Expressed in a simple way, the research field of environmental modelling comprises of two approaches: The empirical and analytical investigation of contexts, influencing factors, and / or differences. While the empirical approach is based on the derivation of findings from collected and measured data, the analytical approach uses proven, mathematically described relationships and combines them into a complex system of algorithms that reflects the functionality of an ecosystem.

In addition to common statistical methods of empirical data analysis (e.g. ANOVA, regression analysis), advanced or more demanding techniques are also applied today. For example, cluster analysis (e.g. k-means method, hierarchical clustering) is often used for classification and categorization of the observed features. The aim is to divide the cases in the data set into groups based on the characteristics of the present variables. The groups should be as homogeneous as possible, but there should be as much difference as possible between the groups.

In this way a classification of soil hydraulic parameters was carried out on the basis of extensive measurements: Depending on the allocation of soil horizons and bulk density, hydrological effects resulted from degradation of peat soils have been verified (DOI: 10.1016/j.geoderma.2017.10.026).

Based on the analytical approach, deterministic and dynamic ecosystem models haven been developed. Such process-oriented ecosystem models are increasingly used to assess management options in forestry and agriculture or to evaluate the effects of climate change on land use systems. For example, the soil type-specific long-term effect of plant residues on the preservation of organic carbon stores was researched in a case study. Plant residues left in the field are considered a key component of soil-based strategies to mitigate climate change, such as the “4 per 1000” initiative (DOI: 10.1088/1748-9326/ab395c).