J.P. Mol

Metafore: The Afforest Deposition-Soil-Water-Vegetation Metamodel

This chapter describes the development of the METAFORE metamodel for the AFFOREST project, focusing on aspects that are important in defining the role of the metamodel in the entire system. Two modes are distinguished: one in which the METAFORE metamodel operates in a batch mode for generating the AFFOREST-sDSS tables for decision support, and another mode of operating with an extended user interface and extended possibilities for evaluating detailed results. The various detailed process-based models are from different sources and each of the institutes had experts on the processes that were modelled. From these models, the individual partners developed meta-descriptions of their parts of the system. The task of the METAFORE metamodel was to combine all this knowledge into a single model executable, and assure a correct calculation of the values needed for the AFFOREST database and the AFFOREST-sDSS. The design of the METAFORE distinguishes a metamodel framework and the model components or submodels. The metamodel framework focuses on the interface and communication between the different submodels and it is responsible for the communication between the submodels. In this design, the submodels are merely servers, waiting to be initialized or called to perform one step of the simulation (i.e. one month or one year of the simulation). To do this, each submodel has only a limited set of exposed methods. Although the detailed process models, as part of their scientific development process, have been extensively validated and calibrated, this does not automatically assure a proper simulation of the processes by the metamodel. During the entire process of the development of METAFORE, the simulation behaviour of the METAFORE modules were constantly tested against the detailed process models. In the end, METAFORE has been developed as a simplification of the detailed models with a lesser demand on data. This means that the detailed behaviour in the process models can at best result in similar but aggregated behaviour in the METAFORE metamodel. It is concluded that the results are satisfactory.

Integrated Assessment of Impacts of Atmospheric Deposition and Climate Change on Forest Ecosystem Services in Europe

Important forest ecosystem services are pollutant filtering relevant for an adequate water quality (regulating service), wood production (provisioning service) with related carbon (C) storage (regulating service) and the provision of a habitat for a diversity of plants and animals (supporting service). Nitrogen (N) and sulphur (S) deposition affect these ecosystem services. In this chapter, we describe the application of the soil model VSD, in combination with the forest growth model EUgrow and the plant species occurrence model PROPS to quantify the impact of N and S deposition on: (i) changes in soil buffering, in terms of accumulation or depletion of the pools of N, base cations (BC) and aluminium (Al), and changes in nitrate (NO3) and Al concentration in soil water, (ii) forest growth and carbon sequestration, and (iii) plant species diversity. Results showed that the depletion of Al and BC pools and the soil water concentrations of NO3 and Al increased strongly between 1950 and 1980, followed by a decrease between 1980 and 2010, reflecting the strong initial increase and subsequent decrease in N and S deposition in both periods, respectively. The impact of future emission reductions on the various parameters in the period 2010–2050 was larger than the climate change impact. Unlike soil and water quality, both N deposition and climate change had on average a positive impact on carbon sequestration. N deposition was calculated to be the dominant driver of changes in forest growth in the past (period 1900–2000) and climate change for the future (period 2000–2050). Plant species diversity changed hardly in scenarios assuming constant climate and low N deposition reduction, significantly at constant climate and strongly decreasing N deposition, and sharply when both climate and N deposition changed, especially in areas with a pronounced temperature change.

Modelling the Afforested System: The Soil and Water Compartment

Existing complex mechanistic models require too many input data, which are generally unavailable for application at a regional scale. For the development of a simplified soil model, however, we used existing complex models to start with. This soil model includes all relevant processes in order to simulate the temporal trajectory of carbon (C) sequestrations, nitrate leaching and water recharge. The soil model was derived from the existing models NUCSAM, SMART2 and SMB. For the water balance the existing model WATBAL was used. The principal question was whether the derived simplified soil model is acceptable for use within the AFFOREST project. To test this, the soil model was applied to two afforested oak chronosequences and three spruce chronosequences in Sweden, Denmark and the Netherlands. Validation of the soil and water model was performed by:

Modelling the Afforested System: The Forest/Tree Model

A forest/tree model has been developed of which the main growth processes are based on the CENW model. The model links the flows of carbon (C)), energy, nutrients and water in trees and soil organic matter. Modelled tree growth depends on physiological plant factors, the size of plant pools, such as foliage mass, environmental factors, such as temperature and rainfall, and the total amount and turn-over rates of soil organic matter, which drives mineraliZation of soil organic nitrogen (N). The forest/tree model has been developed as a generic model for coniferous trees. In addition, the model has been extended generically for deciduous trees by shedding of leaves in autumn and including growth buds in which C can be stored during winter time, and from which re-growth is initiated in spring. In spring the initial C gives the leaf activity an augmentation for photosynthesis. For the purpose of the available input parameters in the AFFOREST project it was needed to change the daily time step of the original CENW-model into a fortnightly time step. In relation to this, the original simple hump functions for N, water and temperature dependencies are replaced by smooth minimum/optimum/maximum functions. The model was validated against data obtained for Quercus robur (Oak) from the AFFOREST chronosequence in Vestskoven in Denmark. The different tree stands covered a chronosequence over a period of 35 years. The forest/tree model was successfully able to simulate all aspects of tree growth, which included water, N and C dynamics reflected in biomass production. An associated structure variable, i.e. mean tree height was also simulated successfully, which is an important input parameter for the amount of atmospheric N deposition in forests. The model includes essentially all relevant pools and processes that determine forest growth under a range of natural conditions.