FR :

La contamination croissante des eaux souterraines par les fertilisations agricoles nécessite une gestion efficace de ces ressources impliquant l'utilisation combinée d'outils informatiques et de suivis de terrain. Le modèle AgriFlux a été développé afin de combler une lacune existant entre les modèles de recherche très complexes et les modèles de gestion peu flexibles. AgriFlux est un modèle de type mécaniste-stochastique, c'est-à-dire utilisant une représentation conceptuelle des mécanismes combinée à une prise en compte de la variabilité des paramètres et processus. Il permet l'évaluation de la contamination potentielle des eaux souter- raines par les fertilisants agricoles. Les modules HydriFlux (bilan hydrique) et NitriFlux (cycle de l'azote) sont actuellement disponibles. Une application du modèle est présentée dans l'optique de la gestion environnementale d'un système agricole. Le site expérimental étudié est localisé près de la ville de Québec (Canada). Il s'agit d'un limon sableux sous culture de maïs sucré (Zea Mays, L.) et recevant des fertilisations inorganiques selon les doses recommandées. Un échantillonnage de l'eau interstitielle a été réalisé sur un réseau de lysimètres avec tension durant deux saisons végétatives ainsi qu'un échantillonnage du sol durant un été. Le contenu en nitrates est déterminé dans les deux cas. Les concentrations en nitrates dans les eaux interstitielles simulées à l'aide d'AgriFlux représentent relativement bien les concentrations mesurées. Les différences observées peuvent être expliquées en partie par les conditions de grande sécheresse ayant prévalu durant la période d'étude. Les contenus en nitrates mesurés dans le sol sont moins bien représentés par le modèle. En début de saison, les variations rapides des contenus en nitrates observées au champ ne sont pas reproduites par le modèle alors que les valeurs de fin de saison sont mieux obtenues par le modèle. Malgré ces différences, la concordance au niveau des ordres de grandeur des concentrations dans l'eau obtenues du modèle et des mesures de terrain confirme l'intérêt d'un tel outil pour la gestion environnementale des contaminations agricoles des eaux souterraines.

EN :

Groundwater contamination by agricultural practices is a problem of growing concern with water resources managers. Considering the environmental importanve of the situation and the complexity of agricultural systems, models are used more than ever in parallel with field investigaffons to assist in the decision-making process. Most available models are either too complicated (many non-measurable parameters) or too simple (semi-empirical or site-specific) to be used as management tools. Such tools should conform to known theory and be structured to enable efficient analysis of field situations with minimal requirements for parameters. The AgriFlux model was developed according to these criteria. It is a mechanistic-stochastic model simulating groundwater contamination by agricultural fertilizers. It combines reliability and conceptual representations with a limited number of parameters

Most mechanistic models are also deterministic, using a single value for every parameter. Because of the important field variability of most parameters, it is pertinent to use a stochastic model, incorporating the variability resulting from field heterogeneity, measurement errors and intrinsic uncertainty related to parameter definition. AgriFlux represents soil water and nitrogen dynamics at the scale corresponding to a homogeneous area (soil type, crops, fertilizers, ..), usually the agricultural field. The soil profile is divided in homogeneous horizons or compartments. A daily tlme step is used in the calculations. The model is based on a modular structure to facilitate the inclusion of future modules describing the fate of phosphorus and pesticides. At present, AgriFlux contains two modules. The first one named HydriFlux, simulates water-related processes (precipitation snowmelt, infiltration, runoff, water uptake by plants, evaporation, percolation and drainage using soil water characteristic functions and the unsaturated hydraulic conductivlty function. The second module, NitriFlux, represents the nitrogen cycle processes (fertilizer inputs, mineralization, nitrification, denitrification, nitrogen uptake by plants and nitrate leaching). Biochemical processes are influenced by soil temperature and humidity.

To illustrate the use of AgriFlux in a resource management perspective, an application of the model was performed on an agricultural field located near Québec City (Québec, Canada). The cultivated soil consists of a loamy sand underlain by a coarse till. Sweet corn (Zea Mays, L) was grown on the plot for ten years using standard inorganic fertilizations. The environmental studies performed on the site consisted of sampling interstitial water through a series of tension Iysimeters (June 1990-November 1991) to determine the nitrate concentration of water reaching the groundwater. Soil samples were collected (June 1991-November 1991) to evaluate the potential for nitrate leaching. The parameters required to represent the field were identified from site characterization (soil data), from available agricultural data (crop and fertilization data), from literature (nitrogen cycle parameters) and from other available values (climate data). Special care was put into the identification of nitrogen cycle parameters because of their importance on the simulated results. The necessary parameters require no adjustement or optimization because they represent physically measurable values.

The simulated nitrate concentrations represent the measured values relatively well with the exception of some periods during the year especially during the summer for which the measured values are slightly overestimated by the model. This discrepancy can be explained partially by the drought of the 1991 summer months which limited the number of water samples thus restricting their representativeness. The spatial variability of measured concentrations was underestimated by AgriFlux. This can be related to the small number of field measurements compared to simulated values (the standard deviation decreases when the number of values increases), to a possible underestimation of parameter variability in the model (mainly nitrogen cycle parameters which are difficult to estima te) and to the influence of macroporosity (matric flow and macropore flow having different nitrate contents). This observation confirms the importance of using a stochastic model and the necessity of sampling at many locations in the field. The soil nitrate contents are represented with less precision by the model. Rapid variations in measured values are not found in the simulated soil nitrate contents. The overestimation of simulated summer nitrate concentrations and underestimation of simulated soil nitrate contents can be related to the drought conditions prevailing during the summer of 1991. When the soil is very dry, water extracted from tightly bonded water can contain more nitrates than freely flowing water, an effect which AgriFlux does not take into account.

The case study shows an example of an application of AgriFlux in a resource management perspective. Overall, simulated results represent adequately measured values, thus confirming the parameter values selected to represent the field under study. A water resources manager could use this set of parameters to represent a large number of scenarios consisting of alternative agricultural practices. These scenarios can be confirmed afterwards by a limited number of field investigations. The use of a model such as AgriFlux in the decision-making process facilitates and accelerates the implementation of governmental intervention .

FR :

L'étude bibliographique montre que l'oxydation de l'atrazine en milieu aqueux par 0³, 0³/H²0², 0³/UV, H²0²/UV et TiO²/UV ne permet qu'une dégradation limitée du pesticide (pas d'ouverture de l'hétérocycle azoté). Ces procédés d'oxydation conduisent aux mêmes sous-produits d'oxydation. Les composés N-déalkylés, les acétamido-s-triazines et l'hydroxyatrazine constituent les premiers sous-produits de dégradation de l'atrazine. Une oxydation plus poussée conduit par des réactions de N-déalkylation, d'hydroxylation et de déamination à la formation de produits finals relativement stables comme la déséthyldésisopropylatrazine, l'amméline, l'ammélide et l'acide cyanurique. La distribution des différents sous-produits en cours d'oxydation dépend du procédé d'oxydation utilisé, des conditions de mise en oeuvre du procédé (dose d'oxydants ou d'UV, longueur d'onde d'irradiation,...), des caractéristiques des eaux de dilution (pH, pièges à radicaux hydroxyles,...).

Les études cinétiques indiquent que l'atrazine est relativement réfractaire à une oxydation par l'ozone moléculaire (constante cinétique de l'ordre de 6 l mol-¹ s-¹ à 20 °C) et est assez réactive vis-à-vis des radicaux hydroxyles (constante cinétique de l'ordre de 2,5 10·9 mol-¹ s-¹ à 20 °C). En ce qui concerne les constantes cinétiques de réaction des radicaux hydroxyles sur les autres s-triazines, les résultats montrent que les méthylthio s-triazines sont beaucoup plus réactives que les méthoxy s-triazines qui sont elles mêmes légèrement plus réactives que les chloro et hydroxy s-triazines. Parmi les sous-produits d'oxydation de l'atrazine, la déséthyldésisopropylatrazine et l'acide cyanurique sont très réfractaires à une oxydation par les radicaux hydroxyles et par l'ozone moléculaire.

EN :

In this paper, oxidation studies of s-triazines in aqueous solution by advanced oxidation processes (O3, O3/H2O2, O3/UV, H2O2/UV, et TiO2/UV) have been reviewed.

Oxidation by-products of atrazine:

Several investigators have shown that N-dealkylated (deethylatrazine and deisopropylatrazine) and acetamido-s-triazines are the primary oxidation by-products of atrazine by O3 and by O3/H2O2 (table 1; fig. 1a). Under conditions which favored the production of hydroxyl radicals (03/H2O2), trace amounts of hydroxyatrazine may also be formed. These primary by-products are subsequently degraded to give complete N-dealkylated, deamined, dehalogenated and hydroxylated s-triazines (deethyldeisopropylatrazine, ammelide ammeline, cyanuric acid,...) (table 1). For example, oxidation of deethylatrazine by O3/H2O2 yields deethyldeisopropylatrazine as the major by-product (fig. 1b).

Identical by-products are produced by photochemical oxidation (O3/UV, H2O2/UV and TiO2/UV) (table 2; fig. 4 and 5). UV photolysis of atrazine at 253.7 nm (monochromatic radiation) yields hydroxyatrazine as the major product (=0.95 -1.0 mole of hydroxyatrazine formed / mole of atrazine photolysed (fig. 4a) whereas N and N,N'-dealkylated, deaminated and hydroxyderivatives are produced by UV irradiation in the presence of ozone, hydrogen peroxide (fig. 4b) or photosensitisers.

The s-triazine ring is found to be resistant to chemical and photochemical oxidation. Pathways for the degradation of atrazine by molecular ozone (fig. 2) and by hydroxyl radical (fg. 3) are proposed.

Kinetic rate constants:

The second-order kinetic rate constants for the reaction of molecular ozone and of hydroxyl radical with atrazine have been determined by several authors from competitive experiments or from kinetic models The rate constants for the reaction of ozone which have been measured (~ 61 mol-¹ s-¹ at ~ 20°C) indicate that molecular ozone is not very reactive towards atrazine. The rate constants which have been determined for the reaction of hydroxyl radical with atrazine by using different modes of generation of hydroxyl radicals (O3 + OH-; O3 + H2O2; PhotoFenton; H2O2 + UV) are in the order of 2 10[exp]9 - 2.5 10[exp]9 l mol-¹ s-¹ at ~ 20°C (table 3).

Rate constants for the reaction of hydroxyl radical with other s-triazines have been determined from competitive kinetic experiments. The relative rate constants show that methylthio s-triazines are far more reactive than methoxy s-triazines, which in turn are more reactive than chloro and hydroxy s-triazines (table 4). The kinetic data also confirm that deethyldeisopropylatrazine and cyanuric acid are very refractory to the oxidation by hydroxyl radical.

FR :

Pour mieux comprendre les modalités d'alimentation des sources de l'Atlas, des campagnes de coloration et d'analyses isotopiques ont été effectuées sur l'ensemble des sources. La disparité des circulations, la variabilité des vitesses et des taux de restitution, révèlent un aquifère complexe ou à plusieurs entrées. La majorité des écoulements convergent vers la source de Ain Asserdoune qui constitue une zone d'abaissement d'axe de plis. L'étude isotopique montre la variation des teneurs en 180 avec l'altitude des zones de recharge des différents exutoires et avec la saison. Les gradients définis autorisent la détermination de l'altitude moyenne des impluviums des systèmes jusqu'alors mal connus, en particulier celui de l'exutoire principal du massif, Ain Asserdoune.

EN :

The chalky massif of Beni-Mellal, which supplies many springs, spreads over a 365 km² area of 500 to 2400 meters elevation (average altitude, 1200 m). Rainfall in this sector varies between 400 mm and 700 mm/year. All springs emerge along the major coastal fault giving the Liassic massif overthrushing the Cenozoic lands of piemont. The most important is Ain Asserdoune spring with an average discharge of about 1.1 m³/s. The mainly chalky and dolomitic formations are very fractured according to preferential directions (N150-180 and N50-70) and show some karstification features at the surface. Some tracing experiments with fluorescein allowed the delimitation of the boundaries of the intake area of Ain Asserdoune Basin and the determination of the groundwater flow velocity. Two types of flow were identified with fast (100 to 231 m/h) and slow circulation (< 100 m/h). Most of circulations converge towards the spring of Ain Asserdoune (fig. 1) with an important intake area. This convergence results from an axis lowering. The variability of the circulation axes, the velocity and the rate of restitution suggest either a complex aquifer or an aquifer with several inputs.

Two water samplings (the first for the Beni Mellal Atlas) were carriet out in March and in September 1993 on the several springs, in order to study the variations of the oxygen-18 isotope. The oxygen-18 contents range between - 5.5 per thousand and - 7.7 per thousand (table 2). The most depleted waters directly exit from the Liassic aquifer or emerge at high altitudes; those which are richer in oxygen-18 emerge at the level of foothills, either from the Tertiary or from Liassic scales. The relationship between the oxygen-18 and the H-2 content of some samples indicates that the points are aligned along the world meteoric line (Craig, 1961). Even if the waters of Dir (piemont) springs are slighly evaporated because they partly originate in seeping Liassic waters, oxygen-18 content is probably mainly controlled by an altitude effect. Oxygen-18 data for local precipitations are not available; however, a local relationship between oxygen-18 content and mean recharge altitude was etablished for eleven (11) springs with small recharge areas whose mean altitude may be estimated from topography and geology. Oxygen 18 ratio is very well correlated to recharge average elevation for spring samples ( [delta]per thousand (O - 18)=- 3 - 0,0026 x altitude, r=0,96, fig. 3), when it is not for autumn samples. This gradient is very close to the regional relationship proposed by Marce (1975) and Kabbaj et al. (1978) (slope: 3 per thousand in oxygen-18 per 1,000 m ). This difference is attributed to the seasonal variations in the origin of water that feeds the springs. In the high water season, the whole basin area contributes to the recharge. In contrast, during low water period, water mainly originates from altitudes higher than the mean altitude of the basin, probably with a longer transit time. The seasonal variations are particularly marked for the Dir (piemont) springs (with a lower flux and intake area range of 800 to 1,200 m). They are weaker for the Lias springs and even non-existent for the main spring of the massif, Ain Asserdoune. In this latter case, the smoothing (- 7.6 per thousand and - 7.5 per thousand) of seasonal variation (which is also observed for temperature: 15.5 to 16°C ) gives evidence of important water reserves. A mean recharge altitude higher than 1,700 m may be proposed for Ain Asseroune.

Based on this analysis, some emergences which in the part were assigned to the same hydrologic system can now be discriminated. On the other hand, with a similar value of oxygen-18 ratio, rather distant springs can be considered as belonging to the same hydrogeological Atlasic unit. So the relationship which was quantified (slope: 2.6 per thousand in oxygen-18 per 1,000 m) is usable for the whole Atlasic range on account of regionaly similar weather. The methodology selected in this work corroborates the results obtained by hydrometric analysis concerning the very important groundwater reserves in the Beni Mellal aquifer.

FR :

Dans cette étude, des mesures électrochimiques ont été réalisées pour caractériser le comportement du laiton Amirauté dans une solution de chlorure de sodium 0,5 M en l'absence d'inhibiteur et en présence d'un mélange d'amines à longue chaîne et de dérivés thiazolés. Cette formulation est utilisée pour le traitement des eaux des circuits de refroidissement.

Les courbes courant-tension stationnaires et les diagrammes d'impédance électrochimique ont été obtenus avec une electrode tournante.

La première partie de cette étude présente les résultats obtenus pour deux concentrations du mélange et deux temps d'immersion préalable au potentiel de corrosion. Les essais électrochimiques montrent que l'effet inhibiteur est d'autant plus marqué que la concentration augmente. Une légère diminution de la protection de la surface du laiton Amirauté est observée avec l'augmentation du temps de maintien au potentiel de corrosion.

La deuxième partie de l'étude compare les effets inhibiteurs de chacun des constituants du mélange, d'une part, pour les valeurs de concentration utilisées lors du traitement des eaux des circuits de refroidissement (0,1 mg/l de solution contenant les dérivés thiazolés et 1 mg/l de solution contenant les alkylamines) et, d'autre part, pour des valeurs de concentration comparables (5 mg/l).

Pour la concentration de 0,1 mg/l l'action inhibitrice des dérivés thiazolés n'a pas été clairement mise en évidence. Ainsi, la protection contre la corrosion des circuits de refroidissement en laiton Amirauté avec le mélange est apportée principalement par les alkylamines. Celles-ci forment un film compact très protecteur à la surface de l'électrode.

Pour des valeurs de concentration plus importantes, l'addition de 0,5 mg/l de dérivés thiazolés à 5 mg/l d'alkylamines améliore la protection du matériau.

EN :

In this study, electrochemical measurements were carriet out to characterize the behaviour of Admiralty Brass in a 0.5 M sodium chloride solution in the absence of inhibitors and in the presence of a mixture of alkylamines and thiazol-derived products. This formulation is employed for water treatment in cooling circuits.

The steady-state current voltage curves and the impedance diagrams were obtained with a rotating disc electrode in order to control the hydrodynamic conditions.

The first part of this study deals with the results obtained for two concentrations of the mixture and for two immersion times at the corrosion potential. It has been shown that the inhibitive effect is more marked when the concentration increases from 10 to 50 mg/l. A slight decrease of the protection of the brass surface is observed when the immersion time increases

The second part of this study is a comparison of the inhibitive effects of each compound of the mixture, on one hand, for concentration values used in the water treatment (0.1 mg/l of tbe solution containing the thiazol-derived products and 1 mg/l of the solution containing the alkylamines) and on the other hand, for identical concentration values (5 mg/l ). For 0.1 mg/l, no inhibitive action of thiazol-derived products could be detected. Thus, the corrosion protection of cooling circuits in Admiralty brass is attributed essentially to the alkylamines. They form a thick, protective film at the electrode surface.

For higher concentration values, the addition of 0.5 mg/l of thiazol-derived products to 5 mg/l of alkylamines improves the corrosion resistance of brass.

FR :

Cette étude présente l'application d'un modèle stochastique de prédiction de la température de l'eau en rivière. L'analyse porte sur les variations imputables aux conditions naturelles et sur une évaluation des performances du modèle une fois appliqué au ruisseau Catamaran au Nouveau-Brunswick (Canada).

Ce modèle stochastique est développé selon l'approche de Box et Jenkins (1976) basée sur les séries temporelles des températures de l'eau et de l'air. Le modèle a été calibré avec des données de 1990. L'évaluation de performance comprend une analyse des séries résiduelles et le calcul des erreurs quadratiques moyennes. Les résultats montrent que l'erreur quadratique mensuelle varie de 0,42 °C en juillet 1990 (année de calibration) jusqu'à 2,96 °C en septembre 1992. Finalement, une discussion est menée pour souligner les avantages et les inconvénients relatifs à cette approche.

EN :

Water temperature is a very important parameter not only in water quality studies but also in biological studies. For instance, salmonids can be adversely affected by natural high stream water temperatures or by those resulting from anthropogenic sources such as deforestation. To predict stream water temperatures, two different approaches have been used; the deterministic and stochastic approaches. The deterministic approach consists of a physical model based on the energy budget (solar radiation, convection, etc.) and the physical characteristics of the stream (water depth, stream cover, etc.). The stochastic modelling approach consists of studying the structure (autocorrelation) of the stream water temperature time series and its dependence on air temperatures (cross-correlation).

The purpose of this study is to develop and test the performanoe of a stream water temperature model using a stochastic approach to predict water temperatures in rivers under natural conditions. The performance of such an approach was tested using data from Catamaran Brook, a small stream in New Brunswick (Canada). It differs from previous studies in that most others were on larger river systems.

This stochastic model incorporates the Box and Jenkins method (1976) which relates the time series data to both water and air temperature residuals. To calculate the residuals of both air and water temperatures, a seasonal component was first estimated using Fourier series analysis. This seasonal component better represents the long-term trend in air and water temperatures for the studied period or season (i.e. increasing water temperatures at first, then reaching a maximum during the early part of August and decreasing again later in the season). The Fourier series with one harmonic was chosen for the analysis as it has been shown in previous studies that the first harmonic represents most of the variation within the stream water temperature variable. The model was calibrated using the Box and Jenkins method and Catamaran Brook data from 1990. This analysis consist of determining a transfer function relating present water temperature residuals to past water and air temperature residuals including present air temperature residuals and a random component. The random component (also called « noise series ») of the model is a normally distributed variable with a standard deviation calculated using the calibration period. After the calibration period, subsequent years or post-calibration years were analyzed to predicted stream water temperatures with the model using air temperature data only.

A study of residuals between predicted and measured stream water temperatures showed very good results during the calibration year (1990) with a calculated root-mean square error of 0.75°C. The predicted temperatures during post calibration years (i.e. 1991 and 1992) were good and the root-mean-square errors were similar to previous studies (e.g. Marceau et al. 1986) with values of 1.45°C and 2.10°C respectively. The measured stream water temperatures during the post-calibration years were only used to estimate the relative performance of the model as opposed to a forecasting model which utilizes actual measured temperatures.

At Catamaran Brook is has been observed that natural variation in air temperatures can have an influence on the performanoe of the model. When air temperatures were recorded higher or lower than the long term values (normal temperatures) calculated by the Fourier series analysis, the predicted water temperatures was not as good. For instance it was observed that during September of 1992, during which time the air temperature was higher that normal, the performance of the model was not as good with a root me an squared error of 2.96°C. However, during July 1992, below normal air temperatures were also recorded and a very good prediction of stream water temperatures in Catamaran Brook was achieved with a root me an square error of 0.98°C. In general, satisfactory prediction of stream water temperatures was achieved using the Box and Jenkins stochastic modelling approach.

FR :

Les processus hydrologiques variant dans l'espace et dans le temps en fonction de la variabilité spatio-temporelle des entrées météorologiques et de l'occupation du sol ainsi que de la variabilité spatiale de la topographie et de la nature du terrain, un modèle conçu pour bénéficier des données provenant de la télédétection et des SIG a été développé. Les principaux objectifs poursuivis étaient: l'application au plus grand nombre de bassins possible, une sélection d'algorithmes permettant de tenir compte des données disponibles, un minimum d'étalonnage, la facilité de transfert d'un bassin à l'autre, la programmation des algorithmes sur micro-ordinateur avec une interface très conviviale.

La structure d'écoulement à l'intérieur d'un bassin versant est obtenue de manière informatisée à partir d'une discrétisation des altitudes rencontrées dans la zone d'intérêt, en mailles carrées de dimensions données, d'où l'on tire les pentes et les orientations de chaque maille, puis le bassin versant en amont de la maille identifiée comme exutoire, le réseau hydrographique, les sous-bassins versants et, finalement, les unités hydrologiques relativement homogènes (UHRH), définies sur la base de ces sous-bassins, en les regroupant ou les divisant au besoin. Un logiciel spécifique pour ce faire a été développé: PHYSITEL.

Selon la conception très modulaire adoptée pour HYDROTEL, chaque sous-modèle offre généralement plus d'une option de simulation, afin de tenir compte des données disponibles sur le bassin versant traité. Les simulations peuvent être réalisées en considérant comme unité de simulation du bilan hydrologique vertical les mailles originales constituant le bassin ou les UHRH.

Des exemples de simulation des débits à l'aide du modèle HYDROTEL sur divers bassins versants situés au Canada (Québec, Ontario, Colombie-Britannique) et dans le sud de la France sont présentés.

Ces simulations indiquent que les différents algorithmes du modèle réagissent bien et qu'il est possible de considérer son application sur des bassins très divers situés sous des conditions climatiques variées. Des tests plus poussés sont en cours afin de mieux cerner la sensibilité des divers algorithmes aux données d'entrée ainsi que leur interchangeabilité.

EN :

As hydrological processes vary both in space and time as a function of meteorological inputs, land-use, topography and soil type, to mention only those, a model able to make the best use of data from remote sensing and geographic information systems (GIS) has been developed. One of the objectives in developing HYDROTEL was to be able to apply the model to as many watersheds as possible, with a minimum of calibration. Also, as the availability of data varies both in type and density from watershed to watershed, it was considered necessary to develop a model offering a choice of algorithms allowing adaptation of the model to data availability on various watersheds. Another objective was to program HYDROTEL on a micro-computer with a user-friendly interface.

The complete drainage structure of a watershed is obtained with PHYSITEL, a software program designed specifically to prepare the watershed database for HYDROTEL. The area of interest is first discretized in square cells allowing the creation of a digital elevation model (DEM), with a pre-determined accuracy, from which the slope and aspect of each cell are obtained next. The aspect of each cell being known, that is the direction of flow from cell to cell, it is necessary to identify the cell considered as the outlet of the watershed to identify all cells upstream of that cell, with a recursive algorithm. When all cells constituting a watershed are identified, together with the drainage structure, it is possible to trace the river network corresponding to cells draining a number of cells greater than a specified threshold. Finally, sub-watersheds are determined, with outlets at the river junctions. Those sub-watersheds can be further sub-divided or grouped to obtain relatively homogeneous hydrological units (RHHU).

A modular approach has been adopted with HYDROTEL allowing easy addition or modification of algorithms. A choice of algorithms, selected when possible for their compatibility with remotely sensed and GIS data, is generally offered for each sub-model. Moreover, it is possible, instead of choosing one of the available simulation options, to read data from disk. This permits using, for instance, rainfall data estimated from weather radars by another program. Also, one can decide to activate only specific sub-models for a run and read data from disk for the others or just ignore the others, if they are not needed for the run. The simulation runs can be done, using as a unit for the estimation of the vertical water budget, the original cells or the RHHU's.

For the precipitation sub-model, solid or liquid precipitations, together with air temperatures, are interpolated to each simulation unit either by the Thiessen method or by a method leading to a weighted average of the measured amounts at the nearest three stations, taking into account, if desired, of the precipitation and temperature lapse rates. Daily variation and metamorphism of the snowpack are estimated by a modified degree-day method in which the energy budget at the snow-air interface is estimated by the degree-day approach but that within the pack by a more physical approach. Four equations are available to estimate potential evapotranspiration, those of Thornthwaite, Linacre, Penman-Monteith and Priestley-Taylor, in order to use the best equation for a given data set. The vertical water budget is simulated by the vertical algorithm of the CEQUEAU model or by a new algorithm more suited to remote sensing and GIS information. A kinematic wave approach is used to estimate downward flow from cell to cell, whereas river routing is simulated with the kinematic or diffusive wave equations. When the vertical water budget is done for a RHHU, internal routing of the available flow within the RHHU is performed through the use of a geomorphological hydrograph derived from its drainage structure.

The HYDROTEL model has been applied to watersheds located in Québec, Ontario and British-Colombia in Canada and to one in Southern France, in order to test its applicability to watersheds of different types and areas in various climat es. The results obtained on those watersheds, using the available algorithms, show that the model does bave a normal reaction to precipitation and temperature impulses on all watersheds. At the same time, those results confirm the need for accurate spatial information, which is likely to be available more from remotely sensed and GIS data. A new version of HYDROTEL is now under development and it will be run on OS\2 and other environments.

Finally, with its simulation options allowing monitoring of various variables during a simulation run, HYDROTEL appears to be a good tool for understanding and managing phenomena related to hydrological processes.

FR :

Incluse dans une étude plus générale de caractérisation de la Matière Organique des eaux naturelles superficicielles, cette partie présente les résultats obtenus par ultrafiltration. Les différents paramètres analytiques suivis concernent le COT, COD, CODB, UV à 254 nm et le potentiel de formation des trihalométhanes (pTHM). Les méthodes expérimentales, mises en oeuvre sur les eaux brutes et clarifiées (notées eb et ec respectivement), comprennent l'ultrafiltration (seuils de coupure: 10, 1 et 0,5 kilodaltons), les demandes en chlore et en bioxyde de chlore ainsi que l'étape de clarification, si sur les usines une préoxydation existe.

Le seuil de coupure de 10 kd ou 1 kd caractérise l'origine d'une eau brute. Les eaux de barrage, contrairement aux eaux « courantes », ont un classement logique des différentes fractions des masses moléculaires apparentes et ceci quel que soit le paramètre analytique suivi (COT ou absorbance UV). Sur toutes les eaux clarifiées ultrafiltrées, la fraction supérieure à 1 kd reste majoritaire. Dans nos conditions experimentales, ce seuil de coupure induit le changement le plus significatif entre une eau brute et son eau clarifiée. La valeur de l'absorbance UVec permet dans un premier temps d'approximer la valeur du COTec.

Les autres paramètres analytiques (CODB, pTHM et demandes en oxydants) aussi bien sur les eaux brutes que clarifiées n'ont pas pû être corrélées avec un ou plusieurs autres paramètres physico-chimiques de ces mêmes eau.

EN :

The following paper presents the work being carried out in a general study of natural organic matter. The three lines of work chosen by SAUR concern:

- The fluorescence of natural waters or humic substances (SlMON et al., 1993);

- The determination of dissolved hydrophobic and hydrophilic substances on each step of a water treatment plant (CROUE et al, 1992 a and b);

- The molecular weight fractionation of dissolved organic matter, subject of this paper.

The different raw or clarified waters (respectively rw and cw) were sampled in SAUR water plants. Some inorganic analytical characteristics of some raw waters are given in table 1. All raw and clarified water samples were characterized according to total organic carbon (TOC), UV absorbance (254 nm) and trihalomethane formation potential (THMFP) under the following conditions: about 20°C, 4 mg Cl2/mg TOC, and a 72-hour contact time in the dark. A Dohrmann DC80 and a Uvikon 930 were used for the determination of TOC (DOC) and UV-absorbance at 254 nm, respectively. When a preoxydation step was employed at the water plant, the clarification treatment was performed with a laboratory apparatus described in another paper (LEFEBVRE and LEGUBE, 1990). Bioeliminable Organic Dissolved Carbon in water was determined by the method described by JORET and LEVI (1986). Cl2 and ClO2 demands of raw and clarified waters were conducted as batch operations with oxidant doses of 1, 2 and 4 mg oxidant per mg TOC. Residual chlorine and chlorine dioxide in solutions were determined respectively by spectrophotometric measurements by two colorimetric methods: the DPD and ACVK methods.

Ultrafiltration was conducted on raw and clarified waters to determine apparent molecular weight distribution (AMW). Organic matter distribution was determined with an Amicon system (stirred cell) with membranes characterized by nominal cutoff limits of 10, 1 and 0.5 kilodaltons. For each cutoff membrane, we took the initial water and a new membrane was used for each filtration. All permeates were characterized according to TOC and UV absorbance.

In the case of clarified waters, the results included all the data, whatever the coagulant salt and its applied dosage.

As shown in figures 1 and 2, no relation exists between UVrw absorbance or the ratio UVrw/TOCrw and TOCrw. For the same raw water, the ratio UVrw/TOCrw fluctuates according to a change between fulvic, humic and hydrophilic acids during the seasons (fg. 3). A better relation between UVcw and TOCcw was obtained: UVcw (254 nm, 5 cm)=0.1091 TOCrw - 0.0231 (r2=0 758), demonstrating that organic matter, which resists coagulation, has quite the same aromaticity whatever the water source (fig. 9). An approximation of TOCcw can be done by a simple measurement of UVcw at 254 nm. Figure 4 shows a correlation between DOCrw and TOCrw (DOCrw (mg/l)=0.888 TOCrw + 0.094; r²: 0.945).

Except for 2 waters, UV absorbance removal was greater than TOC removal (fig. 13).

The AMW distribution of all the studied waters was influenced in the same degree by clarification, with a shift in DOC and UV absorbance to the < 1 kd fraction (fig. 14). Organic matter with AMW above 1 kd was essentially removed during clarification, since TOC concentrations were found slightly to be lower or unchanged in the < 1 kd fraction. But in some cases (e.g.: Charente, table 4) a significant removal of the < 1 kd fraction was obtained. Decrease in the relative UV/TOC ratio between raw and clarified waters was also observed with the different fractions. Neverthless, in the case of 4 clarified waters (table 2), the > 10 kd fraction can represent at most 49.7% and 57.3% respectively of TOCcw and UVcw absorbance. All of the coagulants (ferric chloride, aluminium sulfate or prepolymerised alum salts) were completely in effective in removing dissolved organic matter with an AMW of legs than 500 daltons. The significant removal of high molecular weight compounds is followed by a small reduction in the THMFP expressed in µg per mg of TOC.

The THM formation potential (THMFP) of raw waters is affected by the water source. Chloroform accounted for 85.4% of the THMFPrw and its average value was 63 µg per mg TOCrw. The clarified waters showed a marked difference in their THMFP distribution. Chloroform percentage was very different, ranging between 30 and 93% of THMFPcw.

Clarification treatment slightly reduced the Cl2 and CIO2 demands of the waters: 2.7 mg Cl2/mg TOCrw and 23 mg Cl2/mg TOCcw; 25 mg ClO2/mg TOCrw amd 2.3 mg ClO2/mg TOCcw.

An estimation of TOC removal by coagulation with aluminium salts can be done by ultrafiltration of this unknown water with a nominal cutoff of 10 kd (fig. 15).

The value of the fraction above 10 kd or below 1 kd could be a useful parameter for the determination of the type of an unknown water, whatever the analytical parameters (tables 3a and 3b). In the case of reservoir raw waters, the order of fractions distribution was logical, whatever the analytical parameters (TOC or UV absorbance): the > 10 kd fraction; the 10 kd >> 1 kd fraction and the < 1 kd fraction (fig. 6 and 7). Figures 6 and 7 do not show the same sequence in the case of river water (e.g.: Mayenne, Garonne, Thames,.. .).

No relation exists between BDOCrw and TOCrw or UVrw/TOCrw ratio (Fis. 5a and 5b). BDOCrw represents 0 to 36.4 percent of TOCrw. BDOC removal by coagulation is very different from one water to another (table 5), varying from 35% to 100%.