Analyse des tendances des séries temporelles de la concentration en azote minéral dissous dans les trois principaux estuaires de la façade Manche-Atlantique

Estuaries are areas which, from their structure, their fonctioning, and their localisation, are subject to significant contribution of nutrients. One of the objectif of the RNO, the French network for coastal water quality monitoring, is to assess the levels and trends of nutrient concentrations in estuaries. A linear model was used in order to describe and to explain the total dissolved nitrogen concentration evolution in the three most important estuaries on the Chanel-Atlantic front (Seine, Loire and Gironde). As a first step, the selection of a reliable data set was performed. Then total dissolved nitrogen evolution schemes in estuary environment were graphically studied, and allowed a resonable choice of covariables. The salinity played a major role in explaining nitrogen concentration variability in estuary, and dilution lines were proved to be a useful tool to detect outlying observations and to model the nitrogenlsalinity relation. Increasing trends were detected by the model, with a high magnitude in Seine, intermediate in Loire, and lower in Gironde. The non linear trend estimated in Loire and Seine estuaries could be due to important interannual variations as suggest in graphics. In the objective of the QUADRIGE database valorisation, a discussion on the statistical model, and on the RNO hydrological data sampling strategy, allowed to formulate suggestions towards a better exploitation of nutrient data.

Etude écologique de projet. Site de Flamanville. 1er cycle. Traitements mathématiques, Rapport général, Annexes

The study of a first annual hydrobiological cycle on the site of Flamanville was carried out through 19 campaigns conducted between July 76 and June 77. Twelve of them are considered as "heavy" and took place on the following dates: July 8th, 76; August 6th, September 4th, October 3rd, November 3rd, December 16th, January 5th, 77; February 2nd, March 3rd, April 14th, May 10th and May 24th. Seven campaigns consist only in one sample July 23rd, 76; August 21st, September 16th, November 19th, January 19th, 77; February 17th and June 16th. However, some variables which had not been measured during these "light" missions have not been taken into account for the analysis. A first global analysis groups together the first ten campaigns (in May, the nutritive salts had not been measured) and presents the annual variations of nine parameters: temperature, salinity, oxygen, nitrates (N0 3), nitrites (N0 2), phosphates (P0 4), silicates (Si0 2), chlorophyll and pheopigments. A second study of the annual hydrobiological cycle includes all campaigns but only five variables: temperature, salinity, oxygen, chlorophyll and pheopigments.