No description available.

No description available.

Models for the mean sea surface (MSS) are created by combining and interpolating on a specified spatial grid inhomogenous data sets from different satellites with different ground track coverage. There are various approaches in which the sea surface height (SSH) data from different satellites can be combined to create an accurate reference surface. The orbit errors (especially from the early missions) need to be reduced, and systematic biases between different satellites can be decreased by re-processing them using the improved models and geophysical corrections. In this research, a new method for the data adjustment (or error reduction), which attempts to compensate for both long-wavelength orbit errors and systematic biases, simultaneously and efficiently. The approach is based on using an accurate sea surface profile as a reference surface for the integration process. The new data adjustment technique is based on along-track SSH gradients computed for each satellite, which are integrated along-track with initial values obtained by dual crossover computation with respect to an accurate set of sea surface heights. The accurate Jason-1 SSH data were used to determine the reference surface, and a total of 5 different satellites (Geosat ERM, ERS-2, T/P, Envisat and ERS-1 geodetic mission) data were adjusted to the Jason-1 SSH data. After editing, the new homogeneous SSH datasets were averaged into mean SSH profiles. Then, they were gridded into a 5-minute resolution mean sea surface over the global ocean within ±60º latitudes, as defined by the Jason-1 mean profile, using a 2-D spline interpolation in tension with Green’s function approach. The new gridded mean sea surface, named CSRMSS14 was validated by three comparisons. First, it was compared with two accurate altimeter data sets: 7-year Jason-1 and 8-year Envisat mean profiles. Second, two recent MSS models, DNSC08 and DTU10, were compared to investigate the accuracy of CSRMSS14. Third, a somewhat independent test is obtained by comparing a 2-year Jason-2 mean profile with the three MSS models (CSRMSS14, DTU10 and DNSC08), since Jason-2 data were not used in their construction. These three validations demonstrated that CSRMSS14 mean sea surface model obtained with this new approach is comparable in accuracy to DNSC08 and DTU10. / text

No description available.

No description available.

No description available.

No description available.

No description available.

The statistical prediction of local sea surface winds at a number of locations over the global ocean (Northeast Pacific, Northwest Atlantic and Pacific, tropical Pacific and Atlantic) is investigated using a surface wind statistical downscaling model based on multiple linear regression. The predictands (mean and standard deviation of both vector wind components and wind speed) calculated from ocean buoy observations on daily, weekly and monthly temporal scales are regressed on upper level predictor fields (derived from zonal wind, meridional wind, wind speed, and air temperature) from reanalysis products. The predictor fields are subject to a combined Empirical Orthogonal Function (EOF) analysis before entering the regression model. It is found that in general the mean vector wind components are more predictable than mean wind speed in the North Pacific and Atlantic, while in the tropical Pacific and Atlantic the difference in predictive skill between mean vector wind components and wind speed is not substantial. The predictability of wind speed relative to vector wind components is interpreted by an idealized Gaussian model of wind speed probability density function, which indicates that the wind speed is more sensitive to the standard deviations (which generally are not well predicted) than to the means of vector wind component in the midlatitude region and vice versa in the tropical region. This sensitivity of wind speed statistics to those of vector wind components can be characterized by a simple scalar quantity theta=arctan(mu/sigma) (in which mu is the magnitude of average vector wind and sigma is the isotropic standard deviation of the vector winds). The quantity theta is found to be dependent on season, geographic location and averaging timescale of wind statistics. While the idealized probability model does a good job of characterizing month-to-month variations in the mean wind speed based on those of the vector wind statistics, month-to-month variations in the standard deviation of speed are not well modelled. A series of Monte Carlo experiments demonstrates that the inconsistency in the characterization of wind speed standard deviation is the result of differences of sampling variability between the vector wind and wind speed statistics. / Graduate

This thesis describes the construction of a sea surface temperature (SST) dataset from Along-Track Scanning Radiometer (ATSR) observations suitable for climate applications. The algorithms presented here are now used at ESA for reprocessing of historical ATSR data and will be the basis of the retrieval used on the forthcoming SLSTR instrument on ESA’s Sentinel-3 satellite. In order to ensure independence of ATSR SSTs from in situ measurements, the retrieval uses physics-based methods through the use of radiative transfer (RT) simulations. The RT simulations are based on the Reference ForwardModel line-by-line model linked to a new sea surface emissivity model which accounts for surface temperature, wind speed, viewing angle and salinity, and to a discrete ordinates scattering (DISORT) model to account for aerosol. An atmospheric profile dataset, based on full resolution ERA-40 numerical weather prediction (NWP) data, is defined and used as input to the RTmodel. Five atmospheric trace gases (N2O, CH4, HNO3, and CFC-11 and CFC-12) are identified as having temporal and geographical variability which have a significant (∼0.1K) impact on RT simulations. Several additional trace gases neglected in previous studies are included using fixed profiles contributing ∼0.04K to RT simulations. Comparison against ATSR-2 and AATSR observations indicates that RT model biases are reduced from 0.2–0.5K for previous studies to ∼0.1K. A new coefficient-based SST retrieval scheme is developed from the RT simulations. Coefficients are banded by total column water vapour (TCWV) from NWP analyses reducing simulated regional biases to <0.1K compared to ∼0.2K for global coefficients. An improved treatment of the instrument viewing geometry decreases simulated view-angle related biases from ∼0.1K to <0.005K for the day-time dual-view retrieval. To eliminate inter-algorithmbiases due to remaining RT model biases and uncertainty in the characterisation of the ATSR instruments the offset coefficient for each TCWV band is adjusted to match the results from a reference channel combination. As infrared radiometers are sensitive to the skin SST while in situ buoys measure SST at some depth below the surface an adjustment for the skin effect and diurnal stratification is included. The samemodel allows adjustment for the differing time of observation between ATSR-2 and AATSR to prevent the diurnal cycle being aliased into the final record. The RT simulations are harmonised between sensors using a double-difference technique eliminating discontinuities in the final SST record. Comparison against in situ drifting and tropical moored buoys shows the new SST dataset is of high quality. Systematic differences between ATSR retrieved SST and in situ drifters show zonal, regional, TCWV, and wind speed biases are less than 0.1K except for themost extreme cases (TCWV <5 kgm−2). The precision of ATSR retrieved SSTs is ∼0.15 K, lower than the precision ofmeasurement of the global ensemble of in situ drifting buoys. From 1995 onwards the ARC SSTs are stable with instability of less than 5mK year−1 to 95% confidence (demonstrated for tropical regions).