Аннотация:Fluorescence excitation emission matrices (EEMs) have a trilinear structure, aligning perfectly with the tensor rank decomposition, PARAFAC. Consequently,PARAFAC has become essential for extracting information from freshwater EEMs,pinpointing individual fluorophore groups, and tracking their behaviour acrossdiverse environment. However, EEMs of seawater, with typically low organicmatter, are often dominated by Rayleigh and Raman scattering, which deviatesfrom the trilinear model. Traditional one-dimensional interpolation to eliminatethese interferences varies in outcome based on its matrix application directionand struggles with noisy data. Our proposed techniques, employing Whittakersmoothing and CNN, effectively eliminate scattering signals, even in noise-richscenarios. Notably, CNN adeptly preserves the overall EEM shape across varioussizes and dimensions, establishing itself as an optimal choice for interpolatingscattering zones in EEMs of organic matter-deficient freshwaters.
