Characterization of Earth's magnetic field is key to understanding dynamics of the core. We assess whether Iridium Communications magnetometer data can be used for this purpose since. The 66 Iridium satellites are in 86° inclination, 780 km altitude, circular orbits, with 11 satellites in each of six orbit planes. In one day the constellation returns 300,000 measurements spanning the globe with <2° spacing. We used data from January 2010 through November 2015, and compared against International Geomagnetic Reference Field (IGRF-11) to inter-calibrate all data to the same model. Geomagnetically quiet 24-h intervals were selected using the total Birkeland current, auroral electrojet, and ring current indices. The z-scores for these quantities were combined and the quietest 16 intervals from each quarter selected for analysis. Residuals between the data and IGRF-11 yield consistent patterns that evolve gradually from 2010 to 2015. Residuals for each day were binned in 9° latitude by 9° longitude and the distributions about the mean in each bin are Gaussian with 1-sigma standard errors of ∼3 nT. Spherical harmonic coefficients for each quiet day were computed and time series of the coefficients used to identify artifacts at the orbit precession (8 months) and seasonal (12 months) periods and their harmonics which were then removed by notch filtering. This analysis yields time series at 800 virtual geomagnetic observatories each providing a global field map using a single day of data. The results and CHAOS 7.4 generally agree, but systematic differences larger than the statistical uncertainties are present that warrant further exploration.

中文翻译：

---

用于研究地球磁场的铱通信卫星星座数据

地球磁场的表征是了解地核动力学的关键。我们评估 Iridium Communications 磁力计数据是否可用于此目的。66颗铱星以86°倾角、780公里高度、圆形轨道运行，6个轨道平面各有11颗卫星。在一天之内，星座以 <2° 的间隔返回全球 300,000 次测量。我们使用了 2010 年 1 月至 2015 年 11 月的数据，并与国际地磁参考场 (IGRF-11) 进行了比较，以将所有数据相互校准为同一模型。使用总伯克兰电流、极光电射流和环形电流指数选择地磁安静的 24 小时间隔。该ž- 将这些数量的分数相结合，并选择每个季度中最安静的 16 个区间进行分析。数据和 IGRF-11 之间的残差产生一致的模式，从 2010 年到 2015 年逐渐演变。每天的残差在 9° 纬度和 9° 经度上进行分箱，并且每个分箱中的均值分布是具有 1-sigma 标准误差的高斯分布约 3 nT。计算每个安静日的球面谐波系数，并使用系数的时间序列来识别轨道进动（8 个月）和季节性（12 个月）期间的伪影及其谐波，然后通过陷波滤波去除它们。这种分析产生了 800 个虚拟地磁观测站的时间序列，每个观测站都使用一天的数据提供了全球磁场图。结果和 CHAOS 7.4 大体一致，