Search

The principal challenge in the parameterization of storm flow models for agricultural catchments with an artificial drainage network and fields with different degrees of tillage lies in the parsimonious definition of distributed model parameters in a way that reduces the number of calibration parameters to a justifiable minimum. This paper presents a comprehensive case study for the parameter estimation of a distributed storm flow model applied to an agricultural catchment (0.91 km2) in the Mediterranean region. Model parameterization was combined with procedures for multi-criteria, multi-storm calibration, where we automatically calibrated three parameters related to flow velocity and infiltration, and compared single and multi-storm criteria that are based on discharge volume, peak flow, and the Nash–Sutcliffe coefficient. Multi-storm calibration yielded a set of parameter values for the simulation batch with best multi-storm overall performance, which are close to the median values in the pre-calibration of individual storms. Our results suggest that flow velocities and proportionality of the channel infiltration rate do not vary significantly over the course of 11 years. Copyright © 2012 John Wiley & Sons, Ltd.

International audience; Due to their significant influence on infiltration/runoff partitioning and soil erosion, hydrological soil surface characteristics (SSC) have to be known in a spatially distributed manner. This paper proposes a new approach based on radiometric and spatial remotely sensed information, for the mapping of hydrological SSC classes according to a predefined typology based on infiltration rates. Traditional remote sensing approaches usually focus on single SSC attributes only, where the latter correspond to key structural properties such as micro topography, surface crusting and soil cover. The originality of the method proposed here is considering the composite nature of SSC classes, by combining the aforementioned single attributes. This method makes use of a multiscale image segmentation that allows extracting image objects at two spatial scales of interest. At the fine scale, each Homogeneous Unit (HU) is assigned to an SSC attribute. At the coarse scale, SSC classes are identified by combining HU of SSC attributes assigned at the fine scale. The method was applied on airborne images collected over a Mediterranean vineyard by a small Pixy drone, and validated using intensive ground-based observations. The results showed acceptable performances with an overall accuracy ranging from 63 to 84%, depending on SSC classes and surface conditions. However, unsystematic confusions still remained between SSC classes which significantly differ in terms of hydrological behaviours. Improvements are expected considering richer spectral information, and ancillary information about SSC evolution in the case of intensive temporal monitoring

Spatially distributed hydrological modelling is required to understand and predict erosion, flooding and pollution risks that affect the vine cultivated Mediterranean environment. Previous field studies have demonstrated the dominant influence of soil surface features on overland flow and they therefore constitute an essential input to the hydrological model. In this paper we propose a remote sensing based method to map vineyard soil surface features with a spatial and temporal resolution appropriate for integration into the model. Our goal was to classify each soil surface portion in accordance with a pre-established, field measured infiltration capacity based typology. The radiometric characteristics of the classes of this typology were measured in the field and their Bi-directional Reflectance Distribution Function (BRDF) was modelled. Vineyard sunlit soil surface pixels were automatically extracted from high spatial resolution scanned aerial colour photographs Wassenaar et al., 2001 , Wassenaar et al., 2002 . These pixels are radiometrically classified by comparison of their reflectance with BRDF-based reflectance predictions of each soil surface type for the specific illumination and viewing geometry of the pixel. The results show that most hydrological soil surface classes have distinct bi-directional radiometric properties. For one given geometric configuration however, the predicted reflectance ranges of some classes can considerably overlap (tilled soils and stone layers for example), while others can always unambiguously be identified (bare soil crusts, surfaces covered for more than 50% by weed or litter). We conclude that our fuzzy classification approach and the simple radiometric information used, allow us to identify the majority of the hydrological surface types. The method can easily be transposed in time and space. Its performance quite strongly depends on the radiometric and geometric accuracy of the input data. Significant improvements in soil surface classification precision are expected from considering spatial context information and monitoring the soil surface evolution.

Contact: pare@supagro.inra.fr; International audience; Soil surface characteristics (SSC) influence strongly hydrological processes and are known to vary largely in space and time according to soil characteristics and soil management. Because tillage is a main source of variation, the goal of this study was to present and evaluate a prediction model of the temporal variation of the SSC after tillage at the catchment level. The study focused on bare soils prevailing in spring and summer. A logistic regression approach was used to predict the evolution along three stages, starting from the fresh tillage stage to the crusted soil stage. This method provides the probabilities of occurrence of each stage. The predictor candidates tested were a rainfall characteristic, namely cumulative rainfall depth or cumulative kinetic energy, basic soil properties and tillage features. The results showed that a model based on cumulative kinetic energy since tillage and soil stoniness accurately predicts the dynamics of SSC: the rate of well classified SSC was 91%. However, no significant difference in the prediction performance was found using as predictor either cumulative kinetic energy or cumulative rainfall amount since tillage. In the prediction model, the rainfall characteristic was the most significant predictor for the SSC evolution and the only one during the first stages of crust development since tillage. Stoniness was also shown to influence SSC evolution but only during the last stages of crust development: high stone cover speeds up soil surface evolution. The same approach using logistic regression can be applied elsewhere but will require a re-examination of the most relevant predicting variables. Finally, to be able to predict the soil surface characteristic evolution on an annual scale, weed growth characteristics must be considered in the list of predictor candidates.

Correspondance auteur: Raclot D. raclot@supagro.inra.fr; International audience; Soil erosion surveys and modeling approaches often rely on datasets collected from small experimental plots or fields. Therefore, one important issue is to assess to what extent reference data collected on small areas can be affected by soil tillage practices and extrapolated to larger areas. The objectives of this paper are to discuss: (i) the impact of superficial tillage on the erosion rates at field scale and (ii) the impact of the scale on the erosion rates for areas ranging from small fields to 1 km(2) catchments. We analyzed a dataset of 18 rainfall events including runoff and erosion data from fields of approximately 1200 m(2) and 3200 m(2) consisting of vineyards with two contrasting weeding practices - either superficial tillage or notillage with chemical weeding - and from the outlet of a 0.91 km(2) catchment that was 70% vineyard. Results showed that superficial tillage significantly reduced (4.5 times) soil losses as compared to notillage, with a large event-to-event variability. Considering the natural conditions studied by the authors, no-tillage with chemical weeding is not recommended for the study area. Erosion rates observed at catchment scale were significantly lower than those observed at the scale of individual fields. This trend, which has already been observed by many researchers, confirms that catchment soil loss cannot be estimated by the sum of individual field soil losses. In this context, the scaling transition between the field and the 1 km(2) catchment can be analyzed in terms of the spatial variability of soil management and connectivity between the sediment-producing and deposition areas

International audience; During the last decades, the European loess belt has been confronted with a significant increase in environmental problems due to erosion on agricultural land. Spatially distributed runoff and erosion models operating at the catchment scale are therefore needed to evaluate the impact of potential mitigation measures. Expert-based models offer an alternative solution to process-based and empiricalmodels, but their decision rules are only valid for the local conditions forwhich they have beenderived. The STREAMmodel,whichwas developed inNormandy (France), has been applied in two Belgian catchments having a similar soil texture, as well as in a catchment of southern France differing by soil, land use and climate characteristics. The performance of hydrological models can be assessed for instance by calculating the Nash–Sutcliffe efficiency criterion (ENS).When applied to Belgium, the model results are satisfactory to good after an adaptation of the decision rules (0.90bENSb0.93 for runoff predictions and 0.85bENSb0.89 for erosionpredictions).Given the important environmental differences between Normandy and southern France, the model rules were also adapted for application in the latter environment. Unfortunately, the quality of runoff predictions was insufficient to simulate erosion in southern France. In conclusion, STREAMis a reliablemodel providing satisfactory runoff and erosion predictions in the regionswhere hortonian overland flowdominates.Nevertheless, an adaptation of decision rules based on localmulti-scale (plot, field, catchment) data is needed, before running themodel. STREAMcan then serve as a decision support tool to design for instance flood control measures.

International audience; Spatial patterns of soil surface characteristics (SSC) are important factors for heterogeneity in runoff and infiltration processes in row-cropped fields. The objective of this paper is to analyse the spatial heterogeneity of SSC in the case of a vineyard in order to propose an adequate sampling strategy. A preliminary field survey was carried out to obtain a quantitative assessment of SSC. The experimental layout consisted of 3 wire-trained vine subplots with 3 different soil treatment. 60 observations per subplot were conducted using the line transect method. Their locations were determined according to a stratified sampling design conceived in accordance with the geometric structure of the vineyard system, defined by the alignment of vine rows and the direction of the main slope. The spatial heterogeneity of SSC was analysed at intra-transect and inter-transect levels using measurements of structural crust, sedimentary crust, grass and organic litter. The characterization of intra-transect heterogeneity, using join-count statistics, allowed patches of homogeneous units of attributes to be detected and hence an appropriate sampling resolution and an adequate transect length to be determined. A sampling resolution of 10 cm on a 1.25 m length transect was found to be appropriate for the observation of homogeneous units of attributes. Analysis of inter-transect heterogeneity, using non-spatial statistics, revealed considerable variability between rows and inter-rows that must be accounted for during field surveys. The role of site topography on spatial patterning of SSC attributes across the plot and the need to stratify the plot into areas with homogeneous slope were also demonstrated. Finally, the t-test procedure indicated that the minimum sample size was found to be 200 transect/ha for a minimum precision rate of 80% in the values of SSC attributes. The proposed sampling protocol is described as a step-by-step procedure and could be applied in sampling designs relevant to SSC monitoring in row-cropped fields.

Chapitre 2.3; International audience

Correspondance: voltz@ensam.inra.fr; International audience; Observation and interpretation of spatial soil surface moisture patterns are fundamental to spatially distributed modelling of runoff generation, soil evaporation, and plant transpiration. Compared to natural basins, man-made managements in farmed basins, such as field limits, agricultural practices and the networks of ditches, Lead to great spatial heterogeneity in hydrological processes at the catchment scale. The aim of this study was to identify the factors controlling the spatio-temporal variability of the surface soil moisture in the farmed Mediterranean catchment of Roujan (0.91 km(2)) located in southern France. Intensive measurements of soil moisture patterns were recorded during two drying sequences, respectively, in dry and wet seasons. Results show that the soil surface characteristics (SSC), which result in part from the agricultural practices such as soil tillage, chemical weed control or grass covering, are the main factors controlling the spatio-temporal distribution of the soil surface moisture during both the wet and dry drying sequences. However, in this study, none of the local factors such as the soil insolation (sunlight reaching soil surface through the plant canopy if there is one), the slope, the aspect and the soil texture is correlated to the soil moisture spatial variability. Only local factors control the spatio-temporal variability of soil surface moisture because agricultural operations like tillage influence greatly the local surface runoff by altering soil hydrologic properties. Also, the ditch networks influence the water transfer from the fields to the catchment outlet by routing runoff directly to the catchment outlet without modifying the soil surface moistures of downslope fields. Consequently, in farmed catchments the agricultural managements and practices strongly modify the spatio-temporal soil moisture distribution and must be taken into account in the understanding and in the modelling of hydrological processes

Most runoff simulation and infiltration models have been developed at the global-catchment scale or the local-soil column scale. Few models have been specifically developed at the scale of agricultural fields and there are no guidelines to help modellers choose an adequate model to simulate overland flow and hence analyse the impact of different soil management practices on flood generation. A comparison is undertaken to select and calibrate models that simulate Hortonian overland flow at the field or small plot scale. The proposed methodology couples a runoff production model to a unit hydrograph transfer function. Four different models were tested: Philip, Morel-Seytoux, Horton and SCS. These models differ by their mathematical structure and the parameters to be calibrated while input hydrologic data are the same site data: rainfall/runoff and initial water content. The models are calibrated on 14 events and validated on 14 others. The results of both the calibration and validation phases are compared on the basis of their performance with regards to six objective criteria, three global criteria and three relative criteria representing volume, peakflow, and the root mean square error. The first type of criteria gives more weight to strong events whereas the second considers all events to be of equal weight. The results show that the calibrated parameter values are dependent on the type of objective criteria used. Furthermore, when analysing the performance of the six objective criteria used, it can be seen that the global volume, global RMSE and relative peakflow criteria give the best compromise between bias and precision. Within the selected modelling framework, Morel-Seytoux's model performed better than the other three and the SCS gave the worst results. Horton's model showed to be more consistent in overall performance than Philip's model. Results also highlight problems related to the simulation of low flow events and intermittent rainfall events.

The aim of this study was to assess the toxicity of runoff waters in an agricultural multipollution context through an in-depth assessment of copper bioavailability and toxicity. Runoff waters were screened for major ions, metals and diuron. The potential environmental impact of these runoff waters was evaluated using the conventional 72-h growth inhibition test with the green alga Pseudokirchneriella subcapitata. The results suggested that the toxicity detected in the calcareous vineyard field was due to the presence of diuron, whereas the non-calcareous runoff waters were non-toxic. Chemical speciation modelling by MINEQL revealed that most of the copper present in the non-toxic natural runoff waters was complexed by organic matter. These samples were spiked with copper, and then the toxicity and the electrochemically bioavailable copper fraction were measured. Differential pulse anodic stripping voltammetry (DPASV) was used to detect labile complexes and free copper. This combined approach highlighted the presence of some labile copper complexes in samples reaching the EC10—these could have contributed to the copper toxicity. D 2005 Elsevier B.V. All rights reserved.

Field limits, tillage practices and ditch networks constitute man-made hydrological discontinuities in farmed catchments, and are expected to influence hydrological response during flood events. The purpose of this study is to assess the role of human impact, especially the existence of tillage practices and ditch network, on flood events. The study area is the farmed catchment of Roujan (0.91 km 2 ) located in Southern France for which a spatially distributed hydrological model, MHYDAS, was developed and tested. The model considers the catchment as a series of interconnected field parts linked to the ditch network. Descriptions are provided for the main model procedures: computation of Hortonian excess rainfall on fields using the Green and Ampt approach, conversion of excess rainfall to surface runoff, interaction between ditch network and groundwater using a simple Darcian model and flood routing through the ditch network using the diffusive wave model. To analyse the role of both tillage practices and the ditch network, two sets of sensitivity analysis of the model were applied. The first set studied the role of tillage practices by comparing the actual spatial distribution of tillage practices on the catchment with three hypothetical scenarios. The second set studied the role of the ditch network by comparing the actual man-made ditch network with a hypothetical drainage network automatically extracted from a digital elevation model. Results show the importance of the role of tillage and the ditch network on the form of the hydrograph, the lag time, the runoff volume and the peak discharge. This technique could also be applied to study the impact of land use change on the hydrological behaviour of the catchment.

International audience; In the Mediterranean wine producing region, soil surface remote sensing for hydrological modelling is hampered by the discontinuous canopy of the perennial vine stocks. Very high spatial resolution imagery allows the recording of ground resolution elements that comprise only soil in between vine stocks. To enable the comparison of the resolution element's spectral properties with known bi-directional properties of different types of soil surface, pixels representing the pure and sunlit soil surface have first to be extracted from the imagery. Multi-scale analysis proved to provide a robust and satisfactory segmentation method. Conditional per-field adjustment of a multi-scale image mask allows the extraction of the image segments corresponding to the sunlit soil surface, whatever its state or composition. Moreover, the vegetation cover can be estimated with an error of at most 10% if the geometric configuration is taken into account.; Extraction de la surface du sol éclairée dans des images de télédétection de cultures pérennes et discontinues ; le cas du vignoble méditerranéen. En région méditerranéenne viticole, l'estimation des états de surface du sol par télédétection est perturbée par la présence d'un couvert végétal pérenne et discontinu. L'imagerie à très haute résolution spatiale permet d'obtenir des pixels constitués uniquement de la surface du sol. Pour comparer la radiométrie de ces pixels aux propriétés bidirectionnelles connues de différents types de surface, on cherche à isoler les pixels correspondant à la surface du sol éclairée. L'analyse multi-échelle fournit une approche de segmentation robuste et satisfaisante. L'ajustement conditionnel par parcelle d'un masque obtenu par analyse multi-échelle permet d'accéder à la surface du sol éclairée, indépendamment de son état ou de sa composition. De plus on peut estimer le couvert végétal de la vigne avec une marge d'erreur de 10 % si la configuration géométrique est prise en compte.

Little work has up to now been done on the mechanisms of storm flow generation in Mediterranean cultivated environments. The present work analysed such mechanisms by natural chemical tracing in a small Mediterranean wine-growing catchment (0.91 km 2 ): Roujan, Herault, France. Two autumn runoff events with very different characteristics were studied. The first, a minor one (specific peak flow=28 l/s/km 2 ), was used to evaluate the sensitivity of the environment to low intensity rainfall. The second was significantly larger (specific peak flow=944 l/s/km 2 ) and was used to analyse the response of the catchment to heavy downpours. Tracer concentrations at the catchment outlet, for the groundwater of two distinct geomorphological units (depression and plateau) and in an experimental plot are presented. A mixing model involving three reservoirs and two tracers (chloride and nitrate) is then used to estimate the contributions of the three main storm flow components: (a) the pre-event water deriving from the depression groundwater; (b) the event water of the precipitations; and (c) the pre-event water of the plateau groundwater. The event water end member basically corresponds to infiltration-excess overland flow plus direct precipitation on saturated areas. The imprecision of the calculations was estimated by the Monte Carlo method. During both runoff events, there was little variation in the rate at which the stream was fed by pre-existing water deriving from the groundwater, although the water tables rose rapidly. Overland flow dominated in the rapid storm flow. Its contribution varied between 12 and 82% according to the importance of the event. When the water level rose, particularly in the case of the heavy runoff event, the overland flows concentrated in the man-made network of ditches running down towards the main ditch. This wave of overland flow spread, expelling the pre-event water into the ditches located downstream, which were initially fed by the groundwater. With the rapid rise in the level of water in the ditches, ditchwater infiltrated into the groundwater, and the latter then ceased to contribute to the flow in the ditch network.