The deployment of smart metrological procedures and practices as well as challenging operating conditions for urban water monitoring require ever more efficient tools. In this context, the OTHU IField Observatory for Urban Water Management) rehabilitated its field monitoring stations. This paper aims at demonstrating benefits from innovative OTHU's monitoring stations regarding the reliability of collected data and the decrease of maintenance costs. The current technology offers real opportunities in terms of reduced costs, reactivity and improvement of data pre-processing. Each station produces approximately 12,240 data per day, ensuring continuous monitoring of stormwater quantity and quality in urban environment. The long term monitoring of the dynamics of hydrologicat processes requires 24/7 operational and reliable equipment and devices. The new OTHU monitoring stations are reliable and provide representative data of higher quality than the previous ones.

Optimisation or real time control (RTC) studies in wastewater systems increasingly require rapid simulations of sewer systems in extensive catchments. To reduce the simulation time calibrated simplified models are applied, with the performance generally based on the goodness of fit of the calibration. In this research the performance of three simplified and a full hydrodynamic (FH) model for two catchments are compared based on the correct determination of CSO event occurrences and of the total discharged volumes to the surface water. Simplified model M1 consists of a rainfall runoff outflow (RRO) model only. M2 combines the RRO model with a static reservoir model for the sewer behaviour. M3 comprises the RRO model and a dynamic reservoir model. The dynamic reservoir characteristics were derived from FH model simulations. It was found that M2 and M3 are able to describe the sewer behaviour of the catchments, contrary to M1. The preferred model structure depends on the quality of the information (geometrical database and monitoring data) available for the design and calibration of the model. Finally, calibrated simplified models are shown to be preferable to uncalibrated FH models when performing optimisation or RTC studies.