Assigned Session: #AGM28: Generic Meeting Session
Translating Observation Uncertainty into Model Calibration unsing the Ensemble Kalman Filter
Abstract ID: 28.7268 | Accepted as Talk | Talk/Oral | 2025-02-28 15:15 - 15:30 | Ágnes‐Heller‐Haus/Small Lecture Room
Oskar Herrmann (0)
Groos, Alexander (1), Tabone, Ilaria (2), Guillaume, Jouvet (3), Fürst, Johannes (1)
Oskar Herrmann ((0) Friedrich-Alexander-Universität Erlangen-Nürnberg, Wetterkreuz, 15, 91058, Erlangen, DE)
Groos, Alexander (1), Tabone, Ilaria (2), Guillaume, Jouvet (3), Fürst, Johannes (1)
(0) Friedrich-Alexander-Universität Erlangen-Nürnberg, Wetterkreuz, 15, 91058, Erlangen, DE
(1) Friedrich Alexander Universität, Erlangen, Germany
(2) Universidad de Conception, Conception, Chile
(3) University of Lausanne, Lausanne, Switzerland
(2) Universidad de Conception, Conception, Chile
(3) University of Lausanne, Lausanne, Switzerland
Accurately modeling how glaciers respond to climate change requires including observation uncertainty in the model calibration process. This study uses the Ensemble Kalman Filter (EnKF) to connect observational data with parameter estimation in surface mass balance models. By using elevation change rates, along with surface velocity measurements, the EnKF helps turn uncertainty in observations into constraints on the admissible space for model parameters. The method also considers observation timing and spatially detailed elevation change data to better match observations with model simulations. Applied to glaciers in central Europe with the Instructed Glacier Model (IGM), this approach improves the calibration of both mass balance and ice flow models. This work provides a practical way to systematically determine uncertainties associated to a simulated present-day glacier geometry. This uncertainty can readily be forwarded in future projections of glacier retreat under climatic warming.
![]() |
||||||||
|