Charles-Emmanuel Testut

Assimilation of sea-surface temperature and altimetric observations during 1992–1993 into an eddy permitting primitive equation model of the North Atlantic Ocean

Sea-surface temperature (SST) and sea-surface height (SSH) observations collected from space between October 1992 and December 1993 have been assimilated into a realistic primitive equation model of the North Atlantic Ocean circulation at eddy permitting resolution. The assimilated SST data originate from AVHRR observations gathered and processed within the NASA Pathfinder project; the altimetric data consist of SSH maps computed as the sum of a time-invariant dynamic topography and gridded sea-level anomalies obtained by combining Topex/Poseidon and ERS altimeter data. The assimilation scheme is a reduced-rank Kalman filter derived from the Singular Evolutive Extended Kalman (SEEK) methodology [J. Mar. Syst. 16 (1998) 323], in which the error statistics is represented in a subspace of small dimension. The error subspace is initialized with a truncated series of Empirical Orthogonal Functions (EOFs) of the system variability. The analysis algorithm includes a mechanism to update the forecast error statistics adaptively using all pertinent informations from the innovation vector. Hindcast experiments have been conducted with a 1/3° model of the North Atlantic basin forced with ECMWF atmospheric reanalyses. The impact of the data assimilated during 1993 is assessed by examining how observed (SSH and SST) and nonobserved variables (such as velocity and thermohaline properties in the interior of the ocean) are modified by the assimilation scheme. Finally, the validation of the hindcast experiments with independent XBT measurements is performed in order to evaluate the objective skill of the procedure. The various diagnostics demonstrate the positive impact of the satellite data to hindcast the upper ocean circulation at eddy permitting resolution and the capacity of the scheme to estimate the geographic distribution of the forecast error.

The GODAE/Mercator-Ocean global ocean forecasting system: results, applications and prospects

The Mercator-Océan eddy permitting (1/4) global ocean forecasting system assimilating satellite altimetry is the French contribution to the GODAE project and to the MERSEA project for operational systems. It has run operationally since October 2005 and is forced with daily surface fluxes from ECMWF operational analyses and forecasts. JASON, ERS and GFO altimetry measurements from AVISO were assimilated from January 2005 up to real time. The simulation results are compared with independent in-situ data of the Atlantic, Pacific and Antarctic Ocean basins in order to provide an estimation of the performance of the system. The results are also compared with the Levitus climatology, and with combinations of in-situ and satellite observations since 2005. In the Atlantic basin, the global system is also compared with Mercator-Ocean regional systems that assimilate SST (Sea Surface Temperature), SLA (Sea Level Anomalies) and in situ (temperature and salinity profiles) near real time observations

Efficient Parameterization of the Observation Error Covariance Matrix for Square Root or Ensemble Kalman Filters: Application to Ocean Altimetry

Abstract In the Kalman filter standard algorithm, the computational complexity of the observational update is proportional to the cube of the number y of observations (leading behavior for large y). In realistic atmospheric or oceanic applications, involving an increasing quantity of available observations, this often leads to a prohibitive cost and to the necessity of simplifying the problem by aggregating or dropping observations. If the filter error covariance matrices are in square root form, as in square root or ensemble Kalman filters, the standard algorithm can be transformed to be linear in y, providing that the observation error covariance matrix is diagonal. This is a significant drawback of this transformed algorithm and often leads to an assumption of uncorrelated observation errors for the sake of numerical efficiency. In this paper, it is shown that the linearity of the transformed algorithm in y can be preserved for other forms of the observation error covariance matrix. In particular, quit...

Comparative assimilation of Topex/Poseidon and ERS altimeter data and of TAO temperature data in the tropical Pacific ocean during 1994–1998, and the mean sea-surface height issue

Five years of Topex/Poseidon (T/P) and ERS sea level anomaly (SLA) data (1994–1998) are assimilated every 10 days into a primitive equation model of the tropical Pacific ocean. The data assimilation technique used here is a reduced-order Kalman filter derived from the Singular Evolutive Extended Kalman (SEEK) filter [J. Mar. Syst. 16(3–4) (1998) 323] with an error covariance matrix parameterised by a subset of multivariate 3D global empirical orthogonal functions (EOFs). The assimilation run is compared to the free run and to independent data from the TAO network. The impact of sea-surface height (SSH) assimilation on surface and subsurface temperature and currents is estimated in the equatorial band. In a second stage, temperature data from the TAO array are assimilated in the same conditions as in the first stage. The comparison between the results of the two assimilation experiments is made mainly with a view to gaining insights into the mean sea-surface height (MSSH) for the assimilation of altimeter data, and more generally, into the question of biases. XBT observations and TAO array data are then used to build a physically more consistent mean sea-surface height for assimilation of SLA data. Results from the assimilation of altimeter data referenced to this new MSSH show significant improvements.

Evolution of the Arctic Ocean Salinity, 2007–08: Contrast between the Canadian and the Eurasian Basins

The authors investigate the variability of salinity in the Arctic Ocean and in the Nordic and Labrador Seas over recent years to see how the freshwater balance in the Arctic and the exchanges with the North Atlantic havebeenaffectedbytherecentimportantseaicemelting,especiallyduringthe2007seaiceextentminimum. The Global Ocean Reanalysis and Simulations (GLORYS1) global ocean reanalysis based on a global coupled ocean‐sea ice model with an average of 12-km grid resolution in the Arctic Ocean is used in this regard.Althoughno seaicedataandnodataunderseaiceareassimilated, simulationoverthe2001‐09period is shown to represent fairly well the 2007 sea ice event and the different components accounting for the ocean and sea ice freshwater budget, compared to available observations. In the reanalysis, the 2007 sea ice minimum isdueto anincreaseof thesea iceexportthroughFramStrait(25%)andan importantseaice meltin the Arctic (75%). Liquid freshwater is accumulated in the Beaufort gyre after 2002, in agreement with recent observations, and it is shown that this accumulation is due to both the sea ice melt and a spatial redistribution of the freshwater content in the Canadian Basin. In the Eurasian Basin, a very contrasting situation is found with an increase of the salinity. The effect of the sea ice melt is counterbalanced by an increase of the Atlantic inflow and a modification of the circulation north of Fram Strait after 2007. The authors suggest that a strong anomaly of the atmospheric conditions was responsible for this change of the circulation.

Efficient Adaptive Error Parameterizations for Square Root or Ensemble Kalman Filters: Application to the Control of Ocean Mesoscale Signals

Abstract In Kalman filter applications, an adaptive parameterization of the error statistics is often necessary to avoid filter divergence, and prevent error estimates from becoming grossly inconsistent with the real error. With the classic formulation of the Kalman filter observational update, optimal estimates of general adaptive parameters can only be obtained at a numerical cost that is several times larger than the cost of the state observational update. In this paper, it is shown that there exists a few types of important parameters for which optimal estimates can be computed at a negligible numerical cost, as soon as the computation is performed using a transformed algorithm that works in the reduced control space defined by the square root or ensemble representation of the forecast error covariance matrix. The set of parameters that can be efficiently controlled includes scaling factors for the forecast error covariance matrix, scaling factors for the observation error covariance matrix, or even a...

Efficient Local Error Parameterizations for Square Root or Ensemble Kalman Filters: Application to a Basin-Scale Ocean Turbulent Flow

Abstract In large-sized atmospheric or oceanic applications of square root or ensemble Kalman filters, it is often necessary to introduce the prior assumption that long-range correlations are negligible and force them to zero using a local parameterization, supplementing the ensemble or reduced-rank representation of the covariance. One classic algorithm to perform this operation consists of taking the Schur product of the ensemble covariance matrix by a local support correlation matrix. However, with this parameterization, the square root of the forecast error covariance matrix is no more directly available, so that any observational update algorithm requiring this square root must include an additional step to compute local square roots from the Schur product. This computation generates an additional numerical cost or produces high-rank square roots, which may deprive the observational update from its original efficiency. In this paper, it is shown how efficient local square root parameterizations can b...

Characterisation of errors of a regional model of the Bay of Biscay in response to wind uncertainties: a first step toward a data assimilation system suitable for coastal sea domains

This study is aimed at exploring the errors of a regional model of the Bay of Biscay, a regional zoom of the IBI configuration of the ocean model NEMO, with the ultimate objective of guiding the choice and implementation of a data assimilation system in that region. An ensemble experiment was carried out by randomly perturbing winds along a base of EOFs with the aim to mimic a potential source of error in the model forecasts. A characterisation was attempted with proxy forecast errors by using statistical moments of order 1 to 4. The temporal variability of model correction patterns in a hypothetical data assimilation system was also illustrated. Significant departures from linear/Gaussian response were found, as well as well-marked non-stationarities in the error patterns. Within the limits of the experimental protocol, this could be technically applicable to other coastal areas as the study illustrates the likely limits of stationary/Gaussian data assimilation approaches in the Bay of Biscay.

Variability of the Gulf Stream position and transport between 1992 and 1999: a re‐analysis based on a data assimilation experiment

A multivariate dataset resulting from the assimilation of sea surface height and sea surface temperature data in an ocean circulation model of the North Atlantic, is used to study the Gulf Stream position variability from October 1992 to October 1999. The analysis takes place between 75° W and 50° W and focuses on two regions, west and east of 65° W. By calculating zonal averages of the latitudinal position, a regular annual cycle of north–south oscillations of the Gulf Stream axis is identified in the eastern region, whereas inter‐annual variations dominate in the western region. An empirical orthogonal function (EOF) analysis of space–time position anomalies reveals that, even though the annual cycle dominates over the whole domain, the variability has a spatial consistency within each individual region. The assimilation dataset is further used to analyse the variability of the Gulf Stream transport towards the east. Compared to the latitudinal position analysis, the characterization of the transport variability is more complex. Semi‐annual fluctuations can be evidenced in both the regions; however, no significant correlation was found between the fluctuations of position and transport.