The MHYDAS Modelling of HYdrology in AgroSystems

MHYDAS (Modélisation Hydrologique Des AgroSystèmes) is an hydrological model for water exchanges, pollutants and erosion transport in cultivated landscapes


The MHYDAS model (Moussa et al., 2002, Hydrol. Process. 16) allows the representation of water production and transfer on surface, the surface-subsurface simplified exchanges and the water balance at different scales - from plot to watershed, and from rain event to annual or pluri-annual simulations. It also includes thematic modules as pollutant fate and transport, and erosive transfer.

The MHYDAS model takes into account specific features of agricultural landscapes : the influence of landscape discontinuities (reachs, plots, banks...), runoff-infiltration partition according to land use, exchanges between groundwater tables and ditches, pollutants fate ...

It is available through the OpenFLUID modelling software, as a pluggable set including a graphical extension for the digital landscape representation (Geo-MHYDAS) and several coupled simulators for processes modelling.

Processes modelling

Processes modelling

The dynamics of the MHYDAS model is based on coupled simulators allows the representation of several processes. Currently, the surface hydrology module is available as free download : it is composed of a production function and a transfer function. The production function splits the rainfall into infiltration and runoff on the surface units with the Morel-Seytoux method (1978). The transfer function transfers the runoff onto the different spatial units (surface and reach segments) and is based on the semi-analytical solution of Hayami (1951).
This module requires the rainfall as inputdata, a parametrisation related to the choosen resolution methods and a parametrisation of the spatial units based on geographical and hydrological data (spatial units geometry, soil hydraulic properties and initial moisture).

The actual research focuses on modelling processes such as surface hydrology, surface-subsurface exchanges, erosion transport, contamination and pollutant transport... Several numerical approaches for modelling of a same process are also considered: finite differences methods or (semi-)analytical solutions,...

If you want to use the MHYDAS model for other processes or using other numerical solutions, please feel free to contact us.

Associated references:
- H. Morel-Seytoux. Derivation of equations for variable rainfall infiltration. Water Resources Research , 14:561–568, 1978.
- Hayami, S. 1951. ‘On the propagation of flood waves’, Disaster Prev. Res. Inst. Bull., 1, 1–16.

Landscape representation

Landscape representation for the MHYDAS model is based on taking into account the spatial heterogeneities of the cultivated landscape. They lead to hydrological discontinuities and fast responses at high spatial resolutions.

The digital landscape representation for MHYDAS model requires :

  • the identification of the natural and anthropogenic landscape elements to represent : ditches, rivers, plots, soil horizons, ...,
  • the discretisation of the homogeneous spatial units in agreement with processes modelling,
  • the connection of these spatial units through an oriented topology determined by water paths,
  • the parametrisation of these spatial units according to processes modelling.

In order to help the modelers for the definition of the digital landscape representation for the MHYDAS model, the Geo-MHYDAS tool (Lagacherie et al , 2010) can be used. It automates a set of spatial analysis commands for performing the landscape discretisation as spatial units, with connections and parametrisation. Geo-MHYDAS is available as a plug-in for the OpenFLUID graphical interface.

Landscape representation

Examples of works

The MHYDAS model was used to model the hydrological processes at various scales, from plot to watershed.

It was also coupled to other models :

  • with the WATSFAR model : this coupled model allows to represent more finely the surface-subsurface water and pollutants exchanges. It replaces the production function of the MHYDAS model in order to enhance the computation accuracy of the runoff water height in downstream and of the soil hydraulic state of the plot.

  • with the decision-making model DHIVINE (Paré, 2011) : this coupled model allows to model the impact of cultural practices at watershed scale. The MHYDAS model represents the water and pollutants exchanges according to several climatic and practice scenarios. The DHIVINE model represents the spatio-temporal distribution of cultural practices on each plot and models the technical itineraries of the wine-growers. The DHIVINE model relies on the DIESE environment (Discrete Event Simulation Environment) which allows to model the functionning of the agricultural production systems ( Martin-Clouaire et Rellier, 2003).
    Associated references:
    - Martin-Clouaire, R. and Rellier, J. (2003). Modélisation et simulation de la conduite d'unsystème de production agricole. In Actes de la 4ème Conférence Francophone de MOdélisation et SIMulation : Organisation et Conduite d'Activités dans l'Industrie et les Services, MOSIM'03, pages 699704, Toulouse, France.