Space Weather Forecasting: Sudden Commencements and Machine Learning

Abstract: Rapid changes in the surface geomagnetic field can induce potentially damaging currents in artificial conductors on the ground.  There are a large range of magnetospheric phenomena that can cause such rapid changes, but they are ultimately driven by the interaction between the solar wind and the Earth’s magnetic field.  Forecasting intervals when there is a risk of damaging induced currents would enable crucial mitigating action to be taken.  However, the complexity and range of magnetospheric phenomena that could be responsible for the generation of significant currents complicates efforts to provide precise yet reliable forecasts. 

In this seminar, I will discuss the scientific background, design and development of a pair of forecasting models that are currently being integrated into the UK Met Office’s Space Weather Operations Centre (MOSWOC).  The first model targets a magnetospheric phenomenon known as a Sudden Commencement, caused by the impact of an interplanetary shock on the Earth system, enabling the evaluation of the risk posed by an observed interplanetary shock.  Second, I will discuss a model that is designed to directly predict when large induced currents are likely in the near future at a given location.  When running at MOSWOC, the models will be driven by a near real time feed of data from the L1 point upstream of the Earth.  I will discuss practical considerations encountered during the process of transitioning scientific models to use operational near real time data.