MAIDEN is an ecophysiological model that was created to explore the relationships between climate variability and forest growth-productivity. The simulations can be verified using different data to validate the different processes in the model, including dendroecological growth data, 13C and 18O data and measurements of ecosystem carbon and water fluxes.
Authors and developers
Authors :
Laurent Misson (MAIDEN)
Pierre-Alain Danis (isotopic module)
Developers in alphabetical order :
Etienne Boucher, UQAM, Canada
Guillermo Gea-Izquierdo, INIA, Spain
Fabio Gennaretti, UQAT, Canada
Joël Guiot, CEREGE, France
Laia Andreu-Hayles, Columbia, US
Ignacio Hermoso, UQAM, Canada
Alienor Lavergne, Imperial College, UK
Spatial simulations
Possible but without interaction between pixels
Spatial resolution
1 m2 of forests
Spatial scale
Average tree
Spatial scale of outputs
Forest stand
Temporal resolution
Daily
Temporal scale of outputs
From daily to several decades
Species
The model has already been used and tested for several temperate and Mediterranean species (Pinus spp., Quercus spp., Cedrus atlantica), and for alpine and boreal species from the northern and southern hemisphere (Picea mariana, Picea gluaca, Picea abies , Larix decidua, Pinus sylvestris, Nothofagus pumilio).
Stand types
Mainly monospecific stands
Evolution of the environment
Climate, CO2, δ13C of CO2
Input climate data at daily time step : year, Julian day, maximum temperature (°C), minimum temperature (°C), precipitation (cm), CO2 concentration in the atmosphere (ppm)
Other input variables for the isotopic module : δ18O of precipitation, δ13C of CO2
Consideration of forest management
Resetting the stand parameters (i.e., LAI) is necessary after any forest management intervention
Simulated processes
Transmission and absorption of solar radiation in the canopy
Water fluxes in the canopy and in the soil
Photosynthesis
Evapotranspiration
Phenology
Allocation of carbon to different tree compartments
Isotopic fractionations of carbon, oxygen and hydrogen
Functional traits modeled
Radiation : LAI, coefficients for absorption and reflection of PAR
Water fluxes in the canopy : interception, maximum storage of water in the canopy
Soil : water infiltration, and thickness, grain size and percentage of roots in four soil layers
GPP : photosynthesis (De Pury and Farquhar) and stomatal conductance (Leuning)
Isotopes : multiple fractionations
Phenology : GDD for budburst, photoperiod for senescence
C Allocation : allometric coefficients and C partitioning according to the phenological phase
Main outputs
At daily resolution : water transfers and budget, GPP, NPP, biomass allocated to specific compartments (trunk, roots, reserves, canopy), cellulose stable isotopes
Average computation time
4 seconds per 100 years for a stand
Deterministic or stochastic modeling
Deterministic
Strengths and weaknesses
Strengths :
Use of simple and easily measurable meteorological inputs
Possible comparison with metrics that can be used by dendrochronologists (ring widths, wood density, isotopic ratios of cellulose)
Consideration of sugar reserves (NSCs)
Detailed simulation of C, water and energy fluxes
- The model has many parameters and it makes no sense to calibrate “blindly” (or just statistically) as many parameters as we want until we get a good calibration. A better approach is to set parameters accordingly to physiological knowledge and only calibrate a short list of them according to the goal of the simulation experiment.
Weaknesses :
Lack of a cell differentiation module allowing dimensional comparison with tree-ring width
Lack of modeling of stand ecological processes, including disturbance, mortality, dispersal and recruitment
Lack of spatial interactions between tree individuals
Most recent developments
Development of a snow accumulation and melting module contributing to the water and isotopic balance of the soil. In the past versions, the model only considers the snow cover in the modeling of the terrestrial albedo.