. Left: Observations (full curves) and background model (dashed curves). Legend shows positions in degrees. Right: Observation minus background for the same three occultations, with corresponding colours.



33 of 50

4.2.7 Pressure profile This parameter is a profile and contains the atmospheric pressure as a function of height (on model levels) above a given location on the Earth, for the same set of heights as the derived temperature and humidity values. See [RD.17,20,22]. Quantity Values Remarks Units hectoPascal (hPa) Domain 0–50 km Surface to ~1 hPa Range 0.01–1100 hPa Vertical resolution 250-500 m depends on background

pressure levels

Target accuracy

(Offline) a) 20% b) 0.01 hPa c) 0.6 hPa

* whichever is greatest of (a) and (b) - but not greater than (c); Here “accuracy” is understood as the STDV of the difference between the retrieval and an NWP reference

Figure 4-6 Examples of GRAS pressure profiles, from three different Metop occultations, 15/12 2016. Left: Observations (full curves) and background model (dashed curves). Legend shows positions in degrees. Right: Observation minus background for the same three occultations, with corresponding colours.



34 of 50

4.2.8 Surface pressure The surface pressure is a scalar product. See [RD.17,20,22]. Quantity Values Remarks Units hectoPascal (hPa)

Domain Surface Horizontal location nominally at the location where the straight line connecting GPS and Metop grazes the surface (WGS 84 ellipsoid).

Range 300–1100 hPa Vertical resolution N/A

Accuracy

(Offline) 0.6 hPa

Here “accuracy” is understood as the STDV of the difference between the retrieval and an NWP reference

4.2.9 Tropopause heights The tropopause height is derived from the dry temperature lapse rate using the official WMO tropopause definition. See [RD.17,23]. Quantity Values Remarks Units Meters (m)

Domain global Horizontal location nominally at the location where the straight line connecting GPS and Metop grazes the surface (WGS 84 ellipsoid).

Range 5-30 km Vertical resolution N/A

Target accuracy (Offline) 1 km In a limited tropical range.

4.2.10 Heights This parameter is the vertical coordinate for the refractivity and of the pressure levels for the retrieved temperature and humidity profiles. Height values are derived from the Level 1B impact parameter and local radius of curvature at the location of the occultation and the refractivity profile. The heights are provided in several reference frames. Profiles are given in order of increasing height (both for setting and rising occultations). Quantity Values Remarks Units metres (m)

geopotential metres (gpm)

(a) Heights above MSL (referenced to geoid EGM96)

(b) Geopotential heights Domain -0.15 – 50 km Surface to ~1 hPa Range -150 – 50,000 m Vertical resolution N/A Accuracy N/A Independent coordinate.



35 of 50

4.2.11 Location This parameter is the horizontal coordinate for the refractivity, dry temperature, temperature and humidity profiles, surface pressure, tropopause height and local radius of curvature. A pair of latitude and longitude value is given for each point in the profile, as the tangent point drift can be considerably more than 100 km during an occultation. Quantity Values Remarks Units degrees (°) of latitude and

longitude Geodetic latitude

Range Latitude: ±90° Longitude: ±180°

Positive in N. Hemisphere positive east of Greenwich

Vertical resolution N/A Accuracy 0.01° ~1 km

4.2.12 Quality information The ROM SAF Products contain the following quality information. There is an overall (scalar) quality indicator (PCD) , and then a quality flag with a value for each altitude. In the reprocessing, the latter quality indicator is not used.

- Boolean flags showing the results of quality tests (’Product Confidence Data’, see also Appendix A ). These results are independent, e.g. a profile with a flag indicating nominal processing of refractivity can have a flag indicating non-nominal bending angle processing; Flags include:

Bit Variable Description Meaning If unset (0) set (1)

1 PCD_summary Quality nominal non–nominal 2 PCD_offline Product type NRT off line 3 PCD_rising Occultation type setting Rising 4 PCD_phase Excess phase processing nominal non–nominal 5 PCD_bangle Bending angle processing no nominal non–nominal 6 PCD_refrac Refractivity processing nominal non–nominal 7 PCD_met Meteorological processing non–nominal 8 PCD_open loop Open Loop not used Used 9 PCD_reflection Surface reflections detected no Yes 10 PCD_l2 signal L2P or L2C GNSS signal used L2P L2C 11 PCD_reserved 11 Reserved 12 PCD_reserved 12 Reserved 13 PCD_reserved 13 Reserved 14 PCD_bg Background profile nominal non–nominal 15 PCD_occultation Profile type observed Background 16 PCD_missing PCD missing; bits 1–15… valid Invalid

- A quality indicator value is a 'percentage confidence' value derived from a combination of other values: (0=bad, 100=good);



36 of 50

- Estimated RMS error values for all derived parameters (refractivity, pressure, temperature, humidity profiles and surface pressure);

- Trace-back information on the processing algorithms used (indicating nominal, FSI and/or canonical transform, 1D-Var configuration, data sources etc);

4.3 Supporting data

Supporting data include Level 1A/1B parameters produced by the EPS-CGS (for NRT) or by the ROM SAF (for offline data, CDRs, and ICDRs), though they may have been post-processed within the Level 2 processor to a form more suitable for most end-users. Users requiring the unprocessed support data should access the Level 1A data directly.

4.3.1 Satellite orbits (POD) This parameter is a sub-set of the LEO and GNSS satellite state vectors (POD locations and velocities), as a time-series, produced by the CGS as Level 1A GRAS products, see [RD.4] for details. Positions Quantity Values Remarks Units metres (m) Range ±30,000 km (GNSS)

±10,000 km (LEO)

Precision 0.01 m Vertical resolution 5 Hz

Accuracy 0.2 m (NRT) 0.1 m (Offline)

Level 1b requirement. See [RD.4] for details.

Velocities Quantity Values Remarks Units metres per second (m.s-1) Range ± 5 km.s-1 (GNSS)

± 10 km.s-1 (LEO)

Precision 0.01 x 10-3 m.s-1 Vertical resolution 5 Hz

Accuracy 0.2 x 10-3 m.s-1 (NRT) 0.1 x 10-3 m.s-1 (Offline)

Level 1b requirement. See [RD.4] for details.

The tables for POD are based on requirements for CGS Level 1b data and given as radial position w.r.t the Earth and absolute velocities. The specification of two coordinate systems reflects the use of these data. Velocity POD is required for GNSS-LEO Phase and Doppler determination, which is independent of Earth-based coordinates, so the ECI coordinate system is most appropriate. Position POD is used for e.g. GNSS-LEO ray-tracing using Earth-centred coordinates, where the ECF system is most appropriate.



37 of 50

4.3.2 Local radius of curvature, azimuth, and geoid undulation This parameter is taken from the Level 1b product. Time and location of the local radius of curvature is specified. In addition, the local centre of curvature offset from the Earth's centre is given, together with the azimuth angle of the plane of occultation and the local geoid undulation. Radius of Curvature Quantity Values Remarks Units metres (m) RoC value at one specified

lat/lon representative for the entire profile

Range 6250–6450 km Precision <1 m Vertical resolution N/a Accuracy <5 m

Radius of Curvature Offset Quantity Values Remarks Units metres (m) RoC offset as (X,Y,Z) ECF

coordinates Range ±10 km in each dimension Precision <1 m Vertical resolution N/a Accuracy <1 m

Azimuth Angle Quantity Values Remarks Units Degrees wrt True North

(degT), positive clockwise Azimuth angle of GNSS to LEO line of sight

Range 0-360 Precision 0.1 deg Vertical resolution netCDF files: 50 Hz

BUFR files: One interpolated value for each of the 247 model levels, cf. [RD.5]

Accuracy 0.5 deg Geoid Undulation Quantity Values Remarks Units metres (m) Geoid height (difference

between local geoid (EGM96) and ellipsoid (WGS-84))

Range ±150 m Precision 0.1 m Vertical resolution N/a Accuracy 1 m



38 of 50

4.3.3 Background profiles from ECMWF The ROM SAF products contain meta data on the background (first-guess) profiles of temperature and humidity, which are used to constrain the retrievals. For NRT data this is from an ECMWF forecast and offline data, CDRs, and ICDRs, the background profiles are from ERA-I or ERA5. Background meta-information include: source, validity date & time, forecast period. Background data are not included, except in the NRT BUFR products. The extracted co-located profiles from the background data are made available for research purposes on PARF the day after the measurement.

4.3.4 Other data The ROM SAF products also contain the following meta-data:

- Identifier of receiving LEO satellite (e.g. Metop-A) - Identifier of transmitting GNSS satellite (e.g. GPS-03) - Identifiers for POD type and source (e.g. Predicted & EPS-CGS) - Timestamps of start of occultation and of processing

Data products also contain estimated a priori uncertainty for all derived parameters, such as refractivity, pressure, temperature and humidity. The META data from Level 1 and from the background profile is inherited downstream in Level 2A and 2B, but not in Level 3. Note that the ROM SAF does not include ‘raw’ parameters such as signal-to-noise ratio, excess phase, Doppler or uncorrected bending angles in its Level 2 NRT products. Users wishing to start processing at this level should access the PARF archive or obtain Level 1b products from UMARF.

4.4 Validation data

Validation data consist of summary statistical information on the reliability and quality of the Sounding Products. Validation data include:

- Analysis of observation delay (time differences between the observation and the start of dissemination of the Level 2 sounding product to users from the SAF);

- Analysis of availability (number of Level 2 sounding products made available to users relative to the number of Level 1b occultations received by the SAF from the CGS);

- Analysis of bending angle, refractivity and dry temperature quality (differences in Level 2 values of refractivity from refractivity synthesised from an NWP model and/or other observational data at the same location, expressed as bias and rms or standard deviation);

- Analysis of temperature, humidity, pressure and surface pressure (and tropopause height) quality (differences in Level 2 temperature and humidity profiles and



39 of 50

surface pressure from equivalent NWP model values and/or other observational data, including high quality balloon borne reference measurements, expressed as bias and rms or standard deviation). Validation is done globally and on the full vertical domain of the product, limited only by the availability of the comparison data.

The major source of comparison data is operational NWP global and regional forecast and analysis. Because of the random time of RO data, comparisons use short-period NWP forecasts in order to minimise the time differences to not more than 3 hours for NRT. The NWP fields at the appropriate time are bi-linearly interpolated in the horizontal to the location of the RO data. Two different NWP systems may be used so that temporal differences – such as drifts or jumps in the bias times series – can be attributed to model or RO problems. If there are sufficient quantities of other observational data with the necessary quality and other characteristics collocating with the RO data at the same times and locations (within defined limits, like 100 km and 3 hours) – such as radiosondes, ground-based remote sensing and passive satellite sounding (including other RO missions) – then the Validation Products for Offline Sounding Products use these sources too. Validation Products show the statistics for:

- Global area and whole vertical domain (‘bottom line’ statistics) - Standard vertical levels - Latitude bands (NH, Tropics, SH) - Surface (Land/Sea) - Regional zones (e.g. 5° x 5° latitude/longitude boxes) for mapping - Occultation type (Rising/Setting) - Daily for NRT data - Monthly for NRT and Offline data alone or in selected combinations.

These products are available to users in the form of:

- Summaries in the form of single or tabulated sets of numbers; - Graphical representations – for example histogram of delays, time-series and

geographical maps of RMS temperature differences at selected levels; Validation products are not actively disseminated, but are automatically updated and posted on a regular basis on the ROM SAF web site. The URL for the NRT monitoring is: http://www.romsaf.org/monitoring The URL for the reprocessing validation is http://www.romsaf.org/re1 During the routine Operational Phase, where operational GRAS RO data will be assimilated at ECMWF, it will be normal to validate against 'background' – i.e. a short-term ECMWF forecast (typically 6 hours). Although the RO information will still have some influence on the forecast, this is not as direct as with an analysis, which has used the same RO information at the same time and place, and thus naturally is not appropriate for

http://www.romsaf.org/monitoring

http://www.romsaf.org/re1



40 of 50

independent validation. The use of a short-term forecast for validation is common for all data types, including satellite data, and this methodology also forms the basis for the NWP SAF satellite monitoring systems. The level 2B and level 2C products are validated against ECMWF analysis which has assimilated the RO data and may be viewed as a more advanced state vector retrieval based on the same RO data.

4.5 Error covariance matrices

The error covariance matrix is a data product that specifies the correlations in the observation errors between all possible pairs of vertical observation data levels. It is given as a 2-dimensional array, of size NxN where N is the number of vertical levels in the sounding product. There are two basic Error-Covariance Matrix Products:

- Covariance matrix for NRT Sounding Products - Covariance matrix for Offline Sounding Products

Each matrix is provided in one or more versions reflecting potential variations with geographical areas (e.g. latitude) and with season. The 'Day 1' product is a single time-invariant, globally-applicable matrix; further matrices will be provided if and when further analysis of GRAS operational data shows their necessity. The matrices are not actively disseminated, but are available for download from the ROM SAF web site. These files are expected to change only infrequently (if at all) after the Metop commissioning period, as the statistics become more stable, and are included with the ROPP software package. The latest versions will always be available via the ROM SAF web site. The Error-Covariance Matrices have the form:

�𝐸𝐸11 ⋯ 𝐸𝐸1𝐿𝐿⋮ ⋱ ⋮𝐸𝐸𝐿𝐿1 ⋯ 𝐸𝐸𝐿𝐿𝐿𝐿

�

Where L is the number of vertical profile levels. The Eij values represent the error co-variances between levels i and j for the off-diagonals (i ≠ j) and the diagonal values (i = j) are the error variances at each level.



41 of 50

5. Format descriptions 5.1 Introduction

The ROM SAF products come in two different formats, BUFR and netCDF, cf. [RD.11].

5.2 BUFR format

The BUFR format is described in [RD.5]. The common file naming principle is given Section 5.3.2.

5.3 netCDF format

The data format for the ROM SAF products is netCDF.

5.3.1 Structure The netCDF files in the ROM SAF system have the following structure: • A common set of attributes for all kinds of data, with general information about the data • A dataset for the variables values • Additional datasets for metadata (e.g. quality flags)

5.3.2 File format The ROM SAF products follow the ROPP data format structure (see [RD.9]). An overview of the structure of the netCDF product files is depicted in Figure 5-1. In the ROPP format all parts except the header are optional. Detailed information of each variable available in the netCDF product files can be found in Appendix A, B, and C, taken from [RD.8].

Figure 5-1 Overview of the ROPP netCDF file structure



42 of 50

5.3.3 File names Files for NRT data The NRT file name version is used in the NRT processing chain, for both level 1 and level 2 products. It is made of 6 fields separated by underscores with structure:

<TYPE><DATE>_<TIME>_<MISSION_ID>_<OCC_ID>_<MODE><SWVER>_<FREE>.<EXT> Explanation

• <TYPE> is one of “atm”, “bgf”, “bgn”, “bgo”, “bga”, “dis”, “occ” or “wet” (see Table below). <DATE>_<TIME> is the start date and time of the observation, as YYYYMMDD_HHMMSS.

• <MISSION_ID> is the EPS name of the observing satellite. Other names will be possible at a later date, when more missions are included.

• <OCC_ID> is the occultation id (EPS format which includes id of occulting satellite). As for <MISSION>, other names are possible for other missions.

• <MODE> is the processing mode and one of “N”, “T”, “V” (see below). • <SWVER> is a four digit code denoting the system software version (e.g. “0372”). • <FREE> is a free field, “XXXX”. • <EXT> is the extension,

o “nc” for NetCDF. o “bin” for BUFR for disseminated products.

Example An example of an NRT atm file name from 29 January 2011:

atm20110129_021532_M02_2020404609_N0018_XXXX.nc Files for offline data, CDRs, or ICDRs The level 1 and level 2 file name version is the name that is used in the corresponding processing chain. It is made of 8 field separated by underscores, with structure: <TYPE>_<DATE>_<TIME>_<MISSION_ID>_<OCC_ID>_<MODE>_<SWVER>_<PVER>.<EXT>

Explanation

• <TYPE> is one of “atm”, “bfr”, “bgr”, “dis”, “occ” or “wet” (see Table below). • <DATE>_<TIME> is the start date and time of the observation as YYYYMMDD_HHMMSS. • <MISSION_ID> is an ID that uniquely identifies the observing satellite. • <OCC_ID> is an ID that identifies the GNSS satellite used for the specific occultation. • <MODE> is the processing mode and one of “O”, “R”, “I”, “T”, “V” (see below). • <SWVER> is a four-digit code denoting the system software version (e.g. “0372”). • <PVER> is a four-digit code denoting the data product version (e.g. “0010” for 1.0). • <EXT> is the extension,

o “nc” for NetCDF. o “bin” for BUFR for disseminated products.

Example An example of an offline atm file name from 29 January 2011:

atm_20110129_230533_C004_G010_O_0372_0010.nc



43 of 50

Overview tables: The tables below show the file types for Level 1 and 2, and the possible processing modes. Table 5.1 Description of ROM SAF files types, all file types hold a “header” – a set of ancillary data

useful for processing or describing each occultation. For detailed information see [RD.11].

Type ROPP

levels Format Description

dis

1A 1B 2A 2B 2C 2D

NetCDF

This file type is the input to the BUFR file that is disseminated on GTS and EUMETCast. The content is bending angle and impact parameter originating from CGS and the refractivity profile and 1DVar output derived from this. This file holds LEO and GNSS positions and velocities from the GSN NRT product.

bfr - BUFR This file type is based on the “dis” file. The BUFR file holds a thinned set of the bending angles and refractivity from the “dis” file. Only the first position and velocity sample is contained in this file.

bgf, bgn, bgo, bga

2D NetCDF These file types contain the model background fields used for the 1DVar retrieval.

occ 1A NetCDF This file contains signal-to-noise for the phases, the excess phases and the GNSS/LEO positions and velocities as function of time.

atm 1B 2A NetCDF This file contains latitude, longitudes, impact parameters, bending

angles and refractivity.

wet 2B 2C NetCDF This file contains output from the 1DVar i.e. temperature, pressure

and humidity. zgrid N/A NetCDF Gridded monthly mean of Level 1 and Level 2 data. trace N/A NetCDF Meta-data associated with the zgrid file.

Table 5.2 ROM SAF processing mode acronyms

Mode Description

N NRT processing O Offline processing R Climate Data Record (reprocessed) I Interim Climate Data Record T Test V Validation



44 of 50

6. Data quality and limitations The product quality is assessed in validation reports [RD.18], [RD.19], [RD.20], [RD.21], [RD.22], [RD.25] and [RD.26]. The detailed discussions of the quality of NRT products are given in the validation reports [RD.18], [RD.19], [RD.25], [RD.26], [RD.27], [RD.28], [RD.29], [RD.30]. The detailed discussions of the quality of reprocessed products are given in the validation reports [RD.20], [RD.21] and [RD.22]. For practical reasons the NRT products are solely validated against operational ECMWF analyses, while the reprocessed data is validated against multiple external data sources. Generally for the Level 2B data the major limitation to be considered is effect of inhomogeneity in the background data from ECMWF operational, ERA-I and ERA5 forecasts. The validation report [RD.19] describes to which extent, and in which areas and altitude ranges, the background data have impact on the products.



45 of 50

7. Dissemination of data products The ROM SAF data products are disseminated through different channels:

• GTS/RMDCN network; • EUMETCast; • Web page (HTTP);

The products that are distributed by each channel are identified in Section 2. For access to these data, and other archived data, it is necessary to sign up as a registered user with the ROM SAF. This is done at the web page of the ROM SAF: http://www.romsaf.org. The near real-time distribution of ROM SAF products to the National Meteorological Services (NMSs) of EUMETSAT Member States and Co-operating States is through the GTS/Regional Meteorological Data Communication Network (RMDCN). This requires the ROM SAF products to be compliant with the World Meteorological Organisation standard binary format, the BUFR format. See [RD.5] for the BUFR format descriptions. The NRT data are disseminated via EUMETCast as well. These data are formatted in the netCDF format, see [RD.8]. The dissemination is done by uploading data to the EUMETCast dissemination FTP server. All NRT (BUFR and netCDF) files are also available from the archive for non-real time purposes. The distribution of offline data, CDRs, and ICDRs is done via HTTP. HTTP data delivery is serviced over the internet, not over specialized, guaranteed performance operational lines. Unlike NRT products, which are actively broadcast, these products are passively made available at the archive for retrieval by the user.

Figure 7-1 Overview of file formats and dissemination types

GTS/RMDCN

EUMETCast

HTTP

End

Users

BUFR files

BUFR and netCDF files

BUFR and netCDF files

NRT products

Offline and archived products



46 of 50

8. Summary This PUM describes broadly the ROM SAF L1B and L2 bending angle, refractivity, dry temperature, temperature, specific humidity, pressure, surface pressure and tropopause height products. The description is general, covering NRT, offline, CDR, and ICDR data products at the same time. For more details about quality differences, performance difference and specificities the user should refer to the Algorithm Theoretical Baseline Documents [RD.6, RD.17 and RD.17] and the Validation Reports for the different products [RD.18, RD.19, RD.20, RD.21, RD.22, RD25 and RD.26].



47 of 50

Annex A NetCDF header format From the ROPP User Guide [RD.2, version 9.0]:



48 of 50

Product Confidence Data definition (the %PCD variable in the “Quality”-section above). PCD_summary is a summary bit which is set if any of bits 4, 5, 6, 7 or 14 is set



49 of 50

Annex B Level 1 data NetCDF formats From the ROPP User Guide [RD.2, version 9.0]:



50 of 50

Annex C Level 2 data NetCDF formats From the ROPP User guide [RD.2, version 9.0]: