Modelling in urban hydrology is largely based on the analysis of long time series of data. The quality of the results strongly depends on the quality of the data used. Doubtful or wrong data must be detected and eventually substituted by reliable ones when it is feasible before any further exploitation. This paper deals with the development of an automatic pre-validation procedure that detects doubtful and not reliable data, in order to facilitate their interpretation. This procedure consists in applying a set of seven tests based on the following criteria: the functioning state of the sensor, the physical range of the quantity, the locally realistic range, the duration since the last maintenance of the sensor, the signal's gradient, material redundancy and analytical redundancy. The results of the tests are coded with the letter A for reliable values, B for doubtful values and C for wrong values. After this automatic prevalidation, the ultimate validation of values marked B and C is carried out manually by the operator, with the assistance of specifically developed visual and graphical tools.

Assessing the functioning and the performance of urban drainage systems on both rainfall event and yearly time scales is usually based on online measurements of flow rates and on samples of influent effluent for some rainfall events per year. In order to draw pertinent scientific and operational conclusions from the measurement results, it is absolutely necessary to use appropriate methods and techniques in order to i) calibrate sensors and analytical methods, ii) validate raw data, iii) evaluate measurement uncertainties, iv) evaluate the number of rainfall events to sample per year in order to determine performance indicator with a given uncertainty. Based an previous work, the paper gives a synthetic review of required and techniques, and illustrates their application to storage and settling tanks. Experiments show that, controlled and careful experimental conditions, relative uncertainties are about 20% for flow rates in sewer pipes, 6-10% for volumes, 25-35% for TSS concentrations and loads, and 18-276% for TSS removal rates. In order to evaluate the annual pollutant interception efficiency of storage and settling tanks with a given uncertainty, efforts should first be devoted to decrease the sampling uncertainty by increasing the number of sampled events.

Urban wet weather discharges are known to be a great source of pollutants for receiving waters, which protection requires the estimation of long-term discharged pollutant loads. Pollutant loads can be estimated by multiplying a site mean concentration (SMC) by the total runoff volume during a given period of time. The estimation of the SMC value as a weighted mean value with event runoff volumes as weights is affected by uncertainties due to the variability of event mean concentrations and to the number of events used. This study carried out on 13 catchments gives orders of magnitude of these uncertainties and shows the limitations of usual practices using few measured events. The results obtained show that it is not possible to propose a standard minimal number of events to be measured on any catchment in order to evaluate the SMC value with a given uncertainty.

Two main issues regarding stormwater quality models have been investigated: i) the effect of calibration dataset size and characteristics on calibration and validation results; ii) the optimal split of available data into calibration and validation subsets. Data from 13 catchments have been used for three pollutants: BOD, COD and SS. Three multiple regression models were calibrated and validated. The use of different data sets and different models allows viewing general trends. It was found mainly that multiple regression models are case sensitive to calibration data. Few data used for calibration infers bad predictions despite good calibration results. It was also found that the random split of available data into halves for calibration and validation is not optimal. More data should be allocated to calibration. The proportion of data to be used for validation increases with the number of available data (N) and reaches about 35% for N around 55 measured events.

Stormwater quality modelling is a useful tool in sewer systems management. Available models range from simple to detailed complex ones. The models need local data to be calibrated. In practice, calibration data are rather lacking. Only few measured events are commonly used. In this paper, the effect of the number and the variability of calibration data on models of various levels of complexity are investigated. The study is carried out on "Le Marais" catchment for suspended solids where 40 reliable measured events and good knowledge of the sewer system are available. The method used is based on resampling subsets of measured events among the 40 available ones. Three types of models were calibrated using subsets of events of different sizes and characteristics resampled among the 40 available ones. For each calibration, the model was validated against the remaining events to stand upon the quality of the model. It was found that the models are quite sensitive to calibration data, a problem neglected in practical studies. The use of more complex models does not necessarily improve modelling results since more problems and error sources are to be expected. The findings are specific to "Le Marais" catchment and the models used.