Geomagnetic models

[TODO: intro to models, spherical harmonics; interactive visualisation?; add links for Swarm model descriptions, and how to access and evaluate them]


Overview of models

There are many models which aim to describe the global geomagnetic field (or one of the sources within it). The field itself is complex so naturally there are many decisions to be made when creating a model:

This situation has led to a range of models with different applications and varying levels of cross-compatibility. The following table lists most of the contemporary models available which are applicable near to Earth (between Earth’s surface and at low- to mid-Earth orbit). If you are new to this, then a good place to start is the IGRF, the International Geomagnetic Reference Field, which is collaboratively produced by the scientific community to give a reasonable estimate of the “main field” (i.e. that which is mostly produced by the core).






Combined models:
CM5 Sabaka et al. 2014; [NASA]
CIY4 Sabaka et al. 2018
CHAOS-6 - Finlay et al. 2016; [DTU]
POMME-11 - Maus et al.2010; [CIRES/NCEI]
MEME - Thomson et al. 2010; Hamilton et al. 2015; [BGS]
Main / core field models:
IGRF12 - - - Thebault et al. 2015; [IAGA]
WMM2015 - - - Chulliat et al. 2015; [NOAA/BGS]
EMM2017 - - [NOAA]
GRIMM - - - Lesur et al. 2008; [GFZ]
COV-OBS - - - Gillet et al. 2015; [DTU]
gufm1 - - - Jackson et al. 2000
Crustal field models:
NGDC-720 - - - Maus et al. 2010; [CIRES/NCEI]
MF7 - - - Maus et al. 2008; [CIRES/NCEI]
LCS-1 - - - Olsen et al. 2017; [DTU]
Ionospheric field models:
DIFI-3 - - - Low-mid latitudes; Chulliat et al. 2013; [CIRES/NCEI]
AMPS - - - Polar regions; Laundal et al. 2018; [pyAMPS]; [WebAMPS]
Crustal grids:*
EMAG2v3 - - - Meyer et al. 2017; [NOAA]
WDMAM - - - Lesur et al. 2016; [IAGA]
ADMAP-2 - - - Golynsky et al. 2018
CAMP-M - - - Gaina et al. 2011

* Whereas the other items in this list are smoothly varying models that provide a prediction of the field everywhere, these crustal grids are maps of discrete values which are compiled (and cross-calibrated) from a large number of individual surveys.

Other lists of models

Swarm models

A number of models are produced within the framework of the Swarm mission. These “Level 2” data products themselves are described in the official documentation as well as in scientific papers. The models follow a naming convention according to the physical source they target and the operational processing chain that produces them:

Model of the … C: Comprehensive D: Dedicated F: Fast-track
MCO: Core
MMA: Magnetosphere -
MIO: Ionosphere -
MLI: Lithosphere -

The comprehensive inversion chain aims to co-estimate the four sources together. Dedicated inversions attempt to filter out the effect of other sources and make a better isolation of a particular source. The fast-track products deliver an estimate of the field more rapidly than the other products, so they are available sooner but are less accurate.

Model name Source field Notes (maximum spherical harmonic degree, l, and order, m)
Comprehensive inversion - Updated yearly with 6 week latency
MCO_SHA_2C Core SH l=1-18
MIO_SHA_2C Ionosphere (Sq) SH l=1-60, m=1-12 *
MLI_SHA_2C Lithosphere SH l=16-150
MMA_SHA_2C Magnetosphere SH l=1; 1 hour cadence
Dedicated inversions - Updated yearly with 6 week latency
MCO_SHA_2D Core SH l=1-18
MIO_SHA_2D Ionosphere (Sq) SH l=1-60, m=1-12 *
MLI_SHA_2D Lithosphere SH l=16-150
MIO_SHA_2E Polar ionosphere (the AMPS model) SH model also parameterised by solar wind
Fast track inversions
MCO_SHA_2F Core SH l=1-18; updated every 3 months with 1 week latency
MMA_SHA_2F Magnetosphere SH l=1; 1.5 hour cadence; updated daily
FAC_TMS_2F Field-aligned currents Time series along the orbit
* SH in QD coordinates, modulated by seasonal and daily wavenumbers; valid within ±55° QDLat
In two parts, "primary" and "secondary" for external and internal (induced) fields respectively