Complexity In Microbial Metabolic Processes In Soil Nitrogen Modeling: A Case For Model Averaging

ASU Author/Contributor (non-ASU co-authors, if there are any, appear on document)
Chuanhui Gu Ph.D., Assistant Professor (Creator)
Appalachian State University (ASU )
Web Site:

Abstract: With increasing concerns over environmental problems caused by human disturbances, accurate and reliable predictions of nutrient cycling are essential for sustainable resource management practices. In the last few decades, much effort has been invested in developing biogeochemical models with different complexity that capture nutrient cycling across the scales (Li et al. 1992; Parton et al. 2001; Maggi et al. 2008). Regardless of the detail at which the chemical, biological and physical processes are taken into account, all these models are, in the best case, selective mathematical approximations of processes that are essentially complex. Choosing a level of complexity has been a routine concern by biogeochemical modelers (e.g. Johnson and Omland 2004; Homann et al. 2000) as Biogeochemists have usually struggled to decide what level of complexity in models is actually warranted (e.g., Kimmins et al. 2008). However, this so-called best-model selection really does not solve the significant variability between model performances (Homann et al. 2000). Thus, the standard notion of choosing the right complexity level (e.g., Lawrie and Hearne 2007) doesn’t overcome the model’s structural inadequacy. Furthermore, when there is nearly equivalent support in the observed data for multiple models, it is problematic to choose one model over another. Breuer et al. (2008) recently reviewed some of the nitrogen hydro-biogeochemical models and discussed the importance of addressing various sources of uncertainty in such models, especially the model structural (i.e. complexity) uncertainty. Multi-model averaging can lead to reduction of model selection bias and account for model selection uncertainty.

Additional Information

Gu, Chuanhui, and Newsha K. Ajami (2010). "Complexity In Microbial Metabolic Processes In Soil Nitrogen Modeling: A Case For Model Averaging," Stochastic Environmental Research and Risk Assessment, 24(6) 24.6: 831-44. Version of record available from Springer Verlag. [ISSN: 1436-3240], [DOI: 10.1007/ s00477-010-0381-4]
Language: English
Date: 2010
microbial metabolic process, biogeochemical processes, biogeochemistry, Bayesian Model Averaging

Email this document to