نبذة مختصرة : International audience ; This article aims to understand how to extrapolate in space and time discrete measurements in order to calculate physico-chemical indicators in rivers, which are required by the Water Framework Directive. Linked to this issue, few questions are addressed. Does the French National Basin Network provide enough information in order to make consistent water quality maps? How does the temporal indicator - the 90 percentile - vary in space? The outputs of the ProSe model applied to the Seine River are used to compare two different methods for calculating the 90 percentile: the classical method based on the empirical percentile function and a method that aims to reduce the estimation bias of the 90 percentile. This second method includes temporal weighting and linearization o the empirical percentile function, and therefore its application is a little more complex. But with this method the bias induced by irregular and/or few measurements is reduced. Three methods for spatializing the 90 percentiles have been tested in order to obtain occurrence percentages of the percentiles for each quality class. The first one is based on the "failure principle" and consists in keeping only the worst site for the considered "River Water Body". The second one respects the proportion of percentiles located in each quality class, while the third one allocates an influence segment to each measurement site. Spatializing temporal percentiles in "River Water Bodies" by influence segments leads to a marked improvement of occurrence percentage estimations and reveals the necessity to take into account the spatial configuration of measurement sites when calculating a quality indicator.
Processing Request
No Comments.