Model applications

Prediction of the future response of ecosystems to changes in pollutant loading and environ-mental conditions is necessary from both a scientific and political viewpoint. These predic-tions provide the only basis for the formulation and quantification of remedial measures. In this respect, mathematical simulation models which are capable of predicting system response under future pollution deposition scenarios represent our best tools. These models must be capable of describing the physical, chemical and biologi-cal relationships observed in ecosystems. The degree of damage to an ecosystem can then be estimated provided the models are based upon dose-response principles. Since the output from a model is only as good as the input data used to drive it, a comprehensive monitoring programme to identify the system function and provide adequate data for model calibration is essential.

Currently available models generally focus on one aspect of an ecosystem, notably atmos-pheric deposition, soil/soil solution chemistry or biology. Some of such models suitable to the ICP/IM for scenario testing are given in figure 3 although not all can be adequately parameter-ised at all sites in the ICP/IM. Initially, hydro-chemical models will be utilised as the core of the modelling programme within ICP/IM as they have already been the subject of some quality control and testing. These models have also been validated to some extent and will provide reliable forecasts of the future changes in water quality which might be expected in relation to anthropogenic input of N and S.

Biological models, on the other hand, are still in their infancy and require further development to achieve the mechanistic level of the hydro-chemical and deposition models. Terrestrial models, incorporating plant growth, are cur-rently under construction and will be capable of predicting long-term plant and vegetation re-sponse to changes in pollutant deposition. Aquat-ic models are currently based on empirAquat-ical rela-tionships between species diversity and surviv-al and physicsurviv-al and chemicsurviv-al parameters of water quality. Nevertheless, these models, when linked to predictions from hydrochemical mod-els, provide useful prognoses of future behav-iour.

12

DEPOSITION MODELS EMEP

Harwell/ASAM North American Models

LIK altitude/veget. modifiers (*)(E) Forest filtering modifiers (*)(E)

HYDROCHEMICAL MODELS PROFILE (**)

MAGIC (**) SOIL (*)/SOILN (*) SAFE (*)

BIOLOGICAL MODELS Forest/Terrestrial

Ca/Al relationships (*)(E) Water stress relationships (*)(E) NUCHEM (**)

DAM (**) VEGIE (* *) FORVIT (*)

HYDROLOGICAL MODELS Water budget models (*)(E) PULSE (*)

SOIL (*)

TOPMODEL (**)

Aquatic

Diversity indices (*)(E) PHABSIM (**)

Figure 3. Some examples of existing and potential models. Those marked (*) are generally applicable to all IM sites whilst those marked (**) will only be applicable at certain sites with a comprehensive database. Those marked (E) are simple empirical relationships.

There is some way to go in model develop-ment before ozone and heavy metals are incor-porated as driving variables into ecosystem models, and even the role of nitrogen is not fully understood. These developments must take pla-ce outside the ICP/IM. As new models are developed, however, they could be widely app-lied within the ICP/IM framework, as could all suitable existing models. The ICP/IM provides a unique database for validation and testing of such models, presuming complete data sets from the participating countries.

The comprehensive database from a few sites within the ICP/IM provides a unique op-portunity for establishing links between models of individual ecosystem components. This will provide a powerful tool for the assessment of ecosystem response to future environmental change and the conceptual framework of the

feedbacks and linkages of such a scheme are demonstrated in figure 4.

Sufficient data exists from many nominated ICP/IM sites to apply certain lumped models, for example MAGIC and SAFE. In any case only a small amount of additional soil informa-tion will be required from some of the sites to enable application to more sites. The advantage of applying the same model to many sites is that a consistent approach can be utilised and sensi-ble comparisons can be made. Once established, a model covering many sites can be used to evaluate emission control strategies, and long term changes in policy, and used to investigate trends in the data. It will only be possible to apply the more complex models e.g. PROFILE at sites where a more detailed database describ-ing soil mineralogy is available.

13

Deposition Models

Soil Models

j I

Surface Water Models Aquatic (chemistry, flow) J Biology Models

Figure 4. The potential of linking models within the framework of the IM programme.

The widespread coverage of sites in the ICP/

IM is ideally designed for the application of models rather than model development. This is supported by the benefits of the central database allowing commonality of approach to data ma-nipulation and aggregation for model calibrati-on. Model development requires specific de-sign of sampling and experimentation and the task is better left to more process oriented rese-arch programmes. The strength of the ICP/IM modelling effort lies in scenario assessment through widespread site applications and the development of technologies for linking mo-dels for integrated assessment of environmental change utilising the integrated data sets availab-le.

The critical load concept utilises the conver-se of the scenario asconver-sessment concept within the framework of a dose-response relationship.

Instead of assessing the changes which might occur under a pollutant deposition reduction scenario the critical load is calculated as the maximum pollutant deposition which will not cause damage to the ecosystem, or component of the ecosystem, in question. In other words, instead of setting a dose and predicting response (scenario assessment) the acceptable response is quantified and the required dose needed to achieve that response is calculated (= critical load).

Dynamic models are not necessary for deter-mining critical loads. However, mechanistic soil chemistry models have been necessary for understanding the processes involved in acidifi-cation and soil and runoff chemistry changes, such as weathering, ion exchange and nitrogen cycling. Dynamic models may be of importance for understanding target loads and temporal aspects of e.g. acidification processes. These models provide the theoretical representation of the dose-response relationship with which to assess the sensitivity and likely damage to an ecosystem. The steady state models like PRO-FILE or simple mass balance approaches are sufficient for determining acidification sensiti-vity of forest soils, lakes, streams and ground-water. There is considerable benefit in carrying out these modelling exercises at a number of sites across a wide region, perhaps most impor-tantly to validate the critical load mapping exer-cises currently underway on a national scale across Europe.

The ICP/IM provides an essential database for model validation and prediction of ecosys-tem future response which is not available from other research programmes or sources, namely:

® The internally consistent and integrated hy-drochemical and biological database will en-able the future interaction between global climate change and atmospheric acidic depo-sition to be modelled and assessed.

® The integrated database will provide the plat-form for development of linked ecosystem models.

® The long time series of data will enable trend detection and model validation at a large number of sites and over a wide geographical area.

A number of opportunities exist at this stage of the ICP/IM to provide for an improved data-base from the point of view of modelling activi-ties:

® Biological surveys will be emphasised and must be undertaken on a regular basis if the aim of linking models is to be achieved.

14

• The basic data for applying models will be determined and must be measured at each intensive site and any gaps in the database will be addressed at the beginning of the perma-nent programme.

• The responsibility for model applications will be organised informally by several groups, not by a single modelling centre (see chapter 10).

Links with other international research and monitoring programmes will be formalised and maintained. As new models and process-es are developed and identified this co-opera-tion will ensure that any new relevant param-eters are incorporated into the measurement programmes at IM sites and into the IM data-base.