| | Time-dep. internal | All |
---|

Data subset/Norm | \(N_i\) | Tr(\({\tilde{R}}_i\)) | % | Tr(\(R_i\)) | % |
---|

CHAMP | 1 132 075 | 881.48 | 8.0% | 2993.49 | 13.8% |

Swarm A and B | 1 390 766 | 601.19 | 5.5% | 1443.54 | 6.7% |

Observatories | 1 786 944 | 186.72 | 1.7% | 186.72 | 0.9% |

Cryosat and DMSP | 7 722 827 | 210.44 | 1.9% | 7598.06 | 35.2% |

Subtotal | 12 032 612 | 1879.83 | 17.1% | 12 221.81 | 56.6% |

\(\Lambda\) | 11 820 | 9085.17 | 82.9% | 9 383.19 | 43.4% |

Total | 12 044 432 | 10 965.00 | 100.0% | 21 605.00 | 100.0% |

- For data subsets, \(N_i\) is the number of data in subset
*i*. For the regularization norm \(\Lambda\), \(N_i = \text {rank}(\Lambda )\). The matrices \({\tilde{R}}_i\) refer to the diagonal block of \(R_i\) corresponding to the time-dependent internal field parameters. Tr is the trace operator