Abstract: The PLS-residual bilinearization (PLS-RBL) approach to background correction presented in part 1 of this work is demonstrated with an example from HPLC with diode array detection. Data are also evaluated with generalized RAFA and results are compared. three-way matrix

Abstract: A PLS calibration model is used for quantitative analysis ot bilinear data. In order to reduce the number of independent variables (X block) of the PLS model, the first few principal components for each chromatogram were calculated and the resulting scores were used as X variables. The actual number of principal components was determined by cross-validation. Application of the method is demonstrated by using both simulated and real data, the latter an analysis of phenanthrene and anthracene mixtures by HPLC with an UV-DAD. Results show good predictive properties and stability compared with normal PLS

Abstract: Changes of mean annual net accumulation at the surface on the grounded ice sheets of East Antarctica, West Antarctica and Greenland in response to variations in sea ice extent are estimated using grid-point values 100 km apart. The data bases are assembled principally by bilinear interpolation of remotely sensed brightness temperature (Nimbus-5 ESMR, Nimbus-7 SMMR), surface temperature (Nimbus-7 THIR), and surface elevation (ERS-1 radar altimeter). These data, complemented by field data where remotely sensed data are not available, are used in multivariate analyses in which mean annual accumulation (derived from firn emissivity) is the dependent variable; the independent variables are latitude, surface elevation, mean annual surface temperature, and mean annual distance to open ocean (as a source of energy and moisture). The last is the shortest distance measured between a grid point and the mean annual position of the 10% sea ice concentration boundary, and is used as an index of changes in sea ice extent as well as of mean concentration. Stepwise correlation analyses indicate that variations in sea ice extent of +/-50 km would lead to changes in accumulation inversely of +/-4% on East Antarctica, +/-10% on West Antarctica, and +/-4% on Greenland. These results are compared with those obtained in a previous study using visually interpolated values from contoured compilations of field data; they substantiate the findings for the Antarctic ice sheets (+/-4% on East Antarctica, +/-9% in West Antarctica), and suggest a reduction by one half of the probable change of accumulation on Greenland (from +/-8%). The results also suggest a reduction of the combined contribution to sea level variability to +/-0.19 mm a(-1) (from +/-0.22 mm/a)