Dimensions and indicators for assessing model complexity - Methods and material - iLand model complexity and its niche in the landscape of models

Dimensions and indicators for assessing model complexity

The landscape of models, i.e. the analysis space for model complexity in this exercise, was defined by three dimensions of complexity, i.e. structural (i.e. population dynamics), functional (i.e. ecophysiology) and spatial (i.e. landscape context) complexity. Each dimension was operationally defined by a set of indicators, representing major ecosystem processes and traits (see Table 1). Model complexity was assessed for every indicator using an ordinal scoring relative to the selected reference models (see below), where the most complex representation of the process among the reference models received the highest (5) and the most simple representation the lowest score (1). Aggregation of indicators within the three dimensions was performed by a mean operator.

Table 1: Indicator definition for the three assessment dimensions

structural complexity (population dynamics)structurestructural representation of the ecosystem, its constituents and their interactions; includes the approach to model competition
compositioncompositional representation of the ecosystem (e.g., species)
regenerationresolution in modeling the processes related to regeneration of trees
mortalityapproach to model the processes related to tree death (excluding disturbance- and management-related mortality causes)
managementrepresentation and flexibility in individual- to stand level management
functional complexity (ecophysiology)primary productionrepresentation of gross primary production and respiration in the model
allocationprocesses structuring ecosystem compartments, i.e. allocation of photosynthetic products to tree organs
element cyclingelement cycles modeled explicitly (e.g., C, N, H2O)
atmospheric processeswhich atmosphere-related factors (e.g., climate parameters) are included and how are they represented
belowground processesdetail and process resolution of soil-bound processes in the model
spatial complexity (landscape context)disturbancesaccount of natural disturbances in the model
managementrepresentation and flexibility in management aspects at the landscape level (e.g., patterns, scheduling)
interactionsspatio-temporal interactions at the landscape scale (e.g., disturbance interactions, seed dispersal)

Created by rupert. Last Modification: Monday 31 of May, 2010 00:54:48 GMT by rupert.