Last Updated: 05/20/2024


Version: 2012

LANL* is a tool developed for quickly obtaining L* values, six orders of magnitude (~ one million times) faster than convectional approaches that require global numerical field lines tracing and integration. This model is based on a modern machine learning technique (feed-forward artificial neural network) by supervising a large data pool obtained from the IRBEM library, which is the traditional source for numerically calculating the L* values. The pool consists of about 100,000 samples randomly distributed within the magnetosphere (r: [1.03, 11.5] Re) and within a whole solar cycle from 1/1/1994 to 1/1/2005.

There are seven LANL* models, each corresponding to its underlying magnetic field configuration that is used to create the data sample pool. They are Olson and Pfitzer quiet model (OPQuiet), Pfitzer and Olson dynamic model (OPDyn), Tsyganenko 1989 model (T89), Tsyganenko 1996 model (T96), Tsyganenko 2001 quiet model (T01Quiet), Tsyganenko 2003 storm model (T01Storm), and Tsyganenko and Sitnov 2005 model (T05).


The LANL* model uses solar wind conditions (and (G1, G2, G3), (W1, W2, W3, W4, W5, W6) indices for T01 and T05 models respectively), local pitch angle, position, Mcllwain L shell, and magnetic field at the mirror point. The solar wind conditions (and G, W indices for T01 and T05 respectively) used for each LANL* model are consistent with the input for its underlying magnetic field configuration.

When using the CCMC web interface for instantaneous calculation of L* values, the user chooses the underlying magnetic field configuration, position (XYZ-GSM [Re]), pitch angle, start date/time and duration, and output frequency. The solar wind parameters are internally determined from the Qin-Denton omni2 database downloaded from the ViRBO website. The Mcllwain L and magnetic field at the mirror point are also internally determined from the IRBEM library.

For each underlying magnetic field model, below are the valid ranges for the inputs: OPDyn: density ∈ [5, 50], velocity ∈ [300, 500], Dst ∈ [100, 20], rGEO < 60 RE OPQuiet: rGEO < 15.0 RE T89: Kp ∈ [0, 9], rGEO < 70 RE T96: Dst ∈ [-100, 20], Pdyn ∈ [0.5, 10], |IMF By|< 10. |IMF Bz| < 10, rGEO < 40 RE T01Quiet: Dst ∈ [-50, 20], Pdyn ∈ [0.5, 5], |IMF By| < 5, |IMF Bz| < 5, G1 ∈ [0, 10], G2 ∈ [0, 10], xGSM > -15 RE T01Storm: xGSM > -15 RE T05: xGSM > -15 RE

In addition, the LANLstar model presents its constraint based on its learning/training process: Dst [nT]: (-422, 62) Solar wind velocity [km/s]: (237.1, 1188.5) Solar wind density [cm3]: (0.1, 100) IMF By [nT]: (-47.9, 40.26) IMF Bz [nT]: (-62.6, 47.2) Dynamic pressure [nPa]: (0.03, 69.4)


The output from the LANL* model are L* value at the selected time, position, and pitch angle as well as McllWain L shell, the second adiabatic invariant I, magnetic field at the mirror point, and solar wind conditions (and G, W indices for T01 and T05 respectively).

Model is time-dependent.


  • Geospace
  • Magnetosphere / Inner Magnetosphere / RingCurrent
  • Magnetosphere / Inner Magnetosphere / RadiationBelt

Space Weather Impacts

  • Near-earth radiation and plasma environment (aerospace assets functionality)



Code Languages: Python 2


Publication Policy

In addition to any model-specific policy, please refer to the General Publication Policy.