Abstracts
Résumé
Des échantillons d'eau de rivière ont été prélevés près de l'embouchure de la Venoge, affluent du Léman, au cours de cinq crues survenues entre octobre 1986 et novembre 1987. Les échantillons ont été extraits de grands volumes d'eau à l'aide d'une centrifugeuse à débit continu. La concentration de matières en suspension, et les formes du phosphore de ces échantillons ont été dosés. L'interprétation de ces données, basée sur une méthode de régression progressive, montre que la concentration en phosphore particulaire total peut être estimée par deux paramètres, la fraction de suspensions supérieure à 63 µm et la concentration en phosphore réactif dissous. Par ailleurs, puisqu'il existe des relations statistiques entre le phosphore particulaire total et les autres formes du phosphore particulaire, il est possible d'évaluer la concentration de ces dernières (formes de phosphore apatitique, non apatitique, organique et inorganique). Le phosphore total dissous peut être pareillement estimé en fonction du phosphore réactif dissous. L'estimation des formes du phosphore et ainsi que celle de leur charge en crue peut donc se faire même lorsque le volume des échantillons est limité, c'est-à-dire, lorsque la quantité de matières en suspension n'est pas suffisante pour l'analyse de toutes les formes du phosphore particutaire.
Abstract
In order to gain a better understanding of phosphorus transport in a storm-dependent river system, water samples were collected near the mouth of the River Venoge, Switzerlang, during five storm events between october 1986 and november 1987. Suspended sediment (SS) was extracted from large-volume water samples by continuous flow centrifugation. Soluble and particulate forms of phosphorus were subsequently analysed with the centrifuged and filtered (0.45 µm) waters, and freeze-dried SS in a < 63 µm fraction.
However, the sampling usually performed at more than 4-hour intervals could not guarantee that no information had been missed, for example, the instantaneous fluctuation of soluble reactive phosphorus (SRP) during the events. Although more frequent sampling using an auto-sampler is feasible, the small volume of raw water cannot provide sufficient sediment for all the analyses of phosphorus forms. Thus, it would be useful to find a model capable of estimating phosphorus concentration in different forms, whenever the present measurements are not possible.
A multivariate progressive analysis of the measured phosphorus data set shows that total particulate phosphorus (TPP) concentration can be estimated as a function of two parameters, the percentage of a SS fraction coarser than 63 µm and the SRP concentration measured in filtered water. On the other hand, general statistical relationships exist between the various forms of phosphorus. Total soluble phosphorus (TSP) is dependent upon SRP. Organic phosphorus (OP) and non-apatite inorganic phosphorus (NAIP) can be approximately assessed from TPP, measured or calculated. Then, apatite phosphorus (AP), inorganic phosphorus (IP) and total phosphorus in raw water can be calculated by means of summation/substraction operations.
Modelling is apparently suitable to the storm events during which only a limited volume of water samples could be collected. It also provides a rapid way to estimate the partitioning of phosphorus loads in high flow periods of the river system, thereby reducing the field and laboratory work required.
Keywords:
- Phophorus forms,
- river,
- flooding,
- storms,
- modelling