Using global Bayesian optimization in ensemble data assimilation: parameter estimation, tuning localization and inflation, or all of the above

Abstract

Global Bayesian optimization (GBO) is a derivative-free optimization method that is used widely in the techindustry to optimize objective functions that are expensive to evaluate, numerically or otherwise. We discuss the use of GBO in ensemble data assimilation (DA), where the goal is to update the state of a numerical model in view of noisy observations. Specifically, we consider three tasks: (i) the estimation of model parameters; (ii) the tuning of localization and inflation in ensemble DA; (iii) doing both, i.e. estimating model parameters while simultaneously tuning the localization and inflation of the ensemble DA. For all three tasks, the GBO works 'offline', i.e. a set of 'training' observations are used within GBO to determine appropriate model or localization/inflation parameters, which are subsequently deployed within an ensemble DA system. Because of the offline nature of the technique, GBO can easily be combined with existing DA systems and it can effectively decouple (nearly) linear/Gaussian aspects of a problem from highly nonlinear/ non-Gaussian ones. We illustrate the use of GBO in simple numerical experiments with the classical Lorenz problems. Our main goals are to introduce GBO in the context of ensemble DA and to spark an interest in GBO and its uses for streamlining important tasks in ensemble DA.

An Efficient Estimation of Time‐Varying Parameters of Dynamic Models by Combining Offline Batch Optimization and Online Data Assimilation

“…Although several studies proposed combining data assimilation and offline batch optimization (e.g., Cleary et al., 2021; Lunderman et al., 2021; Tomizawa & Sawada, 2021), there is no contribution to the estimation of time‐varying parameters by combining them. The aim of this study is to develop an efficient and practical model parameter optimization method which can allow parameters to temporally vary by combining online PF, which has been used in earth system sciences (e.g., Abolafia‐Rosenzweig et al., 2019; Mairesse et al., 2013; Qin et al., 2009; Sawada et al., 2015), with offline batch optimization.…”

Enhancing farm-level economics and animal welfare through sensor-based animal health management

“…However, the choice of calibration technique depends on model complexity. Some examples include grid search, simulated annealing (e.g., Dowsland & Thompson, 2012), Bayesian optimisation (e.g., Lunderman et al, 2021), and a suite of evolutionary algorithms (e.g., Petrowski & Ben-Hamida, 2017).…”