Correction of Real-Time Sensor Data in an Urban Stream Environment Reveals Diel Water Quality Patterns

Real-time optical sensors provide high-resolution measurements of water quality in urban streams, enabling detection of contamination events and rapid response to changing conditions. However, in situ sensors are susceptible to environmental and instrument-related interferences, including temperature effects, biofouling, and signal drift, which can obscure true environmental patterns. Effective correction methods are therefore needed to distinguish authentic water quality fluctuations from background variability. In this study, we present a 9-month analysis of humic-like (CDOM) and tryptophan-like fluorescence (TRP) data collected from a real-time fluorometer deployed in Alvarado Creek, an urban stream in San Diego, CA. We developed a multistep postprocessing workflow based on signal analysis techniques to correct for temperature variance, correct biofouling, and remove sensor artifacts. This resulted in a time series that revealed consistent diel patterns in TRP and CDOM concentrations, with morning peaks that closely aligned with fluctuations in streamwater level, suggesting a strong influence of evapotranspiration and water level on organic matter transport during dry weather baseflow. Our data correction workflow and subsequent analysis of corrected data highlight the potential for high-frequency optical sensors to capture recurring, diel dynamics in urban stream systems and differentiate the true environmental signal from sensor interferences.