Massonnet F.,Catholic University of Louvain |
Goosse H.,Catholic University of Louvain |
Fichefet T.,Catholic University of Louvain |
Counillon F.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography
Journal of Geophysical Research: Oceans | Year: 2014
The choice of parameter values is crucial in the course of sea ice model development, since parameters largely affect the modeled mean sea ice state. Manual tuning of parameters will soon become impractical, as sea ice models will likely include more parameters to calibrate, leading to an exponential increase of the number of possible combinations to test. Objective and automatic methods for parameter calibration are thus progressively called on to replace the traditional heuristic, "trial-and-error" recipes. Here a method for calibration of parameters based on the ensemble Kalman filter is implemented, tested and validated in the ocean-sea ice model NEMO-LIM3. Three dynamic parameters are calibrated: the ice strength parameter P*, the ocean-sea ice drag parameter Cw, and the atmosphere-sea ice drag parameter Ca. In twin, perfect-model experiments, the default parameter values are retrieved within 1 year of simulation. Using 2007-2012 real sea ice drift data, the calibration of the ice strength parameter P * and the oceanic drag parameter Cw improves clearly the Arctic sea ice drift properties. It is found that the estimation of the atmospheric drag Ca is not necessary if P* and Cw are already estimated. The large reduction in the sea ice speed bias with calibrated parameters comes with a slight overestimation of the winter sea ice areal export through Fram Strait and a slight improvement in the sea ice thickness distribution. Overall, the estimation of parameters with the ensemble Kalman filter represents an encouraging alternative to manual tuning for ocean-sea ice models. Key Points We use an objective method for parameter calibration in an ocean-sea ice model Simulation of ice dynamics is improved with the new parameters The method can be easily extended to GCMs applications © 2014. American Geophysical Union. All Rights Reserved.
Liu B.,King Abdullah University of Science and Technology |
Gharamti M.E.,King Abdullah University of Science and Technology |
Gharamti M.E.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography |
Hoteit I.,King Abdullah University of Science and Technology
Journal of Hydrology | Year: 2016
An ensemble-based Gaussian mixture (GM) filtering framework is studied in this paper in term of its dependence on the choice of the clustering method to construct the GM. In this approach, a number of particles sampled from the posterior distribution are first integrated forward with the dynamical model for forecasting. A GM representation of the forecast distribution is then constructed from the forecast particles. Once an observation becomes available, the forecast GM is updated according to Bayes' rule. This leads to (i) a Kalman filter-like update of the particles, and (ii) a Particle filter-like update of their weights, generalizing the ensemble Kalman filter update to non-Gaussian distributions. We focus on investigating the impact of the clustering strategy on the behavior of the filter. Three different clustering methods for constructing the prior GM are considered: (i) a standard kernel density estimation, (ii) clustering with a specified mixture component size, and (iii) adaptive clustering (with a variable GM size). Numerical experiments are performed using a two-dimensional reactive contaminant transport model in which the contaminant concentration and the heterogenous hydraulic conductivity fields are estimated within a confined aquifer using solute concentration data. The experimental results suggest that the performance of the GM filter is sensitive to the choice of the GM model. In particular, increasing the size of the GM does not necessarily result in improved performances. In this respect, the best results are obtained with the proposed adaptive clustering scheme. © 2016 Elsevier B.V.
Ait-El-Fquih B.,King Abdullah University of Science and Technology |
El Gharamti M.,King Abdullah University of Science and Technology |
El Gharamti M.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography |
Hoteit I.,King Abdullah University of Science and Technology
Hydrology and Earth System Sciences | Year: 2016
Ensemble Kalman filtering (EnKF) is an efficient approach to addressing uncertainties in subsurface groundwater models. The EnKF sequentially integrates field data into simulation models to obtain a better characterization of the model's state and parameters. These are generally estimated following joint and dual filtering strategies, in which, at each assimilation cycle, a forecast step by the model is followed by an update step with incoming observations. The joint EnKF directly updates the augmented state-parameter vector, whereas the dual EnKF empirically employs two separate filters, first estimating the parameters and then estimating the state based on the updated parameters. To develop a Bayesian consistent dual approach and improve the state-parameter estimates and their consistency, we propose in this paper a one-step-ahead (OSA) smoothing formulation of the state-parameter Bayesian filtering problem from which we derive a new dual-type EnKF, the dual EnKFOSA. Compared with the standard dual EnKF, it imposes a new update step to the state, which is shown to enhance the performance of the dual approach with almost no increase in the computational cost. Numerical experiments are conducted with a two-dimensional (2-D) synthetic groundwater aquifer model to investigate the performance and robustness of the proposed dual EnKFOSA, and to evaluate its results against those of the joint and dual EnKFs. The proposed scheme is able to successfully recover both the hydraulic head and the aquifer conductivity, providing further reliable estimates of their uncertainties. Furthermore, it is found to be more robust to different assimilation settings, such as the spatial and temporal distribution of the observations, and the level of noise in the data. Based on our experimental setups, it yields up to 25% more accurate state and parameter estimations than the joint and dual approaches. © 2016 Author(s).
George M.S.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography |
Bertino L.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography |
Johannessen O.M.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography |
Samuelsen A.,Mohn Sverdrup Center for Global Ocean Studies and Operational Oceanography
Journal of Operational Oceanography | Year: 2010
An eddy-permitting HYbrid Coordinate Ocean Model (HYCOM) configured for the Indian Ocean has been validated using both in-situ and satellite observations. The present work focuses on a detailed study of the model's capability to simulate the major surface and subsurface variables realistically.Weekly data from the model for eight years from 1994 to 2001 are used for the evaluation of the surface data.The model simulation of the circulation patterns in the Indian Ocean for both the monsoon seasons and the transition periods matches well with the observations. Comparisons between model and satellite observations for the sea surface temperature (SST) patterns and its temporal evolution showed that the model produces realistic SSTs.The sea level anomalies (SLA) from the model compared with those from the altimeter data confirmed that the model is in good agreement with the observed SLA. A detailed comparison of results from the daily data of the model with the Argo profiles, for the years from 2002 to 2004 showed that the model has a diffuse thermocline with warming in the subsurface waters, but overall, the model simulates the subsurface temperature and salinity patterns well.The validation of the model indicates that the model results are satisfactory and that with improvements in some of the model configurations, it can be implemented in an operational forecasting system for the Indian Ocean.