09:15 〜 09:30
[SEM16-02] Attempts to produce candidate models for the IGRF-14 (2)
キーワード：国際標準地球磁場、確定国際標準地球磁場、永年変化
The International Geomagnetic Reference Field (IGRF) is a standard mathematical model representing the Earth’s main magnetic field and its secular variation. The model is expressed in terms of spherical harmonic coefficients, known as the Gauss coefficients, updated every five years by the International Association of Geomagnetism and Aeronomy (IAGA). The 14-th generation of IGRF (IGRF-14), which is the latest IGRF revision, consists of the Definitive Geomagnetic Reference Field (DGRF-2020), the IGRF for 2025 (IGRF-2025), and the Secular Variation (SV) from 2025 to 2030 (SV-2025-2030). We submitted a candidate model for SV-2025-2030 on the basis of the machine learning (Sato et al., in preparation) rather than the data assimilation adopted by Minami et al. (2020).
We also planned to submit a candidate model of DGRF-2020, and checked our procedure used to calculate a geomagnetic field model at 2015 by comparing the model with DGRF-2015. We have used geomagnetic data measured by the Swarm satellites, and adopted the Geomagnetic Virtual Observatory (GVO) method (Hammer et al., 2021). We have resampled the geomagnetic field data at the 300 GVOs which are equally distributed at the altitude of 490 km above the Earth’s surface. Then, using this method, we have obtained a geomagnetic field model and compared it with DGRF-2015. However, we found that root-mean-square differences between a geomagnetic field estimated from our model and that calculated from DGRF-2015 at the Earth’s surface, sqrt(dB), was more than 100 nT. This means that our procedure with use of GVO was inappropriate. Hence, we did not submit a candidate model for DGRF-2020. However, we advanced research with respect to DGRF-2020.
In this presentation, we show geomagnetic field data and methods adopted to derive geomagnetic field models, and results for DGRF-2015 and DGRF-2020. We have assumed that the geomagnetic field varies linearly with respect to the time from 1 July 2014 to 30 June 2015 for determination of a geomagnetic field model at the epoch of 2015.0 to eliminate the annual variation. Furthermore, we have selected geomagnetic field data with respect to the local time to avoid effects of ionosphere, the geomagnetic latitudes to avoid effects of field aligned currents, and the Dst index to avoid effects of geomagnetic disturbance due to geomagnetic storm. Then, we have obtained a geomagnetic field model and its temporal variation at 2015 simultaneously. We have found that sqrt(dB) for the model thus obtained is still not small. However, we have found that sqrt(dB) for another model in which geomagnetic field data at polar regions is included is about 13 nT. This means that geomagnetic field data at high geomagnetic latitudes are essential to geomagnetic field modeling. We further investigate other factors for the modeling.
We also planned to submit a candidate model of DGRF-2020, and checked our procedure used to calculate a geomagnetic field model at 2015 by comparing the model with DGRF-2015. We have used geomagnetic data measured by the Swarm satellites, and adopted the Geomagnetic Virtual Observatory (GVO) method (Hammer et al., 2021). We have resampled the geomagnetic field data at the 300 GVOs which are equally distributed at the altitude of 490 km above the Earth’s surface. Then, using this method, we have obtained a geomagnetic field model and compared it with DGRF-2015. However, we found that root-mean-square differences between a geomagnetic field estimated from our model and that calculated from DGRF-2015 at the Earth’s surface, sqrt(dB), was more than 100 nT. This means that our procedure with use of GVO was inappropriate. Hence, we did not submit a candidate model for DGRF-2020. However, we advanced research with respect to DGRF-2020.
In this presentation, we show geomagnetic field data and methods adopted to derive geomagnetic field models, and results for DGRF-2015 and DGRF-2020. We have assumed that the geomagnetic field varies linearly with respect to the time from 1 July 2014 to 30 June 2015 for determination of a geomagnetic field model at the epoch of 2015.0 to eliminate the annual variation. Furthermore, we have selected geomagnetic field data with respect to the local time to avoid effects of ionosphere, the geomagnetic latitudes to avoid effects of field aligned currents, and the Dst index to avoid effects of geomagnetic disturbance due to geomagnetic storm. Then, we have obtained a geomagnetic field model and its temporal variation at 2015 simultaneously. We have found that sqrt(dB) for the model thus obtained is still not small. However, we have found that sqrt(dB) for another model in which geomagnetic field data at polar regions is included is about 13 nT. This means that geomagnetic field data at high geomagnetic latitudes are essential to geomagnetic field modeling. We further investigate other factors for the modeling.