Heinschink, K., Sinabell, F., & Tribl, C. (2015). Decomposition of variable costs in the Austrian agricultural production. In Jahrbuch der ÖGA (Vol. 25, pp. 231–240).
|
Mitter, H., Sinabell, F., & Schmid, E. (2015). Impacts of climate and policy change on Austrian protein crop supply balances. In Jahrbuch der ÖGA (Vol. 23, pp. 131–140).
|
Sinabell, F., Heinschink, K., & Tribl, C. (2016). Explicit cost accounting for analyses on climate change adaptation, mitigation and ecosystem service provision in agriculture. In S. Sauvage, J. M. Sánchez-Pérez, & A. E. Rizzoli (Eds.),. 8th International Congress on Environmental Modelling and Software.
|
Kebreab, E., Tedeschi, L., Dijkstra, J., Ellis, J. L., Bannink, A., & France, J. (2016). Modeling Greenhouse Gas Emissions from Enteric Fermentation. In E. Kebreab (Ed.), Advances in Agricultural Systems (Vol. 6, pp. 173–196). Synthesis and Modeling of Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forest Systems to Guide Mitigation and Adaptation, Advances in Agricultural Systems (6).
Abstract: Livestock directly contribute to greenhouse gas (GHG) emissions mainly through methane (CH4) and nitrous oxide (N2O) emissions. For cost and practicality reasons, quantification of GHG has been through development of various types of mathematical models. This chapter addresses the utility and limitations of mathematical models used to estimate enteric CH4 emissions from livestock production. Models used in GHG quantification can be broadly classified into either empirical or mechanistic models. Empirical models might be easier to use because they require fewer input variables compared with mechanistic models. However, their applicability in assessing mitigation options such as dietary manipulation may be limited. The major driving variables identified for both types of models include feed intake, lipid and nonstructural carbohydrate content of the feed, and animal variables. Knowledge gaps identified in empirical modeling were that some of the assumptions might not be valid because of geographical location, health status of animals, genetic differences, or production type. In mechanistic modeling, errors related to estimating feed intake, stoichiometry of volatile fatty acid (VFA) production, and acidity of rumen contents are limitations that need further investigation. Model prediction uncertainty was also investigated, and, depending on the intensity and source of the prediction uncertainty, the mathematical model may inaccurately predict the observed values with more or less variability. In conclusion, although there are quantification tools available, global collaboration is required to come to a consensus on quantification protocols. This can be achieved through developing various types of models specific to region, animal, and production type using large global datasets developed through international collaboration.
|
Leolini, L., Moriondo, M., De Cortazar-Atauri, I., Ruiz-Ramos, M., Nendel, C., Roggero, P. P., et al. (2017). Modelling different cropping systems (Vol. 10).
Abstract: Grapevine is a worldwide valuable crop characterized by a high economic importance for the production of high quality wines. However, the impact of climate change on the narrow climate niches in which grapevine is currently cultivated constitute a great risk for future suitability of grapevine. In this context, grape simulation models are considered promising tools for their contribution to investigate plant behavior in different environments. In this study, six models developed for simulating grapevine growth and development were tested by focusing on their performances in simulating main grapevine processes under two calibration levels: minimum and full calibration. This would help to evaluate major limitations/strength points of these models, especially in the view of their application to climate change impact and adaptation assessments. Preliminary results from two models (GrapeModel and STICS) showed contrasting abilities in reproducing the observed data depending on the site, the year and the target variable considered. These results suggest that a limited dataset for model calibration would lead to poor simulation outputs. However, a more complete interpretation and detailed analysis of the results will be provided when considering the other models simulations.
|