Statistical modelling can be used to draw conclusions from areas where direct observation data is not available. The models are based on existing data and can be used to assess the characteristics of the seabed over extensive areas. Models can also be used to produce species distribution maps.
Models are built from the collected environmental data
Modelling enables the study of marine nature and its phenomena over large geographical areas. The models so to speak generalise or extrapolate the data collected from the surveyed sites to the surrounding sea areas.
At its core, modelling is always based on real observational data. For example, the data may be the observed sites of a particular species or habitat. Modelling combines data with a map layer depicting environmental conditions, produced for example from satellite imagery. The result is a probability distribution of the species or habitat over a wider range of observed sites.
Such models can be used to estimate not only the distribution of species but also the most important breeding areas of fish, for example. Physical marine research also produces hydrodynamic models. The models can predict, for example, waves, currents, and water levels, as well as variations in sea temperature and salinity.
Human impacts on marine nature can also be modelled
Modelling is also suitable for human impact assessment. By knowing how different human activities are located on the map, and knowing the species and habitats present in the same areas, models can predict the impacts and risks to nature.
Such impact models can be used, for example, in planning the use of marine areas. With the help of models, planners can examine how much the area of influence of the planned activities overlaps with important nature values and how severely the activity would weaken these values.
Impact models also consider local environmental conditions. For example, the depth of a marine area or its vulnerability to wind and waves can either enhance or reduce the environmental impacts.