Uncertainty analysis of complex hydro-biogeochemical models
Bitte beziehen Sie sich beim Zitieren dieses Dokumentes immer auf folgende
Freie Schlagwörter (Englisch):
GLUE , SPOTPY , CMF , LandscapeDNDC , biogeochemistry
Institut fÃ¼r Landschaft-, Wasser- und Stoffhaushalt
Tag der mÃ¼ndlichen PrÃ¼fung:
Kurzfassung auf Englisch:
This thesis is about complex hydro-biogeochemical models and their practical applications. Several modelling practices and their associated uncertainty are investigated in this joined project of the working groups of Prof. Dr. Lutz Breuer, Justus Liebig University Giessen, and Prof. Dr. Klaus Butterbach-Bahl at Karlsruhe Institute of Technology. The aim of the project is to develop strategies for reducing the climate footprint of agricultural production and to quantify uncertainties of model-based strategies for low emission pathways, while at the same time increasing the credibility in model predictions by evaluating not only trace gas emissions, but also plant growth and hydrological fluxes. A motivation that is next to me driven by an increasing demand of the scientific community, governmental and non-governmental organizations.
During the three-yearâ€™s project, I setup different methods to access parameter sensitivity, parameter and structure uncertainty of environmental models. The methods were combined in a statistical parameter optimization tool for python (SPOTPY) to perform various model diagnostics in a straightforward way. Both working groups and others use the tool now in joined as well as individual studies.
The SPOTPY package enabled us to gain a deeper understanding of the underlying processes and limitations of the investigated complex hydro-biogeochemical models. A key result of my study is that the tested models still lack on robustness to generate outputs for multiple ecosystem services. A change of awareness of site and data managers is required as sensors and small-scale variability of site properties can cause low performance in terms of model predicting capability.
Under this impression, I equipped a study area with greenhouse gas emissions and other important water, carbon and nitrogen fluxes measurements on arable land, grassland and forest. I used these measurements in a model-data fusion approach as a final contribution to my dissertation. This study allowed me to derive missing model processes that would potentially increase model simulation performances, if implemented into the biogeochemical model. My findings provide a strong motivation to enhance our understanding of the hydro-biogeochemical system and guide future work.
Veröffentlichungsvertrag für Publikationen ohne Print on Demand