How to choose the most appropriate model for your application?
All three models COUP, JULES and LPJ-GUESS are generally quite flexible with respect to terrestrial ecosystem types. However, they operate at slightly different scales and each has its particular strengths. In combination, this means that the model toolbox provides models that together cover most relevant ecosystem types and land uses, address key questions related to biogeochemical cycling and address questions at different temporal and spatial scales.
Your choice of one or more models therefore depends on your particular scientific focus. In the table below we compare a range of different aspects of the three models, which should help you to make a qualified choice of best model(s) for you needs.
Summary of input requirements for the models is also available.
| Feature / Process | COUP | LPJ-GUESS | JULES |
|---|---|---|---|
| Model structure | |||
| Model structure | Modular with "switches" making exclusion/inclusion of various processes or ecosystem components possible | Modular with switches for a number of ecosystem processes and configurable parameters, extensible interface for dealing with plant functional types (vegetation). Custom input/output module is a responsibility of the user. | Modular, with switches for processes and ecosystem components. Switches to select variables to output. |
| Parameter setting support and guidance | Described in manual. Support for ranges embedded in model | Parameter setting guidelines included in text file embedded in the model | Described in manual. Input parameters (e.g., for PFTs) can be easily specified and varied spatially and temporally. |
| Validation and uncertainty | Uncertainty based methods are available as integrated parts of the model based both on general Monte Carlo Methods and more formal probabilistic methods | No internal validation tools included. Analysis of the model performace (modelled vs. observed values) needs to be done externally. | Validation must be carried out externally. |
| Processes | |||
| Plant growth and carbon uptake | GPP driven by by light, water, Nitrogen availability. Optionally use of Farquhar approach, light use efficiency or water use efficiency approach. Plant development calculated with different dynamic approaches assuming stage of development with different seasonality for allocation and litter fall. Carbon assimilation allocated to the different parts of depending on growth stage and water and N stress of plant. | Growth is annual, based on accumulated NPP over the previous year. Carbon uptake is daily based on plant requirements vs. available resources. | GPP driven by light, water, CO2 on timestep scale. Growth and vegetation competition on timescale of several days. GPP calculated in terms of three limiting rates: N, light and rate of product transport. GPP reduced by soil moisture stress and ozone effects. Several options for leaf to canopy scaling, with the option of including multi-layer radiation interception, sunfleck penetration, and differing reactions to diffuse and direct radiation. |
| Nutrient dynamics | The model includes major nitrogen transformation processes in the soil. No treatment of other nutrients (e.g. P and base cations) | The model includes description of N dynamics in ecosystem, including photosynthesis limitation, accumulation in plant tissues. CENTURY-like soil pools and SOM dynamics for both N and C. | N concentration can vary through canopy. |
| Carbon and nutrient interactions | The model consider the major carbon components (above and below ground plant, SOM, microbes, and DOC) which are directly interacting with macro-nutrients (especially nitrogen), plant nutrient re-allocation is also considered in the model. | C/N ratios constrained to a certain extent, N leaf content limits C assimilation, C/N ratios affect maintenance respiration. | Nitrogen cycle follows C cycle. N limits photosynthesis and affects respiration. |
| Soil organic matter dynamics and microbiology | Dynamic soil organic matter from different pools is considered. Soil micro-organisms are covered and accounted for the fluxes between different compartments in the soil. | Two microbial pools exist for both C and N. | Q10 and RothC soil carbon model both implemented with one soil pool. RothC can also be used with four (decomposable plant, resistant plant, microbial biomass, long-lived humidified) soil carbon pools. |
| Vegetation change and biodiversity | The model does not include explicit species and biodiversity. However a number of functional plant types can be specified with various degrees of interactions. | Vegetation is dynamic, based on pre-defined set of plant functional types (PFTs). PFT parametrisation is static. Nitrogen, water and light competition simulated on individual or PFT basis. Mortality includes both an age and a growth-stress related components that are modelled stochastically. | Vegetation is dynamic, using five PFTs. PFT parametrisation is static except LAI and canopy height, which respond to growth. Competition hierarchy trees-shrub-grass. Two types of trees compete with each other based on relative height, same for the two types of grass. |
| Soil physics, hydrology and energy | The model simulates in details exchanges of water and energy in a vertical profile from the atmosphere through the vegetation to the underlying soil. | Soil hydrology is represented by a simple 2.5- layer bucket model. Otherwise soil profile is not differentiated. No lateral flow. Temperature at 25 cm is calculated, based on water content in the top layer. Soils are separated into texture groups. Energy balance is not calculated. | Soil hydrology represented by layers (default 4). Soil moisture and temperature calculated for each of these layers, along with Darcian flow between them and surface exchange of H20, CO2 and energy in the top layer, including vegetation effects. Sub-surface runoff is drainage from the bottom of the lowest layer. Energy balance is calculated. |
| Drivers | |||
| Climate change | The model simulate climate change (and different climate scenarios), because meteorological data are used as driving forces. The importance of vegetation and snow conditions on the local soil climate is considered. | The model has been used with a wide variety of forcing data sets including various climate projections for the XXI century. There is no fundamental differences between simulations driven by historical climate or a scenario. | Meteorological data used to drive the model, therefore will respond to climate change. |
| Land use change | The model can describe land use change as respond to climate and management. The competition between the different plant functional types are not developed as part of the model. | Land-use change could be simulated by regulating establishment and mortality. | Land use can be provided as an input, so could be changed and the model will respond. If vegetation competition is switched on, then vegetation cover can change in response to changing climate. |
| N deposition | Both dry and wet nitrogen deposition are considered in the model. | Both forms (reduced and oxidised) of dry and wet deposition can be considered, along with the fertilisers. | Not included |
| Management | The model includes some options of soil management including fertilization, tillage and irrigation. Many options for plant management with respect to clearing, thinning and harvest. | Fertilisation is available, other management options are not included in this version of the model. | Not included |
| Extreme events | Continuous response functions for soil moisture and temperature are considered for a wide range of climate conditions. | Stochastic formulation of generic patch-destroying disturbances. Simple fire scheme included. | Continuous response to soil moisture and temperature |