SISMER Systmes dInformations Scientifiques pour la Mer FNODC

SISMER Systèmes d’Informations Scientifiques pour la Mer F-NODC Contribution to GODAR-WESTPAC from FRANCE Catherine Maillard, Michèle Fichaut, Loic Petit de la Villeon, Vincent Bernard SISMER, IFREMER Centre de Brest, BP 70, 29280 Brest Cedex, France www. ifremer. fr/sismer International Workshop for GODAR-WESTPAC Tokyo, 5 -7 March 2002 1

Data available at SISMER q 174 Data sets from French cruises have been transmitted to GODAR: Ø Ø 50 CTD cruises = 2850 stations from 1982 -1999 124 bottle cruises = 4037 stations from 1957 - 1996 total 6887 stations q International real time data archived at ARGO/CORIOLIS data centre in 2001 Ø Ø whole Pacific Ocean: 77 431 profiles from GTS West Pacific (60 N -30 S & W 120 E 150): 54591 profiles 2

2850 CTD Stations 3

4037 Bottle Stations in West. PAC 4

77 431 ARGO Profiles (54591 in West. PAC) 5

Quality Assurance QC 0 : Check of format and completeness of information QC 1 : Check of location, Date, Duplicates QC 2 : Check of the data points Results : Elimination or Quality Flag : GTSPP Flag Scale 1 0 No QC 2 Correct Inconsistent 3 Doubtful 4 5 9 Bad Changed Missing References Ø Ø ICES -GETADE Formatting Guidelines for Oceanographic Data Exchange (http: //www. ices. dk/ocean/) ICES -MDM QA Guidelines - CTD - Marine Data Management working Group (CIEM) Reports UNESCO/COI/IODE : Manual of Quality Control Procedures for validation of Oceanographic Data - Manuals and Guides 26 UNESCO/IOC/IODE : GF 3 A General Formattin System for Geo-Referenced Data - Manual and Guides 17 6

Sea Surface Salinity Project Following OOPC and IOC/IODE XVI Conference (IODE-XVI-10) recommendation, an Underway Sea Surface Salinity Data Archiving Project has been initiated with the objective of : q q q q Swift assembly and distribution of data, ready availability and free access. Data sets and products are comprehensive in terms of the variables Encourage exploration of historical and paleo data sets. Encourage processing methods that eliminate or minimize the production of spurious signals. Ensure that adequate and timely data are available for the initialization and validation of climate forecast systems. Maximize the utility of the sustained observing system for various process or intensive studies. Work with GOOS/GCOS, World Weather Watch and Global Atmosphere Watch in development and implementation of the strategy. The JCOMM-1 meeting supported such a project urging that integration with other data collected at the same time be properly considered. 7

SSSAL Preliminary Workplan q Working Ø Ø Ø Data Collection Transfer, Archiving, Processing Products q Second Ø Meeting Announcement September 2002, 16 -17 MEDS Ottawa, Canada q Review Ø Groups of available products and Information http: //www. ifremer. fr/sismer/program/sssal/ 8

QC 1 Control of the location and Date - Data collected in station Automatic Checks Results - Flag Duplicate data sets Elimination or 5 Elimination or 3 Date Ship/Platform velocity Location/shoreline Bottom sounding (ETOPO 5) 5 = Correction/Interpolation 3 = doubtfull depth 9

QC 1 Control of the location and Date VISUAL CHECKS - Data collected in station 10

QC 2 - Control of the Data points Automatic Checks Results - Flag Pressure + one more observation Out of the regional scale (min & max values) Increasing pressure Data below the bottom depth Coherence with pre-existing statistics Elimination 4 (LEVITUS, REYNAUD, MODB, MEDATLAS) No constant profiles Spikes Vertical stability 4 4 2 4 3 4 11

QC 2 - Control of the Data points Visual Checks 12

QC 2 - Control of the Data points Regional Adjustment for Broad Range Checks 13

QC 2 - Control of the Data points Minimum & Maximum - Broad Range Check 0 2000 6000 14

QC 1: Control of the location and Date - Time series on fixed mooring 15

QC 2 - Stabilité Verticale q Seuil de détection de l ’instabilité : Ø 0. 03 en sigma-t pour stations profondes haute résolution Ø à. 05 à 0. 1 pour stations cotières q 4 points incriminés : 2 niveaux, T, S q Recherche d ’ autres anomalies sur chacun de ces 4 points Ø Ø si oui --> flag=4 à ce point si non --> flag=4 aux 4 points et validation manuelle 16

QC 2 - TESTS DE PIC q Test de gradient sur 2 points en début et fin de profil q Test de pic sur 3 points en cours de profil q seuil de gradient ajusté sur chaque zone géographique et niveau vertical 17

QC 2 - Test de la climatologie q Plusieurs climatologies disponibles : Levitus 94, 98, Reynaud, Medatlas, MODB Ø par défaut Levitus 98 1 x 1 degré q Définition de l ’enveloppe de variabilité Station du plateau continental fond < 100 m : 5 écarts-type Station du talus 200< fond < 500 : 4 écarts-type station profonde fond > 500 m : 3 écarts-type 18

Tests Visuels « Subjectifs » q Test de la cohérence interne Ø QC 1 : test de la route Ø QC 2 : coherence interne des valeurs - décalages q Test de la cohérence externe Ø QC 1 : rappel des positions d ’un autre fichier si au moins une position est dans le voisinage au même moment Ø QC 2 : superposition des profils d ’un autre fichier q Levers de doute Ø stabilité verticale Ø pics Ø enveloppe climatologique mal définie 19

FORMATS q Meta-Données : RDBS - ORACLE q Données - Observations autodescriptive ASCII (MEDATLAS/CORIOLIS ASCII) Ø Net. CDF incluent le meta-données sous forme de mots-clé, obligatoire ou optionnels, également copiés dans la BD ORACLE Ø q Données Ø Ø - données maillées MODB/MEDMEX ARGO 20