Coastal Observation and a Coastal Observation and a Forecasting - PowerPoint PPT Presentation

Coastal Observation and a Coastal Observation and a Forecasting System for the Forecasting System for the German Bight German Bight Emil V. Stanev and Franciscus Colijn emil.stanev@gkss.de Taipei, 23-25 April 2008

North Sea German Bight Wadden Sea How the coastal ocean will be affected by changes in the tidal currents and the sediment supply due to climatic change

Monitoring and forecasting these changes can Monitoring and forecasting these changes can only be addressed by a combination of state combination of state- - only be addressed by a of- -the the- -art observations and modelling art observations and modelling of activities. activities. Objective measures of uncertainties in the uncertainties in the Objective measures of state estimates and forecasts are needed. are needed. state estimates and forecasts I mproving the performance of observing and I mproving the performance of observing and forecasting systems requires a clear forecasting systems requires a clear identification of relevant parameters and and identification of relevant parameters representation of multiple physical and multiple physical and representation of biogeochemical processes in numerical in numerical biogeochemical processes models. models.

Outline Outline • Observations Observations • • Numerical Modelling Numerical Modelling • • Model Validations Model Validations • Aims of this Talk Aims of this Talk • States Estimates States Estimates • Address regional Address regional observations and predictions observations and predictions • Conclusions Conclusions • with a focus on the German with a focus on the German Bight and Wadden Sea Bight and Wadden Sea I dentify further perspectives I dentify further perspectives of (pre)operational of (pre)operational oceanography oceanography

I ntegrated Coastal I ntegrated Coastal Observation Observation Network (I CON) Network (I CON)

Elbe Buoy Data Germ an Bight NSboje I I I NSboje I I FI NO

WERA Data

Other Data Other Data Observations Real–time data DWD product validations

RG- - W ATT Data Station: Data W ATT Data Station: Data RG are Already Extensively Used to are Already Extensively Used to Calibrate the W ATT- - Model Model Calibrate the W ATT Senkenberg ADCP data Uni-Oldenburg m eteo- and hydro physical data

ADCP Transects ADCP Transects

Use New Data: Use New Data: MERI S TSM - - 3 0 0 m Data and 3 0 0 m Data and MERI S TSM num erical m odel sim ulations num erical m odel sim ulations 1 9 Feb 2 0 0 5 9 Sep 2 0 0 4 1 9 Feb 2 0 0 5 9 Sep 2 0 0 4 Gemein et al. (2006)

Ships of Opportunity Ships of Opportunity (FerryBoxes) (FerryBoxes) Data - - Data, but also Data Data, but also Data - - Model Model Data Calibration is Needed Calibration is Needed

Numerical Modelling Can Help a Lot in the Numerical Modelling Can Help a Lot in the Regional Studies. Regional Studies. One- - w ay Nested w ay Nested Modelling Modelling One System System 6 nm 3 nm GETM: flooding and drying 1 km 200 m East Frisian W adden Sea

... Coupled with with ... Coupled Sediment Transport Model Model Sediment Transport • 3 D equations • 3 D equations • Tw o classes • Tw o classes of of sedim ent sedim ent ( ( m ud m ud and and sand sand) ) • Com plex physical param eterizations • Com plex physical param eterizations • Calibrated • Calibrated to to present present day observations day observations • Hydrodynam ic m odel • Hydrodynam ic m odel forced by w aves forced by w aves and tides and tides Areas of Application: Areas of Application: • North Sea (in cooperation with BSH) • North Sea (in cooperation with BSH) • German Bight (HAMSOM • German Bight (HAMSOM- -based) based) • East Frisian Wadden Sea • East Frisian Wadden Sea • H • Hö örnum Bay (North Frisian Wadden Sea) rnum Bay (North Frisian Wadden Sea)

Are Models Enough Mature and Are Models Enough Mature and Validated in Order to be Used in Validated in Order to be Used in Pre- - operational Activities? operational Activities? Pre Germ an Bight W adden Sea ( BSH data) ( U_ Ol data)

Validation of the W adden Sea m odel the W adden Sea m odel Validation of GKSS-ADCP data versus sim ulated data Stanev et al. (2003a, OD)

Circulation Circulation

Conceptual Models and Process Studies Can tidal asymmerties Tidal asymmetry due Can tidal asymmerties Tidal asymmetry due cause net net transport transport to superposition to superposition of M2 and of M2 and cause M4 tide tide of sediment? GROEN sediment? GROEN M4 of ( Neth ( Neth. J. S. Res, 1967): YES! . J. S. Res, 1967): YES! I nward transport I nward transport Tidal asymmetry due Tidal asymmetry due to topograpgy control topograpgy control to ζ = π = ω ' sin(2 t / T) sin( t) dV' 1 ( ) = εω ω + ε ω ω 2 cos t sin( 2 t) d t 2 t flood = 2t ebb t flood = 2t ebb Outward transport Outward transport

Asymmetric transport in in the vertical the vertical Asymmetric transport plane (Stanev et al., 2007, CSR) could (Stanev et al., 2007, CSR) could be be plane another candidate to to explain explain the the another candidate landward transport of of sediment sediment. . landward transport Transport of vertically non- -uniform sediment uniform sediment Transport of vertically non concentrations is another level of complication. concentrations is another level of complication. Stanev et al. (2007, JGR): Without shear Stanev et al. (2007, JGR): Without shear diffusion, the sediment dynamics are largely diffusion, the sediment dynamics are largely governed by the ‘‘ ‘‘displacement mechanism displacement mechanism’’ ’’ governed by the Courtesy B. Flemming

Along Channel Change Are Tidal Basins Only Ebb- - Along Channel Change Are Tidal Basins Only Ebb of Correlation Patterns: Dominated? of Correlation Patterns: Dominated? Asymmetries along the channels would contribute to accumulating sediments enabling feedback between circulation, basin hypsometry and sediment transport

The Response to Extreme The Response to Extreme Events (Storm Surge Brita) Events (Storm Surge Brita) is Different From the One is Different From the One to Tides. to Tides.

The I mpact of Wind I mpact of Wind Waves Waves on on the the The Sediment Transport Sediment Transport

EOF Analysis of Forcing: EOF Analysis of Forcing: Bed Shear Stress due to Surface Waves Bed Shear Stress due to Surface Waves and Tides (Currents) and Tides (Currents)

Bed Shear Stress Due to Tides and Wind Bed Shear Stress Due to Tides and Wind Display Different Patterns Also in the Display Different Patterns Also in the Wadden Sea Wadden Sea Tides Tides W aves W aves Stanev et al. ( 2 0 0 6 , OD)

The Response to Climate Change The Response to Climate Change Affects in a Non- -Trivial Trivial Affects in a Non Way Spatial and Way Spatial and Temporal Temporal Variability Variability Patterns. Patterns. Deposition Deposition Minus Minus Erosion Erosion Stanev et al. ( 2 0 0 6 , OD)

Suspended Sedim ent Sedim ent Suspended ( GETM, 2 0 0 m Resolution Resolution) ) ( GETM, 2 0 0 m The im portance of exchange betw een tidal flats and open ocean MERI S/ ENVI SAT, 0 3 / 2 9 / 0 4 , 9 :5 0 , 3 0 0 m , Gem ein et al. ( 2 0 0 6 )

Validation Validation Against Against MERI S MERI S Data Data Num erical sim ulations MERI S I m age Gemein et al. (2006)

Sediment Sorting (Different (Different Sediment Sorting Transport of of Water Water, , Suspended Suspended Transport Matter and and Sand) Sand) Matter Sediment is advected 1 on average more slowly than fluid mass. 2 3

Towards Data Assimilation: Optimal I ntrrpolation of MERI S Data Wave Model OCEAN: Water temperature WAM Sea surface height Salinity Water temperature Salinity significant waves height waves direction waves period RI VERS: Topography Fresh water discharge SPM Concentration Circulation SPM transport model SPM concentration ATMOSPHERE: Seabed Cloudiness sedimentation field Relative humidity + Solar radiation Evaporation Precipitation Air temperature model Air pressure Satellite initial data Wind boundary conditions data assimilation model output SPM 3D-Concentration field Mikhail Dobrynin and Heinz Guenther

WAM WAM

MERI S Data MERI S Data

Simulated seasonal mean surface SPM concentration, 2003, mg/l 15 April-15 October 16 October – 14 April calm storm No assimilation MERIS data assimilation (Optimum Interpolation)

Optimal I nterpolation is not the Optimal I nterpolation is not the Best Solution for Near- -Coastal Coastal Best Solution for Near Regions. Regions. Compressing Data and Compressing Data and Statistical Forecasts Statistical Forecasts (Tidal Oscillations) (Tidal Oscillations)