The hydrological paradox - why the whole is less than the sum of its parts?

Abstract: Hydrological systems are inherently non-linear and exhibit an enormous structural and functional heterogeneity. Strikingly, we can nevertheless successfully simulate stream flow generation and the water balance of river catchments with rather simple models that are largely incompatible with the frequently reported subscale process heterogeneity and non-linearity. Here we argue that subscale structural heterogeneity and randomness must not prevent the emergence of functional simplicity. On the contrary, we found simplicity to emerge at rather small scales, reflecting self-organization in hydrological functioning not despite but due to subscale small-scale heterogeneity and the dissipative nature of hydrological process. While we acknowledge that hydrological landscapes are heterogeneous, they are by no means a random product. Catchments exhibit a considerable spatial organisation, which manifests through structured patterns of topography, soil, vegetation, self-similar surface and subsurface drainage networks and most prominently through ubiquitous preferential flow phenomena. While this organized “catchment from” does strongly determine present storage, cycling and release of water, energy and chemical species, this catchment form has in turn been shaped by of water, energy, and nutrients of the past. Is this “co-evolution” just chance or manifested self-organization? This question has been inspiring many scientists to search for thermodynamic principles that link form and function in the Earth system. Here we will present evidence that a thermodynamic and information theoretic perspective opens up new avenues for (i) diagnosing and explaining self-organization in hydrological dynamics, (i) upscaling of constituting relations and (i) using thermodynamic optimality for hydrological predictions. Close abstractClose abstract