An iterative process for efficient optimisation of parameters in geoscientific models: a demonstration using the Parallel Ice Sheet Model (PISM) version 0.7.3

Physical processes within geoscientific models are sometimes described by simplified schemes known as parameterisations. The values of the parameters within these schemes can be poorly constrained by theory or observation. Uncertainty in the parameter values translates into uncertainty in the outputs of the models. Proper quantification of the uncertainty in model predictions therefore requires a systematic approach for sampling parameter space. In this study, we develop a simple and efficient approach to identify regions of multi-dimensional parameter space that are consistent with observations. Using the Parallel Ice Sheet Model to simulate the present-day state of the Antarctic Ice Sheet, we find that co-dependencies between parameters preclude any simple identification of a single optimal set of parameter values. Approaches such as large ensemble modelling are therefore required in order to generate model predictions that incorporate proper quantification of the uncertainty arising from the parameterisation of physical processes.

Exploring parameter uncertainty in a model of the Antarctic Ice Sheet

<p>Physical processes within ice sheet models are sometimes described by simplified schemes known as parameterisations. The values of the parameters within these schemes can be poorly constrained by theory or observation. Uncertainty in the parameter values translates into uncertainty in the outputs of the models. Proper quantification of the uncertainty in model predictions therefore requires a systematic approach for sampling parameter space. We demonstrate a simple and efficient approach to identify regions of multi-dimensional parameter space that are consistent with observations. Using the Parallel Ice Sheet Model to simulate the present-day state of the Antarctic Ice Sheet, we find that co-dependencies between parameters preclude the identification of a single optimal set of parameter values. Approaches such as large ensemble modelling are therefore required in order to generate model predictions, such as projections of future global sea level rise, that incorporate proper quantification of the uncertainty arising from the parameterisation of physical processes.</p>

An iterative process for efficient optimisation of parameters in geoscientific models: a demonstration using the Parallel Ice Sheet Model (PISM) version 0.7.3

Physical processes within geoscientific models are sometimes described by simplified schemes known as parameterisations. The values of the parameters within these schemes can be poorly constrained by theory or observation. Uncertainty in the parameter values translates into uncertainty in the outputs of the models. Proper quantification of the uncertainty in model predictions therefore requires a systematic approach for sampling parameter space. In this study, we develop a simple and efficient approach to identify regions of multi-dimensional parameter space that are consistent with observations. Using the Parallel Ice Sheet Model to simulate the present-day state of the Antarctic Ice Sheet, we find that co-dependencies between parameters preclude the identification of a single optimal set of parameter values. Approaches such as large ensemble modelling are therefore required in order to generate model predictions that incorporate proper quantification of the uncertainty arising from the parameterisation of physical processes.

Hybrid Model for Wind Turbine Main Bearing Fatigue with Uncertainty in Grease Observations

Available historical field data shows that wind turbine main bearing failure can lead to major operation and maintenance costs due to unscheduled downtime. For legacy turbines, fa- tigue is one of the major failure modes and, to a degree, can be partially modeled with physics-based formulations. Unfor- tunately, existing bearing fatigue models can potentially be inaccurate due to lack of understanding of the lubricant degra- dation. One way to enhance these models is to track the grease damage along with the bearing fatigue damage. However, the need of grease degradation data can become an impedi- ment for such strategy. In this paper, we will demonstrate that it is possible to calibrate grease degradation models with cost-efficient periodic visual inspections. Knowing that such inspections introduce observation uncertainty to the model, we will use a hybrid physics-informed deep neural networks to quantify such uncertainties within our models. We built a hybrid model that fuses the physics-based understanding of the bearing fatigue failure with the ability of data-driven layers to compensate the missing physics, with respect to the grease degradation. The proposed hybrid model is also ca- pable of decoding uncertain visual grease inspections with a custom designed classifier. We illustrate the merits of the model with the support of case studies, where we test inspec- tion with different levels of conservatism to train the model and compare the predictions of these models on an artificial wind park. Results from the case studies indicate the success- ful prognostic performance of the trained with limited and noisy observations. While grease damage is predicted with 0.3% root mean square error as a result of baseline inspection campaign, bearing life is prediction is conservatively off only by months for aggressive turbines that have 10 years of life.

Threshold harvesting as a conservation or exploitation strategy in population management

Threshold harvesting removes the surplus of a population above a set threshold and takes no harvest below the threshold. This harvesting strategy is known to prevent overexploitation while obtaining higher yields than other harvesting strategies. However, the harvest taken can vary over time, including seasons of no harvest at all. While this is undesirable in fisheries or other exploitation activities, it can be an attractive feature of management strategies where removal interventions are costly and desirable only occasionally. In the presence of population fluctuations, the issue of variable harvests and population sizes becomes even more notorious. Here, we investigate the impact of threshold harvesting on the dynamics of both population size and harvests, especially in the presence of population cycles. We take into account semelparous and iteroparous life cycles, Allee effects, observation uncertainty, and demographic as well as environmental stochasticity, using generic mathematical models in discrete time. Our results show that threshold harvesting enhances multiple forms of population stability, namely persistence, constancy, resilience, and dynamic stability. We discuss plausible choices of threshold values, depending on whether the aim is resource exploitation, pest control, or the stabilization of fluctuations.

The CRPS – used as a robust objective function for groundwater model calibration in light of observation and model structural uncertainty

<p>The Continuous Ranked Probability Score (CRPS) is a popular evaluation tool for probabilistic forecasts. We suggest using it, outside its original scope, as an objective function in the calibration of large-scale groundwater models, due to its robustness to large residuals in the calibration data.</p><p>Groundwater models commonly require their parameters to be estimated in an optimization where some objective function measuring the model’s performance is to be minimized. Many performance metrics are squared error-based, which are known to be sensitive to large values or outliers. Consequently, an optimization algorithm using squared error-based metrics will focus on reducing the very largest residuals of the model. In many cases, for example when working with large-scale groundwater models in combination with calibration data from large datasets of groundwater heads with varying and unknown quality, there are two issues with that focus on the largest residuals: Such outliers are often i) related to observational uncertainty or ii) model structural uncertainty and model scale. Hence, fitting groundwater models to such deficiencies can be undesired, and calibration often results in parameter compensation for such deficiencies.</p><p>Therefore, we suggest the use of a CRPS-based objective function that is less sensitive to (the few) large residuals, and instead is more sensitive to fitting the majority of observations with least bias. We apply the novel CRPS-based objective function to the calibration of large-scale coupled surface-groundwater models and compare to conventional squared error-based objective functions. These calibration tests show that the CRPS-based objective function successfully limits the influence of the largest residuals and reduces overall bias. Moreover, it allows for better identification of areas where the model fails to simulate groundwater heads appropriately (e.g. due to model structural errors), that is, where model structure should be investigated.</p><p>Many real-world large-scale hydrological models face similar optimizations problems related to uncertain model structures and large, uncertain calibration datasets where observation uncertainty is hard to quantify. The CRPS-based objective function is an attempt to practically address the shortcomings of squared error minimization in model optimization, and is expected to also be of relevance outside our context of groundwater models.</p>

Multi-model data assimilation techniques for flood forecasts

<p>Floods are the most common and disastrous natural hazards and due to climate change and socio-economic growth they are becoming more dangerous. Early warning systems are one of the best ways to decrease the effect of floods by increasing preparedness. The European Flood Awareness System (EFAS), part of the European Commission's Copernicus Emergency Management Service, provides medium-range ensemble flood forecasts for the whole of Europe but only calibrates its forecasts locally at river gauge stations where sufficiently long and reliable observations are available. These corrections do not consider the natural relationships that occur between points up and downstream. In this PhD project, data assimilation techniques will be used, in post-processing, to combine the available gauge observations with the forecasts. Using a weighting matrix, the influence of the observations will be extended along the river channel network, taking account of the ensemble and observation uncertainty. EFAS uses meteorological forcings from four numerical weather prediction (NWP) systems, so a multi-model approach will need to be developed.  This requires new data assimilation theory and hydrological process knowledge to ensure consistent updates. Delocalising calibrations will improve the accuracy of forecasts at unobserved locations allowing end-users to make more informed decisions to mitigate flood damage.</p>

Combining multiple data sets to unravel the spatiotemporal dynamics of a data-limited fish stock

The biological status of many commercially exploited fishes remains unknown, mostly due to a lack of data necessary for their assessment. Investigating the spatiotemporal dynamics of such species can lead to new insights into population processes and foster a path towards improved spatial management decisions. Here, we focused on striped red mullet (Mullus surmuletus), a widespread yet data-limited species of high commercial importance. Aiming to quantify range dynamics in this data-poor scenario, we combined fishery-dependent and -independent data sets through a series of Bayesian mixed-effects models designed to capture monthly and seasonal occurrence patterns near the species’ northern range limit across 20 years. Combining multiple data sets allowed us to cover the entire distribution of the northern population of M. surmuletus, exploring dynamics at different spatiotemporal scales and identifying key environmental drivers (i.e., sea surface temperature, salinity) that shape occurrence patterns. Our results demonstrate that even when process and (or) observation uncertainty is high, or when data are sparse, if we combine multiple data sets within a hierarchical modelling framework, accurate and useful spatial predictions can still be made.