Roadmap

  • Disturbance-storm-time (Dst) index: what and why
  • Solar-wind based forecast of Dst
  • Machine-Learning and artificial neural
    networks
  • Modeling of Dst
  • Results and conclusion

Disturbance-storm-time (Dst) index

  • A measure of magnetic disturbance
    • Solar-wind interaction with Earth’s magnetic field generate electric currents
    • Dst index is a measure of ”ringcurrents” in the magnetosphere
    • Hourly Dst index is calculated using four geomagnetic observatories
    • Different flavors: Kyoto Dst, USGS Dst, Rc index

Why to predict Dst ?

  • Important space -weather specification
    • Ring -current is one of the major current systems in the magnetosphere
    • Critical input to magnetospheric specification models
    • Operational Dst forecast provides early warning
    • Augment NOAA/CIRES real -time magnetic disturbance modeling

Forecasting of Dst using solar-wind data

  • Solar-wind forecasting
    • Less-accurate
    • Lead-time
    • Observatory data not needed
  • Empirical relationship
    • Burton et al (1975), Temerin and Li (2002), O’Brien and McPherron (2000)
  • Physics-based models
    • University of Michigan’s Geospace model
  • Machine-learning approach

Artificial Intelligence

  • An “AI”, or Artificial Intelligence is an intelligent code/machine made by human.
  • AI performs cognitive functions such as learning, problem solving, Planning.
  • AI progression
    • Artificial Weak Intelligence
    • Artificial General Intelligence
    • Strong AI
  • Practical applications are limited to Weak-AI
    • Machine-Learning

 ...

View the entire Presentation:

Forecasting the Magnetic Disturbance-storm-time (Dst) Index using Machine-learning

Manoj Nair, Patrick Alken and Arnaud Chulliat

 

More From Meeting 48