FR :

Le devenir de l'azote(N) et du phosphore (P) a été étudié dans un chenal à haut rendement (CAHR) au cours de la période d'adaptation puis en phase stationnaire. En moyenne, la part de N total perdue atteint 34,5% du N admis en période d'adaptation et 24,5% en phase stationnaire tandis que la part assimilée par les algues atteint respectivement 32 et 43,5%. L'azote organique particulaire s'accumule dans le sédiment et subit une minéralisation lente pendant la phase d'adaptation puis rapide en phase stationnaire. Les pertes nettes en N dans le CAHR seraient dues à la sédimentation et à la volatilisation de NH₃ en saison froide et seulement au dernier mécanisme, en saison chaude. Le bilan de l'azote ammoniacal en phase stationnaire montre que l'importance de la biomasse algale produite en saison chaude ne peut s'expliquer qu'en présence d'une minéralisation active du sédiment. Le taux de minéralisation est estimé à 12,4 kg N ha^-1 j^-1 en saison chaude contre 1,3 en saison froide. En moyenne, la part du P total perdue atteint 27% du P admis en période d'adaptation et 17,5% en phase stationnaire alors que la part assimilée par les algues atteint respectivement 25 et 17,5%. En phase stationnaire, l'élimination du P soluble dans le CAHR serait due à l'assimilation algale (54%) et à la précipitation chimique sous forme de sels de phosphates (46%).

EN :

The fate of nitrogen (N) and phosphorus (P) was studied in a high-rate algal pond (HRAP) during the start-up period and under steady-state conditions. The HRAP was first used by Oswald and co-workers in Berkeley (California, USA) in the late 1950s for wastewater treatment with algae collection as a by-product for Single Cell Protein production. The HRAP consisted of a raceway shaped, shallow pond (less than 0.50 m deep) equipped with a continuous and gentle mixing device (paddle wheel) generally operated at 8 rpm. The paddle wheel is not used for aeration purposes.

In this study, the HRAP is included in a wastewater treatment plant aimed at producing good quality effluent for agriculture reuse purposes. The treatment plant includes three components with the HRAP placed in the middle of the pond train. Up-stream, there is a two-phase anaerobic reactor (TAR) and downstream, two maturation ponds in series. Assigned removal tasks for the plant components are as follows: i) organic matter, suspended solids and helminth eggs for the TAR; ii) N and P for the HRAP; and iii) fecal coliforms for the two maturation ponds. The capacity of the plant is 1,500 inhabitants (70 m³/day), total hydraulic retention time is 8 days and the total occupied area is 1,500 m² including walkways and other facilities. The plant was constructed in December 1996 and has been continuously operated since that time.

The present paper focuses on the fate of N and P in the HRAP in order to evaluate its performance and to shed light on the mechanisms behind N and P removal, with the ultimate goal to improve surface and groundwater protection from pollution by wastewater nutrients. Under steady-state conditions and when we consider the soluble part of N and P taken up by the algal cells together with that lost (either by stripping or chemical precipitation), the total removal averaged 70% for N and 40% for P. Such important removal must be highlighted especially because the system relies on solar energy with no electromechanical aeration, is cheap to construct and easy to operate and maintain.

To understand the mechanisms behind such a performance and also to gain experience for the process start-up and for rational operation and maintenance of future plants, an in-depth analysis on the fate of N and P was done based on the four-year follow up data. Total N losses reached an average of 34.5% in the start-up period and 24.5% under steady-state conditions, whereas N assimilated by the algae represented 32 and 43.5%, respectively. Particulate organic nitrogen (PON) accumulated in the bottom of the unit did undergo rapid mineralization under steady-state conditions. Mechanisms involved in N losses might be attributed to settling of PONr and, to much lesser extent, NH3 volatilization in the cold season, whereas the latter mechanism might occur primarily in the hot season. The rates of mineralization of HRAP sediment were estimated to be 12.4 and 1.3 kg of N ha^-1 d^-1 respectively for the hot and the cold season.

On average, total P losses in the HRAP reach 27% in the start-up period and only 17.5 % under steady-state conditions, whereas algae assimilation represented 25 and 17.5% respectively. Under steady-state conditions, algae uptake (54%) and chemical precipitation (46%) were found to be the main mechanisms for P removal in the HRAP.

FR :

Le but de cet article est de présenter une méthodologie de mise à jour des paramètres de modèles pluie-débit en période de crue. Elle a été mise au point afin d'améliorer un des aspects de la gestion des réservoirs dans un contexte opérationnel de protection contre les crues: la réduction des incertitudes sur la prévision des débits. L'originalité de la méthode proposée réside dans le fait que l'on utilise non seulement une information sur les débits mais aussi une information sur l'humidité du sol. L'objectif de l'étude est d'évaluer l'intérêt de l'introduction de cette information supplémentaire. Pour cela, les données d'humidité du sol sont introduites au sein du modèle par l'intermédiaire d'une relation de passage établie entre l'humidité mesurée in situ et l'humidité calculée implicitement ou explicitement par les modèles. Cette méthodologie a été testée dans le cadre du projet européen AIMWATER sur quatre sous-bassins de la Seine en amont de Paris (France). Deux modèles pluie-débit sont utilisés dans cette étude, un modèle conceptuel semi-emprique et un modèle conceptuel couplé à un schéma de surface simulant une interface sol-végétation-atmosphère et permettant de calculer l'évolution de l'humidité du sol à différentes profondeurs. Cette approche comparative étudie l'intérêt d'un tel modèle couplé par rapport au modèle conceptuel semi-empirique sans représentation explicite des phénomènes se produisant à l'interface sol-végétation-atmosphère.

EN :

Improving the accuracy of rainfall-runoff models and in particular their performances in flood prediction is a key point of continental hydrology. Methods have been developed to improve flood prediction in hydrology based on a better compliance of the model with current observations prior to its use in forecasting mode. This operation has been termed updating in hydrology and assimilation in meteorology. The fundamental idea is that if model predictions diverge from observations at a given time, there is little chance that future estimations will approach correct values. The improvement then comes from a correction of the trajectory of the model based on observations during the period preceding the day when a prediction into the immediate or long-term future is desired. This can be dealt with by a correction of model parameters, which is usually called "parameter updating".

The inability of rainfall-runoff models to produce correct streamflow values generally translates into parameter uncertainty. Parameter calibration is the means used by a model structure to adjust to a given set of data. Therefore, a parameter updating methodology seems to be a natural way to amend errors in streamflow values. In this paper, a specific methodology of parameter updating is presented. The main feature of this method is that it does not carry out updating by reference only to recent streamflow observations, as classic procedures do, but also to soil moisture measurements, which can be retrieved daily from TDR probes. Indeed, it appears that the integration of soil moisture data allows better control of the evolution of the model and improves its performances, in particular in terms of forecasting.

The aim of the research was to assess the usefulness of this additional soil moisture information. To this end, an approach has been suggested that gradually introduces additional information thanks to a constraint relationship between observed and modelled soil moisture. In fact, soil moisture can be calculated implicitly or explicitly by the model when extracting step-by-step the values of the model's store contents. This methodology was put forward for use in the European AIMWATER project on four catchments within the Seine River basin upstream of Paris (France).

The other issue addressed in this paper was whether or not it is necessary to use a model that simulates explicitly the evolution of soil moisture at different depths. One can argue that if the model employed does not feature a store that can be identified closely to the observed soil moisture, there would be no possibility of benefiting from such measurements. On the other hand, it can be argued that if soil moisture is a model output, all the information drawn from soil moisture observations will be directed at improving this specific output at the expense of improving streamflow values. To answer this issue, two models were tested. The first model, GR4j, has no explicit counterpart for soil moisture measurements. The second one, GRHum, has been especially developed to introduce a two-layer soil reservoir that simulates the surface and sub-surface soil moisture.

Since the aim of the present research was to analyse different ways of accounting for soil moisture, and to identify the one that offers the best prospects, several tests were carried out, using different relationships between observed and modelled soil moisture. Indeed, TDR probes give point measurements of soil moisture at several depths and several store contents can be taken into account in a constraint relationship.

First, for both GR4j and GRHum models, tests showed that performances for flood forecasting are significantly improved when assimilating in situ measurements of soil moisture at a daily time-step, especially for the basins where poor simulations are obtained. It is also noteworthy that performances are very dependent on the items taken into account in a constraint relationship.

Secondly, the GRHum model did not appear to be more efficient than the GR4j model when assimilating both streamflow and soil moisture data. However, the GRHum model gave the best results when assimilating only streamflow data, and superficial soil moisture seemed to fit the GRHum better than the GR4j model.

Finally, although the tests required perfect foreknowledge of rainfall, the results of the research are encouraging from an operational point of view. Another interesting perspective is provided by the Earth Observation data. Indeed, previous studies have shown that soil moisture can be derived from EO data using, for example, microwave spaceborne Synthetic Aperture Radar (SAR) images (QUESNEY et al., 2000). This type of catchment-scale data could be more relevant than a local measure given by TDR probes (PAUWELS et al., 2002).

FR :

Une étude à l'échelle centimétrique de l'interface redox situé à la limite entre mixolimnion et monimolimnion d'un lac méromictique (le lac Pavin) a permis d'observer très finement l'évolution de la concentration d'un certain nombre d'éléments chimiques. Nous avons choisi de présenter ici des résultats concernant 5 éléments qui présentent des comportements très contrastés : le rubidium, le fer, le baryum, le vanadium et le manganèse. La comparaison avec un élément conservatif, le sodium, montre que Rb est conservatif, que Fe, Ba et V sont précipités et que Mn est dissous dans cette zone.

Une modélisation de ces concentrations en vue de préciser à quelle profondeur et avec quelle vitesse se produisent les réactions concernant ces éléments nécessite la détermination des paramètres de transport au voisinage de cet interface.

Une représentation analytique des concentrations de sodium permet de calculer le coefficient de diffusion turbulente K_z en fonction de la profondeur. Au voisinage de l'interface redox, ce coefficient est très petit (0,0017m²/jour) et inférieur au coefficient de diffusion thermique moléculaire.

Les concentrations des éléments étudiés ont pu être représentés avec précisions par des polynômes en fonction de la concentration en sodium.

Cela permet d'estimer les vitesses des réactions de précipitation dissolution en fonction de la profondeur. Le rubidium n'est affecté par aucune réaction. Le fer précipite entre 63 et 65 m, le baryum entre 68 et 72 m tandis que le vanadium précipite à la fois dans ces 2 zones. Le manganèse réagit dans une zone très étroite : il est précipité entre 61,5 et 62 m et dissous entre 62,8 et 63,1 m.

Une étude similaire de tous les éléments majeurs (y compris pH et COD) pourrait permettre d'élucider les processus qui conduisent à ces comportements complexes.

EN :

Lake Pavin, French Massif Central, is the main meromictic lake in France and has been extensively studied from more than 50 years. The upper part (mixolimnion) at a depth of less than about 60 m behaves as an oligotrophic lake and is oxic during the major part of the year. The lower layer (monimolimnion) has a higher salinity and is permanently anoxic. Unlike the top of the mixolimnion, element concentrations in the monimolimnion can be considered at steady state. The boundary between mixolimnion and monimolimnion is a redox interface. At this interface, an important number of both chemical and biochemical reactions occur.

This boundary, where element concentrations vary greatly, was studied at the centimeter scale between 58 and 64 m depth. The present paper is focused on five elements showing very different behaviour: rubidium, iron, manganese, vanadium and barium. Sodium was used as a reference element. Sodium and rubidium concentrations had similar patterns: a progressive increase began at 61 m depth and the maximal gradient was located at 63 m. They continue to increase towards the bottom of the lake. Iron concentrations were low (< 1 µmol/L) at a depth less than 62.8 m and increased very sharply below this depth. Manganese concentrations were very low in the mixolimnion(<0.01 µmol/L), exhibited a peak between 62.4 and 63.5 m depth (up to 60 µmol/L at 63 m) and reached a value of about 30 µmol/L at 85 m. Barium concentrations began to increase only at depths greater than 65-67 m. Vanadium concentrations in the mixolimnion were about 14 nmol/L, decreased to a minimum below the detection limit at 62.2 m and then increase drastically (150 nmol/L at 85 m).

In order to derive the accurate location of the chemical reactions and an estimation of their rates from the concentration profiles, knowledge of the transport parameters was needed. As advection can be considered to be negligible, the major parameter of interest is the vertical eddy diffusion coefficient K_z. Na is assumed to be unreactive in the studied layer. Its concentrations can be represented by an analytical function

      Cmax - Cmin                 Cmax + Cmin

C = ___________ * th [P(z)] + ___________

             2                                    2

with P(z)=0.0016 * (z-z_o)³ - 0.0493 * (z-z_o)² + 0.5735 * (z-z_o) - 0.4811

This allows the determination of the coefficient K_z.

K_z = λ ch² [P(z)]/ [P'(z)]

λ is determined from the value of K_z at 85 m depth previously obtained from an hydrodynamic study of the lake (Aeshbach-Hertig et al., 1999). This coefficient is about 0.1 m²/day at the bottom of the monimolimnion. It is very low at the redox interface (0.0017 m²/day), far below the molecular thermal diffusion coefficient. It increases very sharply at the bottom of the mixolimnion. The K_z profile is in fair agreement with the results obtained from the earlier hydrodynamic study.

A quantitative study of the dissolution-precipitation reactions at the center of the lake at depths between 55 and 85 m can then be undertaken. The 55 m limit corresponds to a depth where inputs of fresh water can occur. The 85 m limit is about 7 m above the bottom of the lake. Below this depth important inputs from the pore waters occur which are not taken into account by the present modeling. Concentrations of Rb, Fe, Ba and V can be accurately represented by polynomial functions of the Na concentration. The parameter u=th[P(z)] represents the concentrations of these 4 elements by polynomials :

          N

X(u) = Σ a_n * uⁿ

          0

The rate of dissolution-precipitation for each element as a function of depth can be derived.

                                       N

R = - λ [P'(z)] * ch^-2[P(z)] Σ a_nn(n-1) * th^n-2[P(z)]

                                        0

Rb concentrations are a linear function of the Na ones and therefore rubidium is not reactive. Fe concentrations can be related to sodium concentrations by a parabolic relationship. From this relationship, it can be derived that strong iron precipitation occurs in the 63 - 65 m depth layer. V concentrations are related to sodium ones by a 4^th degree polynomial. It can be derived that V deposition occurs at depths of 63-65 m and at 70 m.

Ba precipitates around 70 m depth. Mn concentrations are represented by

[Mn]=a₀ +a₁ u + b₁ exp[-(z-z°)²/z^*2]

and the derivation shows that Mn is strongly dissolved between 62.8 and 63 m and precipitated just above. These results are in good agreement with a previous study of particles fluxes derived from sediment trap analysis (Viollier et al, 1997).

This study shows the complexity of this interface and more comprehensive studies including all major elements, dissolved organic carbon (DOC) and pH are needed.

FR :

En terrain volcanique, les eaux de pluie s'infiltrent jusqu'à leur rencontre avec un niveau imperméable qui correspond le plus souvent au socle cristallin. Ce sont les talwegs et les lignes de crête des paléo- reliefs de ce socle dont la profondeur peut dépasser la centaine de mètres qu'il convient de détecter, parfois avec une précision décamétrique.

La méthode géophysique la plus utilisée en hydrogéologie des terrains volcaniques est la prospection électrique qui fournit des coupes verticales des résistivités électriques. La morphologie du substratum imperméable ou saturé peut aussi être obtenue en mesurant en surface les potentiels électriques de polarisation spontanée (en abrégé PS) qui se forment par la percolation de l'eau infiltrée dans le terrain poreux. La base de la zone non saturée, appelée surface SPS, est calculée par une relation faisant intervenir les données PS, les altitudes et deux coefficients définis à partir des données géologiques. Cette surface indique directement les axes de circulation et les lignes de partage des eaux.

Deux exemples pris sur des sites bien documentés montrent la validité de la méthode pour localiser les axes de circulation de l'eau souterraine et les limites entre bassins versants. Un troisième exemple montre les résultats PS comparés à ceux des méthodes électromagnétiques VLF et AMT. La méthode PS est légère et offre une bonne précision horizontale, mais elle demande au moins un forage d'étalonnage pour préciser la profondeur des interfaces.

EN :

Hydrogeological prospecting poses great problems in volcanic areas, particularly on volcanic islands. Despite a generally high rainfall in these areas, the high mean permeability of volcanic rocks results in only a few easily exploited resources. Volcanic rocks form a particular case of geophysical prospecting with respect to hydrogeological research. Indeed these formations are characterised by specific hydrological and geophysical parameters, which generally possess two features. First, there are heterogeneous systems composed of non-stratified layers. Acceptable modelling of these structures requires an extensive coverage of data. Generally only light equipment and simple methods can be used. Also, the electrical parameters of volcanoclastics are often very different from those of their surrounding rocks, so geoelectrical methods are generally used in volcanic areas, namely the electrical resistivity and self-potential methods.

The Self-Potential (SP) method gives very good results because SP effects in volcanoclastics are strong. In volcanic areas it has been used for the last twenty years for geothermal and hydrogeological research. It is a measure at the Earth's surface of the natural electric potentials generated in the ground by streaming potential, due to ground water flowing through the porous rock medium. This medium becomes polarised. The principles of data interpretation can be summarised as follows. There is a linear correlation between the range of negative SP anomaly and the thickness of the vadose zone when two conditions are satisfied. The first condition is a high ratio between the resistivity of the vadose zone, and the resistivities of the substratum and the water-saturated zone. The second condition is the homogeneity of the vadose zone. If these two conditions are met, it is possible to define a geophysical surface called SPS, calculated from the SP and topographic data. The SPS is both an equipotential SP surface and the interface between the vadose zone and the saturated medium below. Drainage courses and watersheds are located exactly on the valleys and the ridge lines of the SPS respectively. The survey should be carried out using a high density of measurements, because underground water circulation can sometimes be confined to narrow talwegs. Nevertheless, the analysis of profiles across a valley does not always accurately indicate a narrow flow within the valley, and it is then necessary to run several mechanical probe-tests to locate the underground channel.

Two well-documented examples focus on the groundwater flow below a basaltic lava flow, recognised partly by boreholes (Chaîne des Puys, France). In the first example, the water is channelled along a gallery where the contact between volcanic rocks and granite can be clearly observed. The valley of SPS is narrow and corresponds to the end of the gallery to within 10 m, at 75 m below the surface. In the second example, a SP map with a surface of 5 km2 shows the boundary between two basins, the second being intersected by an impermeable zone. A third example concerns the shield volcano of Piton de la Fournaise (Réunion Island, Indian Ocean), considered as a pile of permeable lavas. Four electrical methods, Self-Potential (SP), Audio-Magneto-Telluric (AMT), Very-Low-Frequency (VLF), Electrical Sounding (ES), were used to provide some information on the groundwater system. The comparison among these methods shows that the SP method is the best method to identify in detail the upper vadose zone, whereas the AMT method is able to identify deeper layers, but without much precision. The comparison between electrical resistivity and SP methods was studied from a practical point of view, in terms of horizontal and vertical precision and accessibility.

Finally, the paper describes practical aspects of the SP method, including SPS calculation methods, specific equipment, field measurement procedures, and disruptions due to anthropogenic factors.

FR :

Le but de cette étude est de déterminer si le périphyton et le gastéropode pulmoné Helisoma trivolvis peuvent être utilisés pour évaluer la capacité de rétention des métaux dissous d'un marais construit. Cette étude a été menée dans un marais de la région d'Ottawa-Carleton (Ontario, Canada), qui a été construit en 1995 afin d'améliorer la qualité des eaux de ruissellement provenant d'un bassin versant à usage résidentiel et agricole. Au cours du mois de septembre 1999, des échantillons d'eau ont été prélevés, des escargots (H. trivolvis) ont été recoltés et un substrat artificiel a été utilisé pour faire croître du périphyton, à l'entrée et à la sortie d'eau du marais. En moyenne, les calculs de balance de masse indiquaient une rétention des formes dissoutes du Cu, Mn, Ni et Zn dans le marais. Cependant, les tendances observées pour les concentrations de métaux dissous et pour les concentrations dans les organismes différaient pour certains métaux. Les concentrations de Cd et Ni dans les tissus de H. trivolvis et le périphyton étaient significativement plus élevées à l'entrée qu'à la sortie. Cependant, les concentrations de Cr et Al dans les organismes n'étaient pas significativement différentes entre l'entrée et la sortie alors que celle de Mn était significativement plus élevée à la sortie du marais. Pour tous les métaux sauf le Cd et le Zn, les concentrations dans le périphyton étaient en moyenne plus elevées que celles dans les escargots. Le périphyton peut donc fournir une mesure plus conservatrice de la contamination du milieu pas les métaux. Cette étude montre l'importance de considérer, non-seulement les mesures chimiques, mais aussi les mesures biologiques, dans l'évaluation de l'efficacité d'un ouvrage de contrôle de pollution.

EN :

Urban and agricultural development has had a significant impact on the water quality of rivers and lakes around the world. In the last few decades, constructed wetlands have been designed as wastewater treatment systems to prevent water quality deterioration in natural receiving waters. Constructed wetlands are built because they are considered sinks for many pollutants thereby protecting the water quality of downstream ecosystems. The treatment performance of these wetlands is generally assessed using mass balance calculations. However, the retention of metals by constructed wetlands is highly variable and the factors involved are still poorly understood. If wetlands are sinks for metals, the metal content of organisms should be lower downstream than upstream. In this context, organisms can be useful to assess the retention or transformation of metals by wetlands.

The objective of this study was to determine whether periphyton and the gastropod Helisoma trivolvis could be used to evaluate the retention of dissolved metals in a constructed wetland. H. trivolvis is a freshwater pulmonate snail widespread in ponds across North America. It feeds mostly on periphyton and is more or less sedentary. Snails have been used as biomonitors because several species are metal tolerant. However, compared to snails, periphytic microorganisms may track more closely dissolved metal concentrations as they take up metals principally from the water column.

This study was conducted at the Monahan Pond in Kanata, Ontario (Canada). This wetland was built in 1995 to treat run-off from an agricultural and residential watershed. Water chemistry samples, snails and periphyton grown on artificial substrata were collected at both the inlet and the outlet of the wetland in the fall of 1999. Tissue samples were digested with concentrated nitric acid and metal analyses were done by ICP-MS (Inductively Coupled Plasma Mass Spectrometry). During the experiment several chemical parameters differed between the inlet and the outlet. Alkalinity was significantly higher at the inflow and all major cation concentrations were higher at the inflow. The temperature was on average 3 ºC higher at the outlet. Mass balance calculations showed that the wetland was a sink for most dissolved metals. Snails and periphyton tissue metal concentrations were higher at the inflow than at the outlet for Cd and Ni. However, no significant differences were observed between inflow and outlet tissue concentrations for Cr and Al, whereas Mn was actually significantly higher at the outlet. As a result, the metal content of the organisms did not consistently reflect the dissolved metal concentrations in water. For all metals except Cd and Zn, periphyton concentrations were on average higher than snail metal concentrations. Periphyton analyses can provide a more conservative measure of metal contamination and, when artificial substrates are used, correspond to defined and known periods of exposure.

This study demonstrates that constructed wetlands may lead to increased metal content of downstream organisms even if these wetlands appear to be overall sinks for dissolved metals based on mass balance calculations. It also shows the need to consider not only metal concentrations, but also biological data when assessing the performance of pollution control facilities.

EN :

A very simple model for the flow of a river, obtained through linear regression, is found to give better results for a certain period when compared to the deterministic model currently in use. The comparisons between the two models are based on three important criteria: the correlation coefficient, the sum of the squares of the errors and the peak criterion. The model examined was used when the river was in spate and the forecasting horizon was a three-day period.

FR :

On trouve qu'un modèle très simple pour le débit d'une rivière, obtenu en se servant de la régression linéaire, donne de meilleurs résultats, pendant une certaine période, qu'un modèle déterministe utilisé actuellement. Les comparaisons entre les deux modèles sont basées sur trois critères importants, à savoir le coefficient de corrélation, la somme des erreurs au carré, et le critère de pointe. Le modèle est utilisé pendant la période de crue de la rivière, et les prévisions hydrologiques sont effectuées jusqu'à trois jours d'avance.