Uncertainty quantification in canopy turbulence

Abstract:

Most problems relating to micro-meteorology are typically tackled via numerical simulations based on a deterministic set of model parameters. The specification of such parameters is a task fraught with difficulty, given that an exact measure is often impossible. As a result, despite recent advancements in the physical understanding and modeling capabilities of micro-meteorological flows, parameter uncertainty still remains an important source of error, which challenges the interpretation of measurements and the accuracy of numerical models. The author's research addresses this problem, namely the quantification of uncertainties (and propagation thereof) in problems relating to micro-meteorology, with a lens on canopy flows and Atmospheric Boundary-Layer (ABL) turbulence. Here, an overview on the ongoing research is provided, with a particular focus on a computational fluid mechanics model-validation effort and on the quantification of uncertainties in fluid mechanics models for the representation of the flow over plant canopies. The goal of the model-validation effort is to determine whether general purpose finite-volume-based solvers are capable of accurately representing ABL turbulence. The importance of validating numerical models is highlighted, and key aspects of the validation process for ABL flow applications are discussed. The goal of the uncertainty quantification effort is to quantify uncertainties in model parameters for the simulation of plant canopy flows based on the Reynolds-averaged Navier-Stokes equations, and to determine how parameter uncertainties propagate into selected flow statistics. Various sources of uncertainty are considered, ranging from the collection and post-process of leaf area density measurements to the inherent heterogeneity of the canopy structure. The same uncertainty is then propagated into selected flow quantities.