Sandra Chapman - ‘Data analytics’ approaches to space weather in space and time

Sandra Chapman (University of Warwick) - ‘Data analytics’ approaches to space weather in space and time

Abstract:
Space weather and solar terrestrial physics observations are increasingly becoming a data analytics challenge and there are common approaches with other fields such as earth climate observations. Whilst focussing on specific applications, this talk aims to present generic methodology for inhomogeneous ‘real world’ data.
Single long-term observations over multiple solar cycles: Over the last 14 cycles we have geomagnetic indices such as the aa index which are poorly resolved in amplitude but nevertheless contain information on the likelihood of occurrence of extreme space weather events, and we discuss how this can be quantified, setting the Carrington event in the context of extreme events that have occurred over the last 150 years. Each solar cycle is unique both in amplitude and duration, and we show how the Hilbert transform of daily sunspot number can be used to map each cycle onto a uniform time-base or ‘sun clock’. We can then use this mapping to quantify how activity in space weather relevant parameters, and in particular the likelihood of super-storms, is modulated by the solar cycle. This reveals the phases in the solar cycle where there is a clear ‘switch on’ and ‘switch off’ of activity and these are in principle predictable. Sun clocks for both the Schwabe and Hale cycles will be discussed.
Multiple observations of spatially and temporally varying fields: spatially irregular observations such as the SuperMAG collated 100+ ground based magnetometer stations in the auroral region can be tested for spatio-temporal patterns of correlation using dynamical networks. Whilst networks are in widespread use in the data analytics of societal and commercial data, there are additional challenges in their application to physical time series. We are able to construct dynamical networks direct from SuperMAG. The transient dynamics of the auroral current system is captured by the spatio-temporal patterns of correlation between the magnetometer time-series and can be quantified by (time dependent) network parameters. Cross-correlation lags can be used to construct directed networks which give directions and timescales for propagation. This offers the possibility of characterizing detailed spatio-temporal pattern by a few parameters, so that many events can then be compared with each other and with differing theoretical predictions of the ‘typical’ substorm current system.