The objective of this work package is the development of a handheld optical system for measurement of the excitation–emission matrices (EEM) of aqueous samples under field-conditions. At the Center for Integrated Sensors Systems, we have developed special technologies for high sensitive fluorescence detection in blood and aqueous samples. The used sensor principle is based on confocal fluorescence spectroscopy which should be integrated into a handheld device. Excitation light generated by a broadband or an array of light sources is spectral filtered and directed by a dichroic mirror to the sample. The generated fluorescence light passes the dichroic mirror and several optical filter stages before the light is amplified by a high sensitive photo multiplier (PMT). To avoid an influence of surrounding stray light on the detector signal, the light source is pulsed and the photomultiplier output signal is phase synchronous amplified by a lock-in amplifier. Within the research project the fluorescence detection technology should be made capable for operation in the UV regime and highly integrated into an easy to use handheld device for lab- and field testing of water samples.
For DOM analyses using real-time sensors, the peak picking method has been shown to deliver data quality equivalent to the PARAFAC method. Besides, DOM analysis can be established using neural networks. This signal processing principle is well suited to solve complex (large-scale) problems which are currently intractable. It is planned to implement neural network algorithms using the “Neural Network” Toolbox from MATLAB for an optimized and reduced complexity of DOM analysis. For data validation and quality control, real-time measurements will be compared with measurements performed on water samples collected directly from the stream or the Lysimeters in the project.