TOPAZ the Arctic TEP and the Arctic GOOS

TOPAZ the Arctic TEP and the Arctic GOOS L. Bertino, G. Evensen, K. A. Lisæter, I. Keghouche Arctic GOOS opening, Bergen, 12 th Sept. 2006

Analysis 1 st Sept. 2006 Google Earth

TOPAZ System Description Model components Assimilation method Upgrades

Motivation § Objective: § Provide short-term forecasts of physical and biogeochemical parameters targeted to users needs (primarily the offshore oil and gas industry) § Strategy § Focus on advanced data assimilation techniques § Gradual increase of resolution § Nesting on regions of higher interest § Support § TOPAZ is the Arctic component of MERSEA IP § ESA, Industry, private donation (Frank Mohn AS, Bergen)

The TOPAZ model system § TOPAZ: Atlantic and Arctic § 18 -35 km resolution § 22 vertical layers § Assimilates § SLA (4 satellite altimeters), § SST (from AVHRR), § ice concentrations (SSM/I) § Run weekly, ECMWF forcing

Ice concentration data § Passive MW (NSIDC) § Real-time data set (2 -3 days delay) § NORSEX algorithm (Svendsen et al 1983) § 37 GHz, 19 GHz § ~25 km resolution

Ice extent and ice volume Solid black line - ensemble mean Dashed black line - free run Grey - individual ensemble members

Barents Sea

TOPAZ 2 (400 x 600 x 22) 20 km resolution TOPAZ 3 (800 x 880 x 22) 11 km resolution

TOPAZ Upgrade Stronger W. Spitzberg Current TOPAZ 2 TOPAZ 3

CERSAT - IFREMER

The Arctic TEP Thematic Portal Common Viewing Downloading Downscaling Validation Plan

Live Access Server

The MERSEA metrics

The MERSEA metrics § Class 1: 3 D daily averages § Class 2: Sections and moorings § Class 3: derived quantities (fluxes) § Class 4: validation to observations

Top 3 m layer

100 m depths

Validation procedures against in-situ measurements and climatology Station at the North Pole TOPAZ profiles In-situ data from CORIOLIS (Argo, XBT, …)

Status / Plan § TOPAZ: § next upgrade (TOPAZ 3) Apr. 2007 § Ice drift data assimilation § Arctic TEP § Started during the TOP 1 period (Oct. 2005) § Barents Sea model § Downscaling from TOPAZ. § Real-time since Sept. 2006. § MERSEA TOP 2 period (Apr. – Sept. 07) § Contributions from all Mersea V 2 systems § More validation metrics

Thank You

The ingredients § Models § HYCOM (U. Miami, USA) § Ice model § Biogeoch. model (AWI, D) § Observations § § Altimetry, SST (CLS, F) Sea Ice (NSIDC, USA) Sea Ice drift (Cersat, F) In-situ (CORIOLIS, F) § Data assimilation § Ensemble Kalman Filter [Evensen 1994, 2006]

Ensemble Kalman filtering a stochastic process Forecast 1 Analysis Member 1 Member 2 2 …… Member 99 1. Initial uncertainty 2. Model uncertainty 3. Measurement uncertainty 3 Observations Member 100

Assimilation update Summer Surface temperature update Ice concentration update Surface salinity update

Results from TOPAZ and nested models

Validation Salinity

Validation Temperature

Results from a nested model VOLUME TRANSPORTS BIC AW NCC How good are the boundary conditions ?

Norway-Bear Island Net transport AW: 1. 5 Sv/year (1. 7 Sv in winter and 1. 3 Sv in Summer) [Ingvaldsen et al. , 2004] Net transport NCC: 0. 5 Sv/year [Blindheim, 1989] Net transport BIC: -? ? ?

Bear Island Svalbard Recirculation within the Bear Island Trough is relatively stable at ~ 1. 0 Sv [Ingvaldsen et al. , 2002]

Svalbard – Franz Josef Land In => 0. 4 Sv Out => 0. 1 Sv [Loeng et al. , 1997] (used Russian literature)

Frans Josef Land – Novaya Zemlja In = 0 to 0. 3 Sv in Summer Out = 1. 5 Sv (between 0. 6 Sv in Summer and 2. 6 Sv in Winter) [Schauer et al. , 2002]

Kara Gate In = 0. 1 Sv Out = 0. 05 to 0. 7 Sv [Loeng et al. , 1997] (used Russian literature)

Planned upgrade TOPAZ 3

Expected improvements § Improved dynamics § Better resolution of shelf currents (esp. in the Nordic Seas) § Better fluxes in/out of the Arctic Ocean § More efficient assimilation of altimetry and hydrographic profiles. TOPAZ 2 Spitzberg Barents Sea opening TOPAZ 3 Norway

Future Perspectives § Geographical extension § Indian ocean (under development) § Pacific Ocean / South China Sea (Nansen-Zhu, Beijing) § Exhaustive model validation § In collaboration with the MERSEA gang § Real-time assimilation of more observations § § Temperature and salinity profiles (Argo program) Ice thicknesses (CRYOSAT) Sea surface salinity (SMOS / Aqua) New geoid and Mean Dynamic Topography (GOCE) § Progresses in data assimilation § Biases, non-linearity, parameter estimation § Coupling to a global system § Mercator

The end

System Applications Nested systems in 1. North Sea (N. Winther/C. Hansen) 2. Gulf of Mexico (F. Counillon) 3. Barents Sea (I. Keghouche)

Illustr. I. Kerghouche