CORAL Seminar: Allan Gross, Department of Business Development and Technology, Aarhus University

Title: Parameter Assimilation of Physical/Chemical System

2018.01.15 | Bodil Krog

Date Fri 02 Mar
Time 14:00 15:00
Location Fuglesangs Allé 4, 8210 Aarhus V, building 2621(B), room 122

Presenter: Allan Gross, Department of Business Development and Technology, Aarhus University

Title: Parameter Assimilation of Physical/Chemical System

Abstract: Science and technology problems are often investigated and verified using field measurements, laboratory experiments and/or theoretical simulations. However, all three methods have their limitations. Field measurements and laboratory experiments have measurement uncertainties and detection limits, or do not include sufficient measurements to fully understand the system. Theoretical models can be based on parameters from field measurements, laboratory experiments and/or atomistic theoretical methods. Theoretical models have therefore uncertainties with respect to their model parameters or insufficient information.
    Data Assimilation (DA)is a well know method used in Numerical Weather Prediction models, Atmospherically Chemical Transport Models etc. to make a “perfect” simulation. In this approach it is assumed that the model is perfect, and DA is used to find the optimal “model state” for a given set of observations, i.e. is does not improve the theoretical approach in the model: Thus, how can we improve the theoretical models, i.e. reduce the discrepancies between experimental/observation data and theoretical (micro/macro) models?
    We will present a methodology which can reduce the discrepancies between experimental data and theoretical models using combined parameter assimilation with sensitivity analyses for hundreds of variables and processes simultaneously. The theory has been implemented into the 2. generation MOON model (MOON2), which includes the following two-ways interactions: (1) a process compiler, (2) a differential equation solver, (3) a model optimization method, (4) sensitivity analyses, and(5) a graphical tool.
    Results from test of the new methodology has been validated on nine different Dimethyl Sulphide (DMS) chamber experiments. Results from the DMS study will be presented together with an overview of the general application of the model approach.

Organizer: Lars Relund Nielsen, CORAL

CORAL seminars
14304 / i32