Ewelina Florczak
Gather.town id
MIS02
Poster Title
Are magnetospheric MHD models any good for predicting magnetic field and GIC in the UK?
Institution
University of Edinburgh
Abstract (short summary)
Space weather events cause disturbances of the Earth’s geomagnetic field. Rapid field fluctuations often result in the induction of quasi-direct currents, known as geomagnetically induced currents (GICs) in conductive structures on the Earth’s surface. Since GICs can be damaging to high-voltage power networks, pipelines or railways, a good forecasting capability is important in order to mitigate their impacts.
We used magnetohydrodynamic models of the magnetosphere and ionosphere, (SWMF/BATS-R-US, SWMF/BATS-R-US+RCM, GUMICS-4 and GORGON) to simulate ground magnetic field variations for the 7/8 September 2017 event, based on solar wind parameters propagated to the simulation domain. Modelled values of the northward and eastward magnetic field components show differences in both amplitude and temporal variability compared to the corresponding measurements from three UK observatories: Hartland, Eskdalemuir and Lerwick. The BATS-R-US model produces the closest agreement in terms of the northward component, whilst GORGON performs best in terms of the eastward component. The addition of Rice Convection Model (RCM) tend to overestimate the field values. Results indicate the accuracy of ground magnetic field forecast decreases with increasing latitude.
The resulting northward and eastward geoelectric field is calculated from the magnetic field using magnetotelluric transfer functions, which is then extrapolated to compute the GIC for a high-voltage UK power network. The GIC response to a uniform electric field of 1 V/km shows that substations (nodes) located near coastlines are affected the most. It is found that GICs computed from BATS-R-US modelled values are most accurate for nodes located at higher latitudes, whilst GUMICS-4 prediction performs best at lower latitudes.
We used magnetohydrodynamic models of the magnetosphere and ionosphere, (SWMF/BATS-R-US, SWMF/BATS-R-US+RCM, GUMICS-4 and GORGON) to simulate ground magnetic field variations for the 7/8 September 2017 event, based on solar wind parameters propagated to the simulation domain. Modelled values of the northward and eastward magnetic field components show differences in both amplitude and temporal variability compared to the corresponding measurements from three UK observatories: Hartland, Eskdalemuir and Lerwick. The BATS-R-US model produces the closest agreement in terms of the northward component, whilst GORGON performs best in terms of the eastward component. The addition of Rice Convection Model (RCM) tend to overestimate the field values. Results indicate the accuracy of ground magnetic field forecast decreases with increasing latitude.
The resulting northward and eastward geoelectric field is calculated from the magnetic field using magnetotelluric transfer functions, which is then extrapolated to compute the GIC for a high-voltage UK power network. The GIC response to a uniform electric field of 1 V/km shows that substations (nodes) located near coastlines are affected the most. It is found that GICs computed from BATS-R-US modelled values are most accurate for nodes located at higher latitudes, whilst GUMICS-4 prediction performs best at lower latitudes.
Plain text (extended) Summary
Rapid magnetic field fluctuations associated with space weather events can induce geomagnetically induced currents (GICs) in conductive structures on the Earth’s surface. Since GICs can be damaging to high-voltage power networks, pipelines or railways, a good forecasting capability is important in order to mitigate their impacts.
We used currently available magnetohydrodynamic (MHD) models of the magnetosphere and ionosphere, (SWMF, SWMF coupled with RCM, GUMICS-4 and GORGON) to simulate ground magnetic field variations for the 7/8 September 2017 event, based on solar wind parameters propagated to the simulation domain. Modelled values of the northward and eastward magnetic field components show differences in both amplitude and temporal variability compared to the corresponding measurements, acquired via INTERMAGNET, from three UK observatories: Hartland, Eskdalemuir and Lerwick.
Results indicate the accuracy of ground magnetic field forecast decreases with increasing latitude. The eastward component (By) tends to be predicted more accurately than the northward component (Bx). It was found that the SWMF performs best in northward component forecast. Coupling with the Rice Convection Model (RCM) overestimate the field value causing poor agreement with measurements. The eastward component of B-field is best predicted by the GORGON model. Despite being the most accurate in terms of Bx, the SWMF shows the largest error in By forecast.
The resulting northward and eastward geoelectric field was calculated from the magnetic field values using magnetotelluric transfer functions, which was then extrapolated to compute the GIC for a high-voltage UK power network. The GIC response to a uniform electric field of 1 V/km shows that substations (nodes) located near coastlines are affected the most. Those nodes were then selected for further investigation.
Results suggest that GICs computed from the modelled values are in closer agreement with measurements at nodes at higher latitudes, where the SWMF performs the best. The GICs computed for substations at lower latitudes show larger discrepancies. In this case, the GUMICS-4 tends to be the most accurate, despite its rather average performance in ground magnetic fields forecast.
Since the accuracy of the simulation of ground B-fields by MHD models included in this study is rather unsatisfactory, attempts to improve their prediction ability will be considered. Applying a method, commonly used in climate modelling, known as downscaling may potentially enhance the forecast accuracy by introducing smaller scale local variations in global variables simulated by each model.
We used currently available magnetohydrodynamic (MHD) models of the magnetosphere and ionosphere, (SWMF, SWMF coupled with RCM, GUMICS-4 and GORGON) to simulate ground magnetic field variations for the 7/8 September 2017 event, based on solar wind parameters propagated to the simulation domain. Modelled values of the northward and eastward magnetic field components show differences in both amplitude and temporal variability compared to the corresponding measurements, acquired via INTERMAGNET, from three UK observatories: Hartland, Eskdalemuir and Lerwick.
Results indicate the accuracy of ground magnetic field forecast decreases with increasing latitude. The eastward component (By) tends to be predicted more accurately than the northward component (Bx). It was found that the SWMF performs best in northward component forecast. Coupling with the Rice Convection Model (RCM) overestimate the field value causing poor agreement with measurements. The eastward component of B-field is best predicted by the GORGON model. Despite being the most accurate in terms of Bx, the SWMF shows the largest error in By forecast.
The resulting northward and eastward geoelectric field was calculated from the magnetic field values using magnetotelluric transfer functions, which was then extrapolated to compute the GIC for a high-voltage UK power network. The GIC response to a uniform electric field of 1 V/km shows that substations (nodes) located near coastlines are affected the most. Those nodes were then selected for further investigation.
Results suggest that GICs computed from the modelled values are in closer agreement with measurements at nodes at higher latitudes, where the SWMF performs the best. The GICs computed for substations at lower latitudes show larger discrepancies. In this case, the GUMICS-4 tends to be the most accurate, despite its rather average performance in ground magnetic fields forecast.
Since the accuracy of the simulation of ground B-fields by MHD models included in this study is rather unsatisfactory, attempts to improve their prediction ability will be considered. Applying a method, commonly used in climate modelling, known as downscaling may potentially enhance the forecast accuracy by introducing smaller scale local variations in global variables simulated by each model.
URL
ewflor98@gmail.com
Poster file