Research
>> internal Site (SSL)

Physics of transport processes in undisturbed ecosystems
BITÖKM 5From 01/1995 to 12/1997
Principal Investigator: Michael Hauhs
Staff: Holger Lange
Grant: 0339476 B Vorhersage und Erklärung des Verhaltens und der Belastbarkeit von Ökosystemen unter veränderten Umweltbedingungen
Forested catchments were characterized by analyzing abiotic fluxes from and to them. Time series of many observables have been investigated using different techniques. The concepts of randomness and complexity proved to be useful in this context. The filter hypothesis could be confirmed, if filtering means reduction of randomness. The ratio of complexities of input and output, however, depends on the temporal resolution: to extract maximal information at maximal complexity, e.g., precipitation should be measured every couple of hours, whereas runoff is best sampled daily. The measures (e.g., algorithmic complexity, metric entropy) are also a sensitive tool to detect data manipulations. The comparison of different catchments shows similarities w.r.t. the randomness distance between input and output, but clear differences when considering the history of the data sets (recurrence plots, power spectra, Hurst coefficients).The reconstruction of breakthrough curves from a series of tracer experiments performed by our group led to a qualification of tracer substances. For the conditions of our experiment, bromide cannot be considered as inert tracer; however, deuterium exhibited a BTC with sharply defined transport parameters.An explicit simulation of runoff from the Lehstenbach catchment made clear that there are two distinct flow components. It could be demonstrated that a reliable estimation catchment transport parameters from output observations is impossible.A mechanism of microscopic sulfate accumulation without equilibrium assumptions or values for desorption isotherms based on a reactiondiffusion process was suggested.The effective dimensionality of the runoff chemistry of the Lehstenbach catchment is low, as shown by a principal component analysis. This enables a neural network simulation, using the first four pc to reconstruct 16 ions in the solution.Characterization of ecosystem time series with complexity measures and related quantities turned out to be a valuable approach to obtain insights beyond standard statistics. Within sensitivity limits, inputs are affected by the biota and used to extract information and to export entropy; the input signals are filtered, the output is simpler than the input, but may be richer in structure. (final report 1998)
List of publications of this Project
Lange, H; Lischeid, G; Hauhs, M: Complexity analysis of time series from two headwater catchments in South Germany in European Academy, Bozen: Hydrology, Water Resources and Ecology of Mountain Areas, Tappeiner, U; Ruffini, FV; Fumai, M, 103106 (1998) 
Hauhs, M; Hornung, U; Lange, H: A Model of Multicomponent Diffusion in a Reactive Acid Soil in Bourgeat,AP; Carasso,C; Luckhaus,S and Mikelic,A (eds.): Proceedings of the Conference "Mathematical Modelling of Flow Through Porous Media" Sant Etienne, World Scientific, Singapore, 231247 (1995)  Details 
Lange, H; Hauhs, M; Schmidt, S: LongTerm Sulfate Dynamics at Lange Bramke (Harz) used for testing two Acidification Models, Water, Air and Soil Pollution, 79, 339351 (1995), doi:10.1007/9789401102612_20  Details 
