Revisiting Experimental Catchment Studies in Forest Hydrology (Proceedings of a Workshop held during the XXV IUGG General Assembly in Melbourne, June–July 2011) (IAHS Publ. 353, 2012). Copyright  2012 IAHS Press 117 Merging modelling and experimental approaches to advance ecohydrological system understanding SVEN ARNOLD, ALEX LECHNER & THOMAS BAUMGARTL Centre for Mined Land Rehabilitation, The University of Queensland, Brisbane, Queensland 4072, Australia s.arnold@uq.edu.au Abstract In this paper the need to merge model development and empirical experiments to improve the understanding of ecohydrological systems is emphasised. An ecohydrological model is used as an example to: (1) classify models in ecology and hydrology according to the level of process detail and understanding of system behaviour, followed by (2) a description of the beneficial interactions between model development and manipulative experiments. It is concluded that the development of a grey box model denotes both the starting and “end point” of a cycle of scientific discovery where the design of future empirical and manipulative experiments is governed by model-based hypotheses and uncertainty analyses. Empirical results feedback to the model development process through re-calibrating or re-structuring of the initial model. Key words ecohydrology; model development; experimental design; white box; grey box; black box; scientific discovery INTRODUCTION In recent years there has only been limited progress in merging the methodological strengths of ecohydrology’s parent sciences of ecology and hydrology (King & Caylor, 2011). The strength in ecological methods lies in building knowledge of a system by using controlled manipulative experiments to test hypotheses, whereas hydrology usually focuses on the process-based characterisation of the system. Merging both traditional approaches will “yield rapid advances in our understanding of coupled ecological and hydrological dynamics” (King & Caylor, 2011). To enhance research on understanding ecohydrological system dynamics, empirical and manipulative experiments need to be encouraged to test and refine modelling results. In this paper, the benefits of merging methodological strengths of both ecology and hydrology are described. The range of modelling approaches are reconciled by defining unifying terminology that specifically illustrates the relationship between the two approaches, acknowledging that models exist on a continuum between the two extremes. Finally, following the recommendations on enhancing the practice of ecohydrology stated by King & Caylor (2011) the beneficial interactions between model development and manipulative experiments are described. How the model development process can identify knowledge gaps is demonstrated, which can then be addressed through further empirical or manipulative experiments such as small scale point experiments of plant–soil water interactions (evapotranspiration) or large scale tracer experiments at undisturbed sites of paired catchment study sites. The results of these experiments allow for the re-calibration or, if necessary, rejection of the initial model structure and parameter estimates. This process is discussed in relation to ecohydrological models as depicted through the cycle of scientific discovery (Savenije, 2009). Throughout the paper an ecohydrological model developed by Arnold et al. (2011) is used as an example of this multi-disciplinary approach and the cycle of scientific discovery. MODEL CLASSIFICATION IN ECOHYDROLOGY In a recent study of the interactions between plant communities and soil moisture dynamics in the Brigalow Belt of Australia, Arnold et al. (2012) developed an ecohydrological model which integrated both plant community processes and soil moisture dynamics by linking a simple aggregated plant community model to a storage-based soil water model (Fig. 1). The