A method for reconstruction of gridded fields of sea surface variables from timedependent observations, using sub-regional EOF (Empirical Orthogonal Functions) patterns from models, is presented and tested. Covariance fields, calculated from the model results over long enough time span, are used to find EOF modes. The gravest "observational" amplitudes and their first temporal derivatives are determined from the least-square minimization of fitting errors in relation to the observed values. The field is reconstructed by superposition of continuous model-based mode patterns multiplied by observational amplitudes that meet adopted statistical limits. If the observational amplitude exceeds the limits, gridded fields for this and higher modes are not produced. We applied the method in the northeastern Baltic over the model time series 2010-2015. Daily averages of sea surface temperature (SST) and salinity (SSS) from the highresolution (grid step 0.5 nautical miles) sub-regional HBM model were spatially averaged over bins of 5 × 5 nautical miles. Three first modes cover 99% of variance of temperature and 61.4% of salinity. As shown by experiments with pseudo-observations (model values at these points reconstructed to the model grid and then compared with the original model data), reconstruction performance depends on the configuration of the observation points in the model domain. Still, a few first modes usually produce acceptable results. When removing the SST seasonal cycle prior to EOF analysis, spatial patterns of leading modes remained practically unchanged, share of variance of the three first modes was reduced to 88.6% and reconstruction errors were reduced by about 25%. Sufficient spatial data coverage of the larger basin with ship-born observations usually takes quite long time-of the order of month; therefore, time correction of the amplitudes using the found temporal derivatives improves the accuracy of reconstruction. The method is compared with the Optimal Interpolation (OI) by using the pseudo-observations. Results show that, for SST reconstruction, the OI method is significantly worse than the EOF method. For SSS, OI is slightly better than EOF. The superiority of EOF is that the remote correlation patterns can be used in the reconstruction, which is important when the observations are sparse.
We present an estimate of the main parameters of the wave climate at the eastern coast of the Baltic Proper based on recently digitized data of historical visual wave observations from Ventspils, Latvia (57º24'N, 21º32'E) for 1954-2011. The features of the local wave climate (the average wave properties, frequency of occurrence of waves of different heights and periods, joint distribution of wave heights and periods) match the existing knowledge about the wave properties in this area. The long-term average wave height in the nearshore is about 0.6 m and typical wave periods are 2-5 s. The strongest signal in the data is the seasonal course in wave heights. Relatively strong decadal changes and especially interannual variability in the average wave height largely mask long-term trends. The annual mean wave height decreased considerably in the 1950s and the 1960s and gradually increased from 1970 onwards. The formal trends are statistically insignificant. Long-term and decadal variations in the mean significant wave height are compared with similar data from other wave observation sites in the north-eastern and south-eastern parts of the Baltic Sea (Vilsandi and Pakri in Estonia, Nida in Lithuania).
Abstract. The tested data assimilation (DA) method based on EOF
(Empirical Orthogonal Functions) reconstruction of observations decreased
centred root-mean-square difference (RMSD) of surface temperature (SST) and
salinity (SSS) in reference to observations in the NE Baltic Sea by 22 %
and 34 %, respectively, compared to the control run without DA. The method
is based on the covariance estimates from long-term model data. The
amplitudes of the pre-calculated dominating EOF modes are estimated from
point observations using least-squares optimization; the method builds the
variables on a regular grid. The study used a large number of in situ FerryBox
observations along four ship tracks from 1 May to 31 December 2015, and
observations from research vessels. Within DA, observations were
reconstructed as daily SST and SSS maps on the coarse grid with a resolution
of 5 × 10 arcmin by N and E (ca. 5 nautical miles) and
subsequently were interpolated to the fine grid of the prognostic model with
a resolution of 0.5 × 1 arcmin by N and E (ca. 0.5 nautical
miles). The fine-grid observational fields were used in the DA relaxation
scheme with daily interval. DA with EOF reconstruction technique was found
to be feasible for further implementation studies, since (1) the method that works
on the large-scale patterns (mesoscale features are neglected by taking only
the leading EOF modes) improves the high-resolution model performance by a
comparable or even better degree than in the other published studies, and (2) the
method is computationally effective.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.