Search

International audience; Maize production is essential for global food security and represents a major supply in several value chains. However, the projected effects of climate change are likely to decrease drastically water availability for crops in many regions, affecting yield. AgriVoltaics (AV) systems are an innovative solution that may improve maize resilience in water-scarce regions mainly by protecting plants from excessive radiation and by reducing irrigation needs. However, shade from panels may also affect crop development and production. This study addresses the interplay between radiation transmission, crop development and irrigation needs of maize cropping in field conditions, by the description of crop development dynamics, distinguishing between fixed and dynamic panels. We showed that maize crop responded to both independent and combined stresses (shade and water deficit), with a significant decrease in leaf area index, total dry matter and grain yield. Concerning water use, we showed the potential of AV to reduce irrigation inputs (by up to 19-47% compared to unshaded plots) via reduced soil water depletion and reference evapotranspiration. The crop development was impacted by shade by increasing phyllochron and causing a generalized delay in phenology. At a finer temporal scale, we concluded that maize leaves react to shade by reducing stomatal conductance, net assimilation of CO2 and leaf temperature in a correlated way to radiation, opening the possibility to use this behavior to optimize water use and shading strategies. The spatial heterogeneities of radiation in fixed AV systems, compared to dynamic AV systems, were identified as a second-order effect at the plot level on leaf area index and phyllochron, compared to the effect of radiation reduction. Moreover, dynamic AV showed their ability to reduce the spatial heterogeneities in soil water depletion, showing the importance of controlled shade strategies in AV systems concerning water use.

International audience

The objectives of this paper are twofold: (i) to calibrate and validate the PILOTEN crop model for tomato (Solanumlycopersicum L.), and (ii) to use the model to examine the response of tomato yield to different irrigation and fertilization schedulingscenarios under varying climatic conditions in order to draw up management recommendations for smallholder farms. Thecalibration and validation were performed by using data collected from field experiments carried out in the north-east of SouthAfrica, on shallow Ferrosol and Luvisol soils. The scenarios dealt with: (i) the initial soil-water availability (fulfilling or not thesoil-water reserve at planting); (ii) the irrigation dose frequency (high versus low irrigation doses ranging from 10 to 40 mm perapplication); (iii) the fertilization dose frequency (full versus split fertilization application). The results indicated that only thefertilization scheduling may significantly affect tomato yields, while irrigation scheduling has a minor role. However, comparedto the local official irrigation guidelines, it is recommended to slightly increase the irrigation doses during the first 4 weeks ofgrowth for tomato grown between early September and mid-December, and to slightly decrease the doses in the second partof the growing cycles starting in February and May. It is also suggested to avoid applying the entire fertilizer dose after seedlingplanting for tomato grown between mid-February and late June; Ce travail a un double objectif: (i) calibrer et valider le modèle Piloten pour la tomate (Solanum lycopersicum L.) et (ii) utiliserle modèle pour analyser la réponse du rendement de la tomate à différents scénarios d’irrigation et de fertilisation, sousdifférentes conditions climatiques, afin de proposer des recommandations de conduites culturales aux petites exploitations.Le calage et la validation du modèle ont été effectués en utilisant des données recueillies par des expérimentations menées dansle nord-est de l’Afrique du Sud, sur des Ferrosols et des Luvisols peu profonds. Les scénarios ont traité: (i) de la disponibilitéinitiale en eau du sol (réserve en eau du sol reconstituée ou non à la plantation); (ii) du couple dose-fréquence d’irrigation(doses d’irrigation élevées à faibles allant de 10 à 40 mm par application); (iii) du couple dose-fréquence de fertilisation.Les résultats ont indiqué que seul le calendrier de la fertilisation peut affecter de manière significative les rendements de latomate, tandis que le calendrier d’irrigation joue un rôle mineur. En comparaison avec les recommandations locales de lagestion de l’irrigation et de la fertilisation de la tomate, il est recommandé d’augmenter légèrement les doses d’irrigation pendantles 4 premières semaines de croissance pour la tomate cultivée entre le début septembre et la mi-décembre, et de diminuerlégèrement les doses dans la deuxième partie du cycle de croissance commençant en février et en mai. Il est également suggéréd’éviter l’application de la totalité de la dose d’engrais après les semis de tomate cultivée entre mi-février et fin-juin

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI [ADD1_IRSTEA]Gestion intégrée de la ressource et des infrastructures; International audience; Agriculture in developing countries will have to transform and increase production by an estimated 70% in order to meet demands by 2050. Although well-managed commercial farms offer little manoeuvring space for increasing agricultural water productivity, smallholder farms usually operate at low input costs and therefore provide ample opportunities to reduce the potential yield gap through agricultural intensification. The aim of this paper is to analyse and compare yields and water productivities obtained in field and modelling experiments in Ethiopia (maize, garlic, onion), South Africa (tomato) and Tunisia (tomato, potato, wheat). Innovative agricultural practices were introduced on smallholder farms: irrigation scheduling and NPS Zn fertilization in Ethiopia; high-yielding cultivar, drip irrigation, mulching and organic amendments in South Africa; and crop water modelling in Tunisia. In general, crop yields increased up to eight times with innovative practices compared to current conventional farming practices. Crop water productivities were fairly stable within the same experiments, but increased with innovations, indicating that intensive farming can be more environmentally sustainable than conventional farming. Intensive farming systems in a resource-rich environment (high radiation levels, relatively fertile, deep and well-drained soils), combined with technology transfer and capacity building could be seen as viable strategies to secure food for smallholders and communities in African rural areas, as well as to improve water utilization in water-scarce catchments.

SI; International audience; Boosting the productivity of smallholder farming systems continues to be a major need in Africa. Challenges relating to how to improve irrigation are multi‐factor and multisectoral, and they involve a broad range of actors who must interact to reach decisions collectively. We provide a systematic reflection on findings from the research project EAU4Food, which adopted a transdisciplinary approach to irrigation for food security research in five case studies in Ethiopia, Mali, Mozambique, South Africa and Tunisia. The EAU4Food experiences emphasize that actual innovation at irrigated smallholder farm level remains limited without sufficient improvement of the enabling environment and taking note of the wider political economy environment. Most project partners felt at the end of the project that the transdisciplinary approach has indeed enriched the research process by providing different and multiple insights from actors outside the academic field. Local capacity to facilitate transdisciplinary research and engagement with practitioners was developed and could support the continuation and scaling up of the approach. Future projects may benefit from a longer time frame to allow for deeper exchange of lessons learned among different stakeholders and a dedicated effort to analyse possible improvements of the enabling environment from the beginning of the research process. © 2020 The Authors. Irrigation and Drainage published by John Wiley & Sons Ltd on behalf of International Commission for Irrigation and Drainage

International audience; Surface soil moisture (SSM) is a key variable for many environmental studies, including hydrology and agriculture. Synthetic aperture radar (SAR) data in the C-band are widely used nowadays to estimate SSM since the Sentinel-1 provides free-of-charge C-band SAR images at high spatial resolution with high revisit time, whereas the use of L-band is limited due to the low data availability. In this context, the main objective of this paper is to develop an operational approach for SSM estimation that mainly uses data in the C-band (Sentinel-1) with L-bands (ALOS/PALSAR) as additional data to improve SSM estimation accuracy. The approach is based on the use of the artificial neural networks (NNs) technique to firstly derive the soil roughness (Hrms) from the L-band (HH polarization) to then consider the L-band-derived Hrms and C-band SAR data (VV and VH polarizations) in the input vectors of NNs for SSM estimation. Thus, the Hrms estimated from the L-band at a given date is assumed to be constant for a given times series of C-band images. The NNs were trained and validated using synthetic and real databases. The results showed that the use of the L-band-derived Hrms in the input vector of NN in addition to C-band SAR data improved SSM estimation by decreasing the error (bias and RMSE), mainly for SSM values lower than 15 vol.% and regardless of Hrms values. Based on the synthetic database, the NNs that neglect the Hrms underestimate and overestimate the SSM (bias ranges between −8.0 and 4.0 vol.%) for Hrms values lower and higher than 1.5 cm, respectively. For Hrms 2.0 cm and SSM between 8 to 22 vol.%, the accuracy on the SSM estimation improved and the overestimation decreased by 2.2 vol.% (from 4.5 to 2.3 vol.%). From the real database, the use of Hrms estimated from the L-band brought a significant improvement of the SSM estimation accuracy. For in situ SSM less than 15 vol.%, the RMSE decreased by 1.5 and 2.2 vol.% and the bias by 1.2 and 2.6 vol.%, for Hrms values lower and higher than 1.5 cm, respectively.

A possible way to save agricultural water is to improve global irrigation efficiency, defined as the ratio between the irrigation water volume used for crop transpiration and the total irrigation water volume. Higher irrigation efficiency leads to higher irrigation productivity ("more crop per drop") and profitability in relation to irrigation water costs: higher financial expectations for the same water and energy costs or identical financial expectations for lesser water and energy costs.Improvements of irrigation efficiency may be sought either from better performing material (contextual relevance, technical quality or durability) or from optimized irrigation strategies with multiple levers of action (dates, doses, scheduling or trigger criteria, leaning on weather forecasts or not) and expected constraints (availability, quota, prefecture decrees).It is even possible to handle these two issues at once by(i) evaluating the irrigation water losses attributable to material (e.g. accidental pipe rupture or unavoidable intrinsic losses when using rainguns, spatial heterogeneity of water delivery),(ii) evaluating the losses due to inadequate irrigation strategies (drainage or evaporation of irrigation water, excessive irrigation water storage in soils, losses due to wind drift)(iii) gathering all losses in a "cascade scheme" that organizes them in phenomenological and chronological manner, somehow assuming successive losses "from the canal to plant roots",(iv) exploring numerous irrigation scenarios that would allow reducing losses, improving efficiency and finally finding the minimal irrigation water amount required to reach the target agricultural yield (or fulfil a typical set of contradictory constraints, e.g. irrigation quota vs. objectives in crop yield profitability and possibly no drainage) This was the scope of the recent development of the "Efficiency" module of the Optirrig model built for the generation, analysis and optimisation of crop irrigation scenarios.

[Notes_IRSTEA]105682 [Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI [ADD1_IRSTEA]Gestion intégrée de la ressource et des infrastructures; International audience; In order to reduce irrigation water withdrawal, the European Commission provides grants to farmers for investments in irrigation techniques that save water. However, little is known about the real extent of water savings at plot scale resulting from change in irrigation application equipment or the adoption of scheduling devices in a given agro-pedo-climatic context. The aim of this study was to evaluate irrigation water savings achievable by switching from sprinkler to localized irrigation system or by using soil hydric status probes. For this purpose, we used the French metropolitan context for our case study and compiled all available studies conducted over the past 30 years. A total of 93 records were collected from experimental field trials representative of a wide range of pedo-climatic conditions (25 different sites) and crops (field crops, fruit and vegetable production). Each record represents the water consumption of two different irrigation systems (sprinkler system vs localized system) or two scheduling systems (without soil probe vs with soil probe) at plot scale and is used to assess the water saving made when comparing the most water consuming system to the least consuming one. Results show that water savings are highly variable, ranging from 0% to more than 75% of the initial consumption. They originate in both irrigation technology and management. Their key features are the following. (1) Water savings made with localized systems, when compared to sprinkler irrigation, significantly decrease when the hydric deficit of the cropping season increases and when soil water holding capacity rises. Moreover, they tend to be higher when irrigation is managed with soil probes. (2) Water savings obtained with irrigation scheduling using soil probes (when compared to scheduling without probes) seem, on the contrary, not to be influenced by hydric deficit and soil water holding capacity. The type of soil probe (tensiometric or capacitive) used has no influence on the water savings obtained. (3) Water savings achieved with either localized systems or soil probes result in increased irrigation water productivity and are only marginally influenced by crop type. This frame of reference for irrigation water savings can guide public policies encouraging and financially supporting the implementation of water saving systems, not only on the subject of irrigation devices, but also on irrigation scheduling tools.

Floods are the highest-impact natural disasters. In agricultural basins, anthropogenic features are significant factors in controlling flood and erosion. A hydrological-hydraulic-erosion diagnosis is necessary in order to choose the most relevant action zones and to make recommendations for alternative land uses and cultivation practices in order to control and reduce floods and erosion. This chapter first aims to provide an overview of the flow processes represented in the various possible choices of model structure and refinement. It then focuses on the impact of the spatial distribution and temporal variation of hydrological soil properties in farmed basins, representing their effects on the modelled water and sediment flows. Research challenges and leads are then tackled, trying to identify the conditions in which sufficient adequacy exists between site data and modelling strategies.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI [ADD1_IRSTEA]Gestion intégrée de la ressource et des infrastructures [ADD2_IRSTEA]Sûreté alimentaire; International audience; In this paper we present PILOTE-N, a crop model devoted to the calculation of crop production from the joint water and nitrogen soil status effects. It is the extension of PILOTE, for contexts in which water might not be the sole limiting factor for crop yield, but the same model structure and credo of parsimonious parameterisation have been kept, assuming simplified descriptions of the physical processes at play. One original aspect of PILOTE-N is the calculation of Leaf Area Index (LAI) and cumulative nitrogen plant demand from similar logistic-type functions. LAI is controlled by specific temperature sums, shape parameters and the occurrence of water and nitrogen stresses, while the time average of LAI values over critical phenological periods also affects the predicted harvest index and crop yield. As specific plant parameters are known from PILOTE, calibration was devoted to nitrogen parameters. Which governing the daily-averaged mineralization rate in a first step, then the two shape parameters of the potential nitrogen plant demand (from the dilution curve) in a second step and, at last, which allowing the link between nitrogen uptake and nitrogen of the soil. Model performance was evaluated using multiple initial soil conditions, irrigation and fertilization strategies for corn, durum wheat and sorghum, over a climatic series of 14 years, at the experimental plot of Lavalette (Montpellier, South-East of France), hence in a Mediterranean context characterized by severe, water stresses during summer, typically exceeding nitrogen stresses. Crop yield as well as the dynamics of nitrogen budget were correctly simulated (R2 > 0.94 for grain yield, total dry matter and nitrogen in plant). The robustness of such a simple and easy-to-calibrate tool is expected to facilitate its use and implementation in other agro-pedoclimatic contexts, to decipher the effect of abiotic stresses and improve irrigation and fertilization scenarios when included in dedicated tools. © 2018 Elsevier B.V.

International audience; The installation of tilting-angle solar panels above agricultural plots provides renewable energy and means of action to dampen some of the effects and hazards of climate change. When the panels are properly operated, their drop shadow reduces water consumption by the plants, as a consequence of alternating shade and sun bands with a short-term impact on the stomatal conductance and a global decrease of gas exchanges. This urged the development of a new model for crop growth and water budget, adapted here from existing literature to handle such transient conditions, characterized by short-term (infra-day) fluctuations. The main difficulty was to combine short-term fluctuations in the climatic forcings (radiation interception and rain redistribution by the panels) and long-term agronomic evaluation, hence the coexistence of several calculation time steps in model structure. All field experiments were conducted on purpose in the agrivoltaic plot of Lavalette (Montpellier, France). Specific adaptations consisted in describing the stomatal behavior of the plants for fluctuating solar radiations and varied water status, aiming at improving both the piloting of the solar panels and water management, i.e. the choice of irrigation amounts. Model simulations have been able to reproduce the expected benefits from agrivoltaic installations, for example showing that it is possible to improve land use efficiency and water productivity at once, by reducing irrigation amounts by 20%, when tolerating a decrease of 10% in yield or, alternatively, a slight extension of the cropping cycle. Agrivoltaism appears a solution for the future when facing climate change and the food and energy challenges, typically in the rural areas and the developing countries and especially if the procedure presented here proves relevant for other crops and contexts.

In this study, we assessed the sensitivity analysis of the LandSoil model (Ciampalini et al., 2012), a landscape evolution model simulating soil erosion at fine spatial resolution (1–10 m) and midterm temporal scales (10–100 years). Our aim was to determine the most significant variables in the model and identify the crucial parameter combinations through their interaction. The methodology we adopted consists of a local sensitivity analysis, also termed “one-factor-at-a-time,” calculating soil loss on a topographical entity—a virtual hillslope—inspired by the virtual catchment framework proposed by Cheviron et al. (2010, 2011). Afterward, a multicategory analysis was set to control the effects and hierarchy of hydrological factors, soil-terrain parameters, and the main topographical configurations of the catchment (i.e., hillslope profile). The results show the classical predominant relevance of rainfall amount (no rain means no water erosion). In a second level, the results highlight the interactions between parameters such as soil roughness, soil crusting, and soil cover responsible for sediment production function, and the influence of the topography describing water flow at the catchment scale.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI [ADD1_IRSTEA]Systèmes aquatiques soumis à des pressions multiples [Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI [ADD1_IRSTEA]Systèmes aquatiques soumis à des pressions multiples; International audience; High sensitivity temperature sensors (0.001 K sensitivity Pt100 thermistors), positioned at intervals of a few centimetres along a vertical soil profile, allow temperature measurements to be made which are sensitive to water flux through the soil. The development of high data storage capabilities now makes it possible to carry out in situ temperature recordings over long periods of time. By directly applying numerical models of convective and conductive heat transfer to experimental data recorded as a function of depth and time, it is possible to calculate Darcy's velocity from the convection transfer term, thus allowing water infiltration/exfiltration through the soil to be determined as a function of time between fixed depths. In the present study we consider temperature data recorded at the Boissy-le-Châtel (Seine et Marne, France) experimental station between April 16th, 2009 and March 8th, 2010, at six different depths and 10-min time intervals. We make use of two numerical finite element models to solve the conduction/convection heat transfer equation and compare their merits. These two models allow us to calculate the corresponding convective flux rate every day using a group of three sensors. The comparison of the two series of calculated values centred at 24 cm shows reliable results for periods longer than 8 days. These results are transformed in infiltration/exfiltration value after determining the soil volumetric heat capacity. The comparison with the rainfall and evaporation data for periods of ten days shows a close accordance with the behaviour of the system governed by rainfall evaporation rate during winter and spring.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI; The objective of this review is to remedy the lack of knowledge about the expected relationship between the irrigation techniques (flood irrigation, sprinkler irrigation, surface and subsurface drip irrigation) and the nitrogen transformations (fixation, mineralization, immobilization and nitrification) or fluxes (denitrification, runoff, volatilization and lixiviation) in agricultural contexts. This study investigates thus the various controls on the nitrogen cycle and budget, either site-specific (soil C/N ratio, pH, salinity, texture and temperature) or corresponding to deliberate strategies in water and nitrogen management. The comprehensive view gained from the gathered literature elements identifies the local variations (in space and time) of soil water content profile as both the strongest control and the control most directly related to the irrigation technique and to the practitioner’s decisions. In the overall picture, flood irrigation and sprinkler irrigation are the techniques in which most transformations or fluxes may be enhanced or reduced, also with the risks associated with non-optimal practices. By contrast, subsurface drip irrigation seems the technique with the least unwanted impacts. Besides the academic aspects (bringing together scattered literature elements) an outcome of this review is thus to facilitate decision-making regarding the choice and/or use of irrigation-fertilization techniques and strategies (doses and scheduling) for given agro-pedoclimatic contexts, also for combined agricultural production, economic and site preservation objectives.

We use high precision Pt 100 sensors and apply numerical models of convective and conductive heat transfer to temperature experimental data recorded versus depth and time along a vertical soil profile. After having determined the thermal diffusivity and the convective flux rate it is possible to calculate the water content and the Darcy’s velocity using empirical soil properties models. The results obtained at Boissy-le Châtel site between April 16, 2009 and March 8 2010 are presented here. The ten day period results are reliable in accordance with local rain and potential evapotranspiration data.

This paper discusses an innovative framework to use crop models which combines sensitivity analysis, uncertainty analysis and constrained optimisation runs for irrigation optimisation purposes, facing competing constraints on several agricultural variables (e.g. crop yield, total irrigation amount, financial expectations). For simplicity, this ex-post optimisation relies on direct calculations only, exploiting the dispersions on the target variables. The screening of the parameter space for sensitivity analysis yields a reference dispersion which is expectedly reduced by reducing the uncertainties in the sensitive parameters and/or climatic forcings. Additional dispersions are calculated to evaluate if the management controls on irrigation strategies (amounts, triggers, periods) are more influential on model predictions than the remaining uncertainties on the soil, plant, irrigation and climatic inputs, eventually allowing optimisation. As a case study, the Optirrig model is used. A discussion proposes future ways to convert diagnostics into real-time near-optimal decision rules, for example through learning algorithms. This ex-post optimisation of irrigation strategies involves direct calculations only.The Optirrig model (Irstea Montpellier, France) is chosen for application.Both the feasibility and the achievement of irrigation optimisation are handled.Dispersion is reduced through sensitivity, uncertainty then optimisation runs.Perspectives are discussed to convert diagnostics into real-time decision rules.

The aim of this study is to assess the effects and interests of integrating remote-sensing-derived parameters (LAI, harvest and irrigation dates) in a crop model (PILOTE) that simulates vegetation growth for hay crops. The target variable is the prediction of Total Dry Matter (TDM) production in each of the three growth cycles. Two scenarios are employed to process the available remotely sensed LAI values, predicting TDM values when forcing in PILOTE either the initial and maximal optical LAI-values, or the initial, maximal and daily interpolated LAI values. The predictions show low deviations compared with the in situ TDM values (RMSE of 0.44 t/ha, MAPE of 23%). The feasibility of using harvest dates that are derived from optical data is examined by feeding the model with randomly perturbed harvest dates. The magnitude of the perturbations is equal to the revisit times of the current optical sensors. Optical images with revisit times lower than 16 days are adequate to feed PILOTE with remotely sensed harvest dates. Emphasis is placed on the forcing of “uncertain” irrigation dates, derived from Synthetic Aperture Radar images either replacing all true irrigation dates by randomly perturbed dates (using 3-day perturbation magnitudes) or hypothesizing one or several irrigations are “missed” (undetected). The results show negligible errors for the TDM predictions when noisy irrigation dates are used (RMSE of 0.17 t/ha and MAPE of 4.2%). Disregarding one or two irrigations within a period with important rainfalls does not induce significant errors for the predicted TDM values; however, it causes noticeable underestimations in drier periods (maximum of 1.55 t/ha, reference TDM of 3.43 t/ha). This study enables the identification of a series of conditions in which remote-sensing-derived parameters are suitable to feed the PILOTE model without endangering the reliability of its predictions.

This review paper investigates the determinants of modelling choices, for numerous applications of 1-D free-surface flow and morphodynamic equations in hydrology and hydraulics, across multiple spatiotemporal scales. We aim to characterize each case study by its signature composed of model refinement (Navier–Stokes: NS; Reynolds-averaged Navier–Stokes: RANS; Saint-Venant: SV; or approximations to Saint-Venant: ASV), spatiotemporal scales and subscales (domain length: L from 1 cm to 1000 km; temporal scale: T from 1 s to 1 year; flow depth: H from 1 mm to 10 m; spatial step for modelling: δL; temporal step: δT), flow typology (Overland: O; High gradient: Hg; Bedforms: B; Fluvial: F), and dimensionless numbers (dimensionless time period T*, Reynolds number Re, Froude number Fr, slope S, inundation ratio Λz, Shields number θ). The determinants of modelling choices are therefore sought in the interplay between flow characteristics and cross-scale and scale-independent views. The influence of spatiotemporal scales on modelling choices is first quantified through the expected correlation between increasing scales and decreasing model refinements (though modelling objectives also show through the chosen spatial and temporal subscales). Then flow typology appears a secondary but important determinant in the choice of model refinement. This finding is confirmed by the discriminating values of several dimensionless numbers, which prove preferential associations between model refinements and flow typologies. This review is intended to help modellers in positioning their choices with respect to the most frequent practices, within a generic, normative procedure possibly enriched by the community for a larger, comprehensive and updated image of modelling strategies.

Knowledge of sediment exports from continental areas is essential for estimating denudation rates and biogeochemical cycles. However, the estimation of current sediment fluxes to the sea is often limited by the availability and quality of sediment discharge data. This study aims to quantify the relative contributions of French rivers to the sediment discharge to the ocean. Sediment fluxes were assessed using the French river quality database, which is characterized by a low temporal resolution but long-term measurement periods. An improved rating curve approach (IRCA) using daily discharge data, which allows the estimation of mean annual sediment loads from infrequent sediment concentration data, was used to calculate sediment fluxes. The resulting mean annual sediment loads show that French rivers export c. 16.21 Mt yr-1 of sediments to the sea. Among the 88 defined French rivers flowing to the sea, the four largest basins (Loire, Rhone, Garonne and Seine) export 13.2 Mt yr-1, which corresponds to 81.3% of total exports. No relationship was found between the mass of exported sediment and the size of the drainage basins. This is due to the variety of river basin typologies among these rivers, including lowland rivers in temperate climates, such as the Seine on the one hand and rivers draining mountainous areas in Alpine/Mediterranean areas on the other hand, such as the Rhone. The latter contributes 60% to the total sediment export for France while its drainage area is only 19% of the total area considered. Differences between the river basins considered are also shown by temporal indicators describing the duration of the exports, which may be linked with sediment production processes over drained areas.

Among the various methods used to determine vertical water seepage in unsaturated soils, thermal convection presents significant advantages: temperature measurements are simple to perform and record, and a wide range of time scales can be considered. The authors analysed the data recorded by the meteorological stations of the Seine river basin, at three different depths: 20, 50 and 100 cm. As the measurement sensitivity was limited to 0.1 K, long series of data needed to be stacked in order to obtain sufficient precision to quantify the convective component of heat transfer, in a predominantly conductive context. For the period from 1984 to 2001, it was possible to determine the average recharge at each station, and the recharge variation between groups of three-year periods. By interpolating these data over the whole basin, a global assessment has been made and compared to the exported flow rate at the river mouth: the resulting value of 94 mm yr −1 lies between the lowest annual rate, 52 mm yr −1 , and the mean total exported value of 252 mm yr −1 .

The aim of this study was to develop an inversion approach to estimate surface soil moisture from X-band SAR data over irrigated grassland areas. This approach simulates a coupling scenario between Synthetic Aperture Radar (SAR) and optical images through the Water Cloud Model (WCM). A time series of SAR (TerraSAR-X and COSMO-SkyMed) and optical (SPOT 4/5 and LANDSAT 7/8) images were acquired over an irrigated grassland region in southeastern France. An inversion technique based on multi-layer perceptron neural networks (NNs) was used to invert the Water Cloud Model (WCM) for soil moisture estimation. Three inversion configurations based on SAR and optical images were defined: (1) HH polarization, (2) HV polarization, and (3) both HH and HV polarizations, all with one vegetation descriptor derived from optical data. The investigated vegetation descriptors were the Normalized Difference Vegetation Index "NDVI", Leaf Area Index "LAI", Fraction of Absorbed Photosynthetically Active Radiation "FAPAR", and the Fractional vegetation COVER "FCOVER". These vegetation descriptors were derived from optical images. For the three inversion configurations, the NNs were trained and validated using a noisy synthetic dataset generated by the WCM for a wide range of soil moisture and vegetation descriptor values. The trained NNs were then validated from a real dataset composed of X-band SAR backscattering coefficients and vegetation descriptor derived from optical images. The use of X-band SAR measurements in HH polarization (in addition to one vegetation descriptor derived from optical images) yields more precise results on soil moisture (My) estimates. In the case of NDVI derived from optical images as the vegetation descriptor, the Root Mean Square Error on My estimates was 3.6 Vol.% for NDVI values between 0.45 and 0.75, and 6.1 Vol.% for NDVI between 0.75 and 0.90. Similar results were obtained regardless of the other vegetation descriptor used.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI; International audience; Macroscopic empirical root water uptake (RWU) models are often used in hydrological studies to predict water dynamics through the soil-plant-atmosphere continuum. RWU in macroscopic models is highly dependent on root density distribution (RDD). Therefore, compensatory uptake mechanisms are being increasingly considered to remedy this weakness. A common formulation of compensatory functions is to relate compensatory uptake rate to the plant water-stress status. This paper examines the efficiency of such compensatory functions to reduce the sensitivity of simulated actual transpiration (Ta), drainage (Draina) and RWU patterns to RDD. The possibility to replace the compensatory RWU functions by an adequate description of RDD is also discussed. The study was based on experimental and numerical analysis of two-dimensional soil-water dynamics of 11 maize plots, irrigated using sprinkler (Asp), subsurface drip (SDI) systems, or rainfed (RF). Soil-water dynamics were simulated using a physically-based soil-water flow model coupled to a macroscopic empirical compensatory RWU model. For each plot, simulation scenarios involved crossing 6 RDD profiles with 6 compensatory levels. RDD was found to be the main factor in the determination of RWU patterns, Ta and Draina rates, with and without the compensatory mechanism. The use of a water-tracking RDD, i.e., higher uptake intensity in expected wetter soil regions, was found a surrogate for compensatory RWU functions in surface-watering simulations (Asp and RF). However, in SDI simulations, a water-tracking RDD should be combined to a high level of compensatory uptake to satisfactorily reproduce real RWU patterns. Our results further suggest that the compensatory RWU process is independent of the plant stress status and should be seen as a response to heterogeneous soil-water distribution. Our results contribute to the identification of optimum parameterization of empirical RWU models as a function of watering methods.

International audience; DBMS is a traditional technology for the storage of business application data. In this chapter, we show that this technology can be of interest in scientific fields. We present a survey of the emergence of the concept of simulation result database. Scientific simulation models become more complex, use more data and produce more outputs. Stochastic models can also be simulated. In this case, numerous simulations are run in order to discover a general trend in results. The replications of simulation increase again the amount of produced data, which makes exploration and analysis difficult. It is also often useful to compare the results obtained in using different model versions, scenarios or assumptions (for example different weather forecasts). In this chapter, we provide several examples of projects of simulation result databases. We show that the database technology can help to manage the large volumes of simulation outputs. We also illustrate this type of projects on environmental databases storing pesticide transfer simulation results. We conclude in highlighting some trends and future works.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI; International audience; Abstract: The spatial distribution of land management practices (LMPs), such as the use of vegetated filters, may have a strong impact on their efficiency in trapping sediments and pollutants. Distributed water erosion models help managers, planners, and policymakers optimize the efficiency of these LMPs regarding their location relative to water and sediment pathways. In this work, the authors analyzed the impact of the spatial distribution of LMPs using an existing distributed model and sensitivity analysis procedures. The distributed model that was used is a distributed single-event physically based water erosion model developed to calculate erosion rates and sediment flow for small (less than 10 km2) agricultural catchments. To measure the impact of the spatial distribution of LMPs, the authors developed a stochastic model that generates LMP locations over the entire catchment. The stochastic model has three input parameters: the density of LMPs, their downslope/upslope location probability, and the probability density function shape controller. Because of its ability to account for the cross effects between parameters, the variance-based Sobol method was used to calculate the sensitivity of the soil loss ratio of a typical Mediterranean agricultural catchment (Roujan, southern France) to the LMP location model parameters. Three measurement points (two subcatchment outlets and the main outlet) were used to examine the spatially distributed effects of the LMP locations. The simulation results indicated that 70% of the variation of the net erosion is explained by variations in LMP density for the main outlet catchment, making LMP density the most sensitive parameter. However, the total Sobol sensitivity indices indicate a strong interaction among the three parameters when the density values are low (few LMPs are applied). Thus, although the density of the LMPs is the most sensitive parameter, their location may influence their global trapping efficiency in (real) cases where few LMPs are applied.

International audience; This chapter presents a state of the art review and research challenges in modeling flood propagation and floodplain inundation. The challenges for flood inundation models are directly linked to the representation of flow processes, to the formulation of theoretical physical laws and to practical considerations. First, we review the various structures of coupled spatially distributed hydrological-hydraulic models and the corresponding spatial representation of flow processes. Second, we present the theoretical basis of 1-D and 2-D Saint-Venant "shallow water" equations with overbank flow, the approximation of Saint-Venant models such as the Diffusive Wave and the Kinematic Wave models and then discuss the domains and limits of applications of each type of models. Practical considerations linked to numerical solution schemes, boundary conditions and model parameterization, calibration, validation and uncertainty analysis were also considered. Finally, the discussion addresses the research challenges for guiding the modeler, according to the principle of parsimony, in seeking the simplest modeling strategy capable of (i) a realistic representation of the physical processes, (ii) matching the performances of more complex models and (iii) providing the right answers for the right reasons.

The temperature variations recorded at several points of a vertical shallow-depth profile are governed both by conductive and convective heat transfers and can be used to calculate the vertical component of the Darcy velocity and thermal diffusivity in the soil. This paper describes such calculations when transient variations over tens of days are considered and tests them using data collected at Voyons (Aube, France). The temperature was recorded during a year and a half period with a 1 h sampling time step at three different depths: 0·2, 0·4 and 0·75 m. By processing the annual variation of temperature, we obtained a value of the Darcy velocity in good agreement with the value of actual/potential evapotranspiration ratio. By processing transient variations, despite the limitation of the calculations due to the lack of sensitivity of the sensors, results obtained at Voyons were in good correlation with tensiometric data. Copyright © 2005 John Wiley & Sons, Ltd.

International audience; The objective of this study was to analyze the sensitivity of radar signal in X-band to irrigated grassland soil conditions. Time series of radar (TerraSAR-X and Cosmo-SkyMed) images were acquired at a high temporal frequency in 2013 over a small agricultural region in South Eastern France. Simultaneously to satellite data acquisitions, ground measurements were conducted during several growing cycles of the grassland in order to monitor evolution in soil and vegetation characteristics. Results show that radar signal is clearly dependent on the soil moisture with a higher sensitivity to soil moisture for biomass lower than 1 kg/m². HH and HV polarizations showed almost similar sensitivity to soil moisture. The penetration depth of the radar wave in X-band was high even for dense and high vegetation: flooded areas were clearly visible on the images with higher detection potential in HH polarization than in HV polarization even for vegetation heights reaching 1 m. These results showed that it is possible to track gravity irrigation and soil moisture variation from SAR X-band images acquired at high spatial resolution and medium incidence angle.

In this work we present a case study of the multi-scale calibration and validation of MHYDAS-Erosion applied to a Mediterranean vineyard. The calibration was performed using expert knowledge in linking physical parameters to land uses with the automatic parameter estimation software PEST. MHYDAS-Erosion was calibrated and validated using spatially distributed observations on total discharge and soil loss. Calibration has been performed within six rainfall events; both hydrological and erosion parameters were calibrated using RMSE, R2 and the modified version of the Nash-Sutcliffe model efficiency criteria. Calibration results indicate there was good agreement between simulated and observed total discharge and total soil loss at the seven observation points (modified Nash-Sutcliffe efficiency (mNSE) ranging between 0.89 and 0.95). Acceptable results were obtained in terms of parameter values, identification of their physical meaning and coherence. However, some limitations were also identified, and could be remedied in more detailed studies involving (i) spatially-distributed rainfall on the catchment, (ii) a description of groundwater exfiltration and (iii) spatially-distributed properties of the ditches over the catchment. Validation results were quite satisfactory for three of the four validation events. The results from this case study suggest that MHYDAS-Erosion may need a specific calibration when applied to another catchment, but once it is calibrated, it could be used for multi-scale soil loss forecasting., Ce travail présente une étude de cas qui traite du calage et de la validation multiéchelle du modèle MHYDAS-Erosion appliqué au bassin versant de Roujan, en France, occupé principalement par des vignobles. Pour le calage, les valeurs des paramètres touchant les cultures ont été estimées à partir des connaissances d’experts et du logiciel d’estimation des paramètres PEST. Des mesures de débits en rivière et de pertes de sol ont ensuite servi à caler le modèle. Six événements ont servi à effectuer le calage du modèle, et quatre ont été utilisés pour valider ce calage. L’intégrité physique des valeurs des paramètres a été respectée. Les résultats du calage indiquent que le modèle est très fidèle aux données observées pour sept points de contrôle distribués sur le bassin versant de Roujan. Les résultats de validation de MHYDAS-Erosion sur le bassin versant de Roujan sont acceptables pour trois des quatre événements choisis. Quelques éléments dans la procédure de calage du modèle pourraient être améliorés au cours d’études subséquentes. Notamment, nous avons constaté que les précipitations varient beaucoup dans le bassin versant de Roujan, et qu’il serait important de considérer cette variation spatiale pour des applications futures du modèle. Aussi, comme les écoulements hypodermiques représentent une grande partie du débit à l’exutoire du bassin versant pour certains événements, il serait préférable de spatialiser les propriétés du réseau de fossés de drainage plutôt que de les considérer uniformes sur l’ensemble du bassin, comme dans la présente étude. L’utilisation du modèle MHYDAS-Erosion sur un autre bassin versant nécessiterait un calage préalable de la valeur des paramètres. Une fois calé, le modèle permettrait d’estimer les pertes de sol pour différents secteurs du bassin versant. Il pourrait aussi servir à tester des scénarios de pratique de gestion bénéfique.

[Departement_IRSTEA]Territoires [TR1_IRSTEA]SYNERGIE [Axe_IRSTEA]TETIS-ATTOS; International audience; The objective of this study was to analyze the sensitivity of radar signals in the X-band in irrigated grassland conditions. The backscattered radar signals were analyzed according to soil moisture and vegetation parameters using linear regression models. A time series of radar (TerraSAR-X and COSMO-SkyMed) and optical (SPOT and LANDSAT) images was acquired at a high temporal frequency in 2013 over a small agricultural region in southeastern France. Ground measurements were conducted simultaneously with the satellite data acquisitions during several grassland growing cycles to monitor the evolution of the soil and vegetation characteristics. The comparison between the Normalized Difference Vegetation Index (NDVI) computed from optical images and the in situ Leaf Area Index (LAI) showed a logarithmic relationship with a greater scattering for the dates corresponding to vegetation well developed before the harvest. The correlation between the NDVI and the vegetation parameters (LAI, vegetation height, biomass, and vegetation water content) was high at the beginning of the growth cycle. This correlation became insensitive at a certain threshold corresponding to high vegetation (LAI ~2.5 m2/m2). Results showed that the radar signal depends on variations in soil moisture, with a higher sensitivity to soil moisture for biomass lower than 1 kg/m². HH and HV polarizations had approximately similar sensitivities to soil moisture. The penetration depth of the radar wave in the X-band was high, even for dense and high vegetation; flooded areas were visible in the images with higher detection potential in HH polarization than in HV polarization, even for vegetation heights reaching 1 m. Lower sensitivity was observed at the X-band between the radar signal and the vegetation parameters with very limited potential of the X-band to monitor grassland growth. These results showed that it is possible to track gravity irrigation and soil moisture variations from SAR X-band images acquired at high spatial resolution (an incidence angle near 30°).

[Departement_IRSTEA]Eaux [TR1_IRSTEA]ARCEAU; This project led to the development of a method in order to assess the surface water contamination potential by pesticides at the small watershed scale. This method considers both spatial and temporal dimensions. The central idea is to aggregate at the catchment scale water and pesticides fluxes simulated at the plot scale, taking into account landscape elements influence (roads, ditches, embankments, buffer zones) on the transferred flux. The goal is to elaborate statistical descriptors of contamination at the outlet. Each catchment component is treated as a typical element, according to its characteristics (soil, slope, land use, agricultural practices ... ), which limits the number of local scale simulations and will facilitate the transposition to larger watersheds (several tens of square kilometers). This approach profits on the knowledge that we have of the processes governing the fate pesticides at plot level or within the landscape features, without having the complexity of a “complete” pesticides transfer model at the catchment scale. The work was carried out on two small catchments where data are numerous but of different quality. They highlighted the need to carefully define the parameters set for the MACRO model, which was used at the plot level. When possible, data should be site specific. A tool making it possible to simulate a large number of simulations was developed. It was applied in order to study the factors which govern pesticides transfers in MACRO for each situation. A method which delineates the catchment in areal or linear homogeneous elements was developed. On top of that, it defines hydrological connections between these elements, which makes it possible to identify which plots are connected or on the contrary isolated from the hydrographic network. Thus a first diagnosis of vulnerability at the catchment scale is possible. Finally, a catchment scale aggregation method of locally simulated water and pesticides fluxes was developed. An analysis method, based on data warehouses, was developed, which makes it possible to perform a thorough analysis of the simulation results. The aggregation method still has to be validated and then applied to a large set of scenarios. Anyway, it will finally allow to qualify more soundly the risk of contamination at the catchment scale, and to identify the most contributive areas or the most problematic configurations.; Ce projet a conduit à l’élaboration d’une méthode d’évaluation du potentiel de contamination des eaux de surface par les produits phytosanitaires à l’échelle du petit bassin versant (BV), qui inclut les dimensions spatiales et temporelles. L’idée centrale est d’agréger à l’échelle du BV les flux de phytosanitaires modélisés à l’échelle de la parcelle, en intégrant l’influence des éléments du paysage (routes, fossés, talus, zones tampons) sur les flux transférés, pour aboutir à des descripteurs statistiques du risque de contamination à l’exutoire. De plus, les éléments qui composent le BV sont chacun assimilés à un élément « type » en fonction de ses caractéristiques (sol, pente, occupation du sol, pratiques culturales ….) ce qui permet de limiter le nombre de simulations à l’échelle locale et de faciliter le passage ultérieur à l’échelle d’un plus grand bassin versant (quelques dizaines de km²). Une telle approche permet de valoriser la connaissance que l’on a des processus régissant le devenir des pesticides à l’échelle de la parcelle ou au sein des éléments du paysage sans avoir la complexité d’un modèle « complet » de transfert de pesticides à l’échelle du bassin versant. Les travaux ont été menés sur deux petits BV où les données sont nombreuses mais de qualité variable. Ils ont permis de mettre en évidence la nécessité de paramétrer soigneusement le modèle MACRO, utilisé à l’échelle de la parcelle, si possible à partir de données acquises localement. Un outil permettant de simuler un grand nombre de simulations a été élaboré, et a permis d’étudier les facteurs déterminant les transferts pour chaque situation dans MACRO. Une méthode de découpage du BV en éléments homogènes, surfaciques ou linéaires, et d’identification des connexions hydrologiques entre ces éléments a été développée. Elle permet d’identifier les parcelles connectées ou au contraire isolées du réseau hydrographique, et donc un premier diagnostic de vulnérabilité à l’échelle du BV. Enfin, une méthode d’agrégation à l’échelle du BV des flux simulés à l’échelle locale a été élaborée, et une méthode d’analyse des résultats, basée sur les entrepôts de données, développée. La méthode d’agrégation reste à valider et appliquer sur un panel de scénarios, mais elle permettra à terme de mieux qualifier le risque de contamination à l’échelle du bassin versant et d’identifier les zones du bassin ou les configurations les plus problématiques.

[Departement_IRSTEA]Territoires [TR1_IRSTEA]SYNERGIE [Axe_IRSTEA]TETIS-ATTOS; International audience; Soil and vegetation biophysical parameter retrieval using synthetic-aperture-radar images requires radiometrically well-calibrated sensors. In this letter, a comparison of signal levels between TerraSAR-X (TSX) and the COSMO-SkyMed (CSK) constellation (CSK1, CSK2, CSK3, and CSK4) was carried out in order to analyze the ability to use jointly all current X-band sensors. The analysis of the X-band signal over forest stands showed a stable signal (variation lower than 1 dB) over time for each of the studied sensors, but a significant difference was observed between the different X-band sensors. Differences between radar signals were higher in HH than in HV polarization. TSX and CSK4 showed similar backscatter signals, with signal level differences of 0.6 dB in HH and 1.4 dB in HV. The CSK3 signal was observed to be lower than those from TSX and CSK4 by about 2.1 dB and 1.5 dB in HH against 3.2 dB and 1.8 dB in HV, respectively. Moreover, CSK2 and CSK1 which showed slightly different backscatter signals (within 1.1 dB in HH and 1.9 dB in HV) had signal levels lower than those obtained from TSX (2.2-3.3 dB in HH and 3.2-5.1 dB in HV for about 29° incidence angle). These results show that it is currently difficult to use jointly the available X-band satellites (CSK and TSX) for estimating the biophysical parameters of soil or vegetation. This is due to the significant difference in the radar signal level between some of the analyzed satellites, which will cause a high overor underestimation of biophysical parameters.

[Departement_IRSTEA]Eaux [TR1_IRSTEA]GEUSI; International audience; This paper addresses feasibility issues in the calculation of fluxes of suspended particulate matter (SPM) from degraded-quality data for flow discharge (Q) and sediment concentration (C) under the additional constraints of infrequent and irregular sediment concentration samplings. A crucial setting of the scope involves establishing the number of data required to counterbalance limitations in the measurement accuracy and frequency of data collection. This study also compares the merits and drawbacks of the classical rating curve (C= aQb) with those of an improved rating curve approach (IRCA: C= aQb+. a1. ?S) in which the correction term is an indicator of the variations in sediment storage, thus relating it to flow dynamics. This alternative formulation remedies the known systematic underestimations in the classical rating curve and correctly resists the degradation in data quality and availability, as shown in a series of problematic though realistic cases. For example, monthly concentration samplings (in average) with a random relative error in the [-30%, +30%] range combined with daily discharge records with a systematic relative error in the [-5%, +5%] range still yield SPM fluxes within factors of 0.60-1.65 of the real value, provided that 15. years of data are available. A shorter 5-day time interval (on average) between samplings lowers the relative error in the SPM fluxes to below 10%, a result directly related to the increased number of Q-C pairs available for fitting. For regional-scale applications, this study may be used to define the data quality level (uncertainty, frequency and/or number) compatible with reliable computation of river sediment fluxes. Provided that at least 200 concentration samplings are available, the use of a sediment rating curve model augmented to account for storage effects fulfils this purpose with satisfactory accuracy under real-life conditions.

Sediment load estimations are often limited by the difficulty in finding an adequate calculation method adapted to infrequent sediment concentration data. In this article, classic calculation methods (averaging estimators, ratio estimators and rating curves) are tested using 61 river basin datasets from the French river quality database (infrequent but long-term datasets). The analyses and results show that averaging methods are inadequate in deriving sediment fluxes from datasets with daily discharge (Q) data and infrequent sediment concentration (C) data. Considering classic power-rating curves, analyses of C(Q) relations for the 61 datasets show significant differences between seasons, flow variations and in function of the antecedent conditions of flows. The subdivision of datasets is thus essential before performing any C(Q) fitting. Moreover, the C = aQb relation only relates instantaneous C and Q values, disregarding the influence of temporal flow dynamics. Rating curve equations, including the dynamic of discharge time series, are established and tested. One equation and approach is selected: an improved rating curve approach (IRCA), which combines a fitting based on the description of sediment stock dynamics with the flow variation subdivision [between rising limbs (RL) and falling limbs (FL) of the hydrograph]. This method was validated with daily datasets from the US Geological Survey (USGS) database. This rating curve equation can be used to quantify sediment exports from infrequent data, for numerous river basins in temperate areas. Copyright © 2010 John Wiley & Sons, Ltd.

[1] Using a previously defined framework, we performed a comparative sensitivity analysis of four very different distributed erosion models (MHYDAS, STREAM, PESERA, and MESALES). We investigated their sensitivities to input fluxes, hydrological submodels, and specific erosion parameters gathered into equivalent slope and equivalent erodibility for each model, thus allowing explicit comparisons between models. Tests involved multiple combinations of rain intensities and runoff conditions for selected screenings of the equivalent parameter space, resorting to one-at-a-time displacements and Latin hypercube samples. Sensitivity to spatial distributions of erosion parameters was calculated as a normalized index of numerical spread of soil loss results, obtained at the outlet of a nine-cell virtual catchment endowed with a fixed flow pattern. Spatially homogeneous or distributed parameterizations yielded responses of comparable magnitudes. Equivalent erodibility was often the key parameter, while sensitivity trends depended on input fluxes and the propensity of soils for runoff, affecting continuous and discrete models in clearly dissimilar ways.

Contact: gumiere@supagro.inra.fr; International audience; The sedimentological connectivity of agricultural catchments may be affected by anthropogenic structures (land management practices) established to reduce sediment exportation from agricultural plots to water streams. Distributed erosion models may in theory provide information about where and how these structures should be installed in catchments to reduce sediment exportation. The interaction between sediment exportation and land management practices is very complex from both theoretical and experimental points of view. Vegetated filters are a widely used land management practice. They interact with water flow, change turbulence conditions, and ultimately affect sediment transport and deposition processes. Experimental results have shown that the efficiency of sediment trapping in vegetated filters is influenced by flow characteristics, sediment size, and vegetation type, as well as by the slope and width of the filter in the streamwise direction. At the catchment scale, the spatial organisation of management practices is crucial for the global sedimentological connectivity. Present-day erosion models propose different approaches to simulate the influence of management practices on soil loss and sediment export for agricultural catchments. Some of them use the Sediment Delivery Ratio (SDR) or P-factor to describe sediment transport from source to sink areas. Others, such as in the TRAVA and VSFMOD, rely on process-based descriptions involving changes in roughness and infiltrability along flow paths to study the effect of management practices. From the literature review conducted herein, we identified the lack of an approach of intermediate complexity, that would be more physically relevant than SDR and P-factor approaches, but simpler and easier to spatialise than TRAVA and VSFMOD-type models

Contact: silvio.gumiere@gmail.com; International audience; In this paper, we present MHYDAS-Erosion, a dynamic and distributed single-storm water erosion model developed as a module of the existing hydrological MHYDAS model. As with many catchment erosion models, MHYDAS-Erosion is able to simulate sediment transport, erosion and deposition by rill and interrill processes. Its originality stems from its capacity to integrate the impact of land management practices (LMP) as key elements controlling the sedimentological connectivity in agricultural catchments. To this end, the water-sediment pathways are first determined by a specific process-oriented procedure defined and controlled by the user, which makes the integration of LMP easier. The LMP dynamic behaviours are then integrated into the model as a time-dependent function of hydrological variables and LMP characteristics. The first version of the model was implemented for vegetative filters and tested using water and sediment discharge measurements at three nested scales of a densely instrumented catchment (Roujan, OMERE Observatory, southern France). The results of discharge and soil loss for simulated rainfall events have been found to acceptably compare with available data. The average R(2) values for water and sediment discharge are 0.82 and 0.83, respectively. The sensitivity of the model to changes in the proportion of LMP was assessed for a single rain event by considering three scenarios of the Roujan catchment management with vegetative filters: 0% (Scenario 1), 18% (Scenario 2, real case) and 100% (Scenario 3). Compared to Scenario 2 (real case), soil losses decreased for Scenario 3 by 65% on the agricultural plot scale, 62% on the sub-catchment scale and 45% at the outlet of the catchment and increased for Scenario 1 by 0% on the plot scale, 26% on the sub-catchment scale and 18% at the outlet of the catchment.

A sensitivity analysis of the HYDRUS-1D model was performed based on scenarios that represented the fate of a nonvolatile pesticide in three soils. We used the latest dual-porosity, transient mobile–immobile (MIM) version of HYDRUS-1D that includes water and solute transfers between mobile and immobile domains. We considered flat bare soils under atmospheric boundary conditions with identical initial pressure head profiles. We targeted instantaneous and mean annual pesticide concentration values at 1-m depth as relevant output variables. Reference values for most of the parameters were obtained from in situ or laboratory measurements. Automation of parameter set generation and input formatting was achieved using SENSAN. We used a one-group-at-a-time sensitivity analysis to isolate and identify the effect of single parameter variations whose ranges were potentially different between topsoil layers (Group 1) and deep layers (Group 2). Sensitivity was analyzed by calculating the time-dependent ratio of variation, i.e., the ratio of relative variation in output concentration to relative variation in parameter input values. As found for other single- or dual- permeability models, transient-MIM HYDRUS-1D was highly sensitive to parameters related to pesticide degradation and sorption, especially the Freundlich exponent associated with nonlinear sorption. Variations in the parameter n of the Mualem–van Genuchten functions had also a very significant impact. Among the MIM parameters, only saturated immobile and mobile water contents were very sensitive parameters, especially those of the topsoil layers. Parameter sensitivities hardly varied among the three soils, but this could be due to their limited contrast.

This report presents the results of the research project ESHEL "Characterization of the risk of contamination of Beauce groundwater by pesticides". The ESHEL project had two main objectives: understanding the mechanisms of pesticide fate in the soils of the Beauce region, and developing a methodological basis for the spatialization of the risk assessment of Beauce groundwater contamination by pesticides. The research considers two spatial scales: the agricultural field, and the hydrological catchment. A detailed pedological study of the selected agricultural field, together with a geophysical prospection of soils, highlighted the strong heterogeneity of the soil cover, in relation to the complexity of the spatial relationships between the various geological materials composing the soil substrate. The strong heterogeneity of the soil cover reflects also in the high spatial variability of the soil water and solute transport properties. Study of the degradation of isoproturon, a cereal herbicide found in the Beauce groundwater, showed that the rate of degradation was highly dependent on soil water content and that isoproturon was highly persistent in subsoil horizons. Mobile /immobile water partitioning was found to play a significant role in solute transport. Isoproturon transp-ort modelling confirmed the importance of this preferential flow process that could explain the contamination levels observed in the Beauce groundwater. The spatial variability of isoproturon fate parameters atthe ouarville catchment scale had characteristics similar to those measured at the field scale. This variability is essentially local (at the metric scale) and does not allow for a geographical definition of risk (i.e. based on soil zoning) at the catchment scale. The spatial variability of isoproturon fate parameters could not be related to the spatial variability of soil characteristics, except the classical relationship between the adsorption coefficient and the organic matter content, which could be spatialised at catchment scale by teledetection. The underdetermined nature of all the other pesticide fate parameters will imply the use of a stochastic risk assessment of the Beauce groundwater contamination by pesticides.; Ce rapport présente les principaux résultats du projet de recherches ESHEL "Caractérisation du risque de contamination de la nappe de Beauce par les pesticides". Le projet ESHEL vise deux objectifs : la compréhension des mécanismes de dissipation des pesticides dans les sols de Beauce, et le développement de bases méthodologiques pour la spatialisation de l'évaluation du risque de contamination des eaux souterraines de Beauce par les pesticides. Le travail de recherche est mené à deux échelons spatiaux : celui de la parcelle agricole, et celui du bassin hydrologique. Une étude pédologique approfondie de la parcelle, aidée d'une prospection géophysique des sols, a mis en evidence une forte hétérogénéité des sols, liée à la complexité des relations spatiales entre matériaux superficiels. Cetté hétérogénéité des sols se traduit par une forte variabilité spatiale des propriétés hydriques et de transport des solutés. L'étude de la dégradation de l'isoproturon, herbicide utilisé sur céréales d'hiver et retrouvé parmi les contaminants de la nappe de Beauce, montre une forte dépendance de la vitesse de dégradation vis-à-vis de l'état hydrique du sol et une persistance élevée de l'isoproturon dans les horizons de subsurface. L'existence d'un mode de transport préférentiel des solutés dans les sols de Beauce, impliquant la partition de 1'eau du sol entre une fraction mobile et une autre immobile, a pu être démontrée. La modélisation du transfert de l'isoproturon dans ces sols confirme l'importance de ce transport préférentiel, qui peut expliquer les niveaux de contamination observés en nappe de Beauce. La variabilité spatiale des propriétés de dissipation de l'isoproturon à l'échelle du bassin hydrologique de Ouarville présente les mêmes caractéristiques que celle évaluée à l'échelle intra_parcellaire. Il s'agit essentiellement d'une variabilité à l'échelle locale (métrique) qui ne permet pas d'approche géographique du risque (i.e basée sur un zonage préalable du sol) à l'échelle du bassin hydrologique. La variabilité spatiale des propriétés de dissipation de l'isoproturon n'a pas pu être reliée à celle des caractéristiques de constitution des sols hormis la relation classique entre coefficient d'adsorption et teneur en matière organique, cette dernière pouvant être spatialisée à l'échelle du bassin par télédétection. Le caractère indéterminé de la variabilité spatiale des nombreux autres paramètres intervenant dans la dissipation des pesticides implique donc l'utilisation d'une approche stochastique pour l'évaluation du risque de contamination de la nappe de Beauce.

[1] Vertical water seepage in soils results in convective heat transport that modifies the temperature profiles and their variations with time; consequently, there is a relationship between temperature profile variations with time and the vertical Darcy velocity associated with the seepage. Considering the annual sinusoidal time variation of the temperature at the soil surface, it can be shown that convective heat transport has a significant effect on the amplitude damping with depth and a negligible effect on the phase lag with depth of the temperature time signal. Standard meteorological stations constitute a relatively dense network, and we show that their routinely collected data can be used to determine an average value of the vertical Darcy velocity, uz, representing the effective annual recharge over long time periods (several years). A new procedure for determining uz from these temperature records is presented. First, the layering of the medium is determined by an electrical sounding. Then the thermal properties of each layer are inferred from the phase lag with depth. Finally, uz is calculated from the amplitude damping. After having tested this approach with synthetic data, we used the 1984-2001 Abbeville (Somme, France) data to determine the average recharge over six 3-year periods. The results are in good agreement with classical meteorological recharge estimates and show a significant increase in the recharge during the last 3-year period, consistent with the observed phreatic 2001 flood event. Specific temperature measurements at appropriate depths and time steps could drastically improve the sensitivity of the method.