GRAS Level 1 Processing and Products JuhaPekka Luntama

GRAS Level 1 Processing and Products Juha-Pekka Luntama and Julian Wilson EUMETSAT Am Kavalleriesand 31, D-64295 Darmstadt, Germany GRAS SAF User Workshop 11 - 13 June 2003 1

Itinerary 1) Introduction 2) GRAS measurement system 3) GRAS level 1 data processing 4) GRAS level 1 products 5) Conclusions GRAS SAF User Workshop 11 - 13 June 2003 2

Introduction • EUMETSAT Polar System (EPS) is the European component of the joint European/US polar satellite system (IJPS) • EPS mission consists of three Metop satellites with a payload of eight meteorological instruments • The first Metop satellite is launched in second half of 2005 • The total duration of the EPS mission is 14 years => NRT products for NWP applications and continuity of radio occultation data for climate monitoring • GRAS receiver has been developed especially for radio occultation measurements within the EPS framework by ESA and EUMETSAT GRAS SAF User Workshop 11 - 13 June 2003 3

Metop spacecraft GRAS GOME AVHRR HIRS IASI AMSU-A 1 ASCAT MHS GRAS SAF User Workshop 11 - 13 June 2003 AMSU-A 2 4

NRT Challenge • EPS GRAS is the first RO mission that has been from the beginning designed for operational applications • NRT requirements for GRAS product dissemination are 2 h 15 min for level 1 b and 3 h for level 2 from the observation • Analysis of the NRT feasibility has revealed two main risks: – Timely availability of the auxiliary data – NRT Precise Orbit Determination of the LEO satellite • Mitigation of these risks is incorporated in the GRAS Measurement System design • There has been no reason to compromise the occultation data processing due to the NRT requirements GRAS SAF User Workshop 11 - 13 June 2003 5

GRAS NRT Product requirements GRAS SAF User Workshop 11 - 13 June 2003 6

GRAS measurement system GRAS SAF User Workshop 11 - 13 June 2003 • GRAS GSN provides GPS POD products and ground based measurements • Level 1 b products are processed and disseminated by the EPS CGS • Level 2 products are processed and disseminated by the GRAS Meteorology SAF • Archived products include raw data, level 1 b, level 2 and all GSN products • Off-line user access to archived products via EUMETSAT UMARF 7

GRAS receiver • • • GRAS SAF User Workshop 11 - 13 June 2003 12 bi-frequency channels Codeless-mode capability Oven stabilised USO (Allan deviation 10 -12) Directional high gain occultation antennas for minimising local multipath Onboard DSP => autonomous operations 8

GRAS GSN Service • GRAS GSN is a service to provide GPS state vectors, clock offset estimates, and clock differencing data for GRAS data processing • GSN will contain a network of 25 fiducial stations • GSN coverage for the GPS constellation is > 200 % to ensure service availability via redundancy • The GSN Service is designed to support the GRAS NRT data processing requirements GRAS SAF User Workshop 11 - 13 June 2003 9

GRAS GSN products • • • GPS NRT state vectors (15 min sampling) GPS NRT clock offset estimates (1 s sampling) Fiducial station clock offset estimates (1 s sampling) Troposphere Zenith Delay (TZD) Sounding Support Data (SSD) for clock correction: – L 1 and L 2 carrier phase – L 1 and L 2 pseudorange – L 1 and L 2 SNR • GSN Configuration and Status database • EOP GRAS SAF User Workshop 11 - 13 June 2003 10

GRAS level 1 processing GRAS SAF User Workshop 11 - 13 June 2003 11

GRAS level 1 a processing GRAS SAF User Workshop 11 - 13 June 2003 12

GRAS level 1 NRT data products GRAS SAF User Workshop 11 - 13 June 2003 13

GRAS level 1 NRT data products (cont. ) GRAS SAF User Workshop 11 - 13 June 2003 14

Level 1 product accuracy • Combined bending angle error at 30 km for azimuth 30 : – GRAS dependent errors: – GRAS independent errors: – RSS 0. 420 rad 0. 662 rad 0. 784 rad • When azimuth angle is > 45° the GRAS dependent errors increase to 0. 651 rad => Total RSS for 90 % of the measurements < 0. 935 rad. • Analysis by the instrument manufacturer indicates that statistically about 8 % of the observations are outside the accuracy specification => 40 occultations out of the daily 500 have errors larger than 1 rad GRAS SAF User Workshop 11 - 13 June 2003 15

GRAS level 1 processing summary • GRAS does not directly produce phase, amplitude, SNR, or pseudorange measurements => they have to be derived from the raw data using a measurement reassembly function • GRAS level 1 processing will include GO, WO, and raw sampling mode data processing chains • Partially overlapping bending angle profiles from all processing chains are disseminated to the users • A software tool for combining and smoothing the bending angle profiles for NWP is planned by the GRAS SAF • Metop NRT POD problem has been solved by using a SRIF algorithm together with the NAPEOS S/W and by solving the Metop orbit in short (< 10 min) arcs GRAS SAF User Workshop 11 - 13 June 2003 16

Conclusions • The objective of the EPS GRAS mission is to provide NWP users operationally radio occultation data products • The EPS mission starts in 2005 and provides data continuously at least for 14 years • The stringent NRT timeliness requirements have made the development of the GRAS measurement system and the data processing chain challenging • GRAS data processing and dissemination are performed by the EPS CGS (level 1 b ) and by the GRAS Meteorology SAF (level 2 ) • All archived GRAS products (level 0, level 1 b, and level 2) shall be made available via the EUMETSAT UMARF facility • Analysis and study results indicate that the NRT data processing with the required accuracy is feasible for about 92 % of the observations GRAS SAF User Workshop 11 - 13 June 2003 17