What causes the "seasons" in space weather?

Wednesday, November 15, 2017

It is widely known that the Sun has 'spots'--dark areas on its surface that are well known to contain strong magnetic fields, much stronger than the Earth's field. The number of spots seen on the Sun waxes and wanes with a period of roughly eleven years, the 'solar cycle'. Spots are just one manifestation of 'solar activity', that includes sudden bursts of high energy radiation and charged particles, called 'flares' , as well as huge eruptions of material and magnetic fields from the solar corona, called coronal mass ejections (CME's). Generally both flares and CME's occur more often and are more intense when there are more spots present, near 'solar maximum', the peak of the solar cycle.

Graphic depicting oscillation between differential rotation kinetic energy (solid red curve) and perturbation Rossby waves kinetic energy (red dashed curve)
Left panel (a) shows oscillation between differential rotation kinetic energy (solid red curve) and perturbation Rossby waves kinetic energy (red dashed curve). Typically up to 42 modes in longitude were included in these nonlinear simulations. The units in x and y axis are dimensionless. 100 dimensionless time units correspond to approximately one year; thus the TNO has a period of about 6 months in this case. Frames (b-d) show perturbation flow patterns (in arrow vectors), and thickness of tachocline fluid shell (in color shades, red shade representing swelling, blue depression). Tilts of perturbations are eastward in (b), extracting energy from differential rotation until Rossby waves’ energy is at a maximum; perturbations then go through neutral tilts (c), and then overshoot to acquire westward tilts (d). Enhanced bursts of activity (shown by a semi-transparent gray arrow pointed towards a local peak (yellow-filled ellipse) in sunspot number curve in panel (e)) occurs when perturbation Rossby wave kinetic energy is at its maximum, followed by a relatively quiet season (second semi-transparent gray arrow pointing towards a local dip (the second yellow-filled ellipse) in panel (e)).

Flares and CME's propagate outward through the interplanetary medium; some of them hit the upper atmosphere of the Earth, causing 'space weather' that can be very disruptive to the world's technology-based economy. For example, a few decades ago, a CME about the size of 36 Earths erupted from the Sun and ripped through space at a million miles per hour. Two days later it crashed against the Earth's 'magnetosphere' , and most importantly Canada's Hydro-Quebec power grid crashed, knocking out electricity to six million people for nine hours. In July 2012, the Earth narrowly escaped the effects of both flares and CME's in an event that is estimated to be much more powerful than the one that knocked out power in Canada; Earth was about 1 week away from passing through the path of this extreme event. We know how powerful it was because it was intercepted by the STEREO spacecraft that was able to continue making measurements. This 'superstorm' was the most powerful to leave the Sun since the famous 1859 storm observed by Carrington, that destroyed parts of the 'internet' of that era, the telegraph system.

Recently it was demonstrated from many observations of solar activity, that even within one solar cycle, the Sun has 'seasons' (McIntosh et al. 2015, Nat. Comm. 6, 6491) lasting several months up to a year or so, during which flares and CME's are much more frequent and more intense, which are then followed by relatively quiet seasons of similar duration. If the timing and intensity of a 'bursty' season could be skilfully predicted up to a year in advance, then much better preparations could be made to mitigate the effects of strong space weather events. At present, such an event can be forecast only after the CME is seen to erupt from the Sun, leaving no more than 2 days to prepare, very much in the paradigm of "closing the stable door after the horse bolted". Forecasting the amplitude, duration and timing of the peak of a sunspot cycle remains important, but forecasting the timing and amplitude of a bursty season within the spot cycle is even more important for alerting society to take measures to protect critical electric and electronic infrastructure and mitigate the damage that could be caused by a major space weather event. As seen with the 1859 and 2012 space weather events, even relatively weak cycles can produce extreme events. Such extreme events are substantially more likely to occur during a bursty season, whatever the overall strength of that cycle.

In a paper just published in the Nature Scientific Reports [https://www.nature.com/articles/s41598-017-14957-x], an international team led by NCAR/HAO scientist Dr. Mausumi Dikpati, discovered for the first time the physical origin of the "seasons" in space weather. They demonstrated, using a global shallow-water model of the Sun's shear-layer (called the tachocline), that a bursty season on the Sun followed by a quieter season could occur due to oscillatory nonlinear exchange of energies among the Sun's Rossby waves, differential rotation and sunspot-producing magnetic fields. Under a wide variety of conditions, the period of these Tachocline Nonlinear Oscillations (TNOs) is within 2–20 months, closely matching with the range of quasi-periodic solar seasons observed on the surface.

The figure shows how the TNO's generated in the team's simulation model can be matched up with a sequence of bursty and quiet seasons observed, as well as the physical mechanism of the oscillation between the Sun's differential rotation and Rossby waves in the solar tachocline. Bursty periods are more likely to occur when the Rossby waves are at or near peak amplitude in the oscillation, because that is when the the tachocline top surface is maximally deformed and hence most strongly bulges into the convection zone above, allowing magnetic fields to more easily penetrate the convection zone and rise buoyantly or be carried to the solar surface to erupt as solar activity.

The significance of this theoretical discovery is that if surface observations of the location, timing and strength of these bursty and quiet seasons can be connected by the model to the tachocline nonlinear oscillations (TNOs), then they can be predicted 1–2 years ahead of their occurrence. Dikpati is hard at work collaborating with her NCAR colleagues (Jeff Anderson and Nancy Collins of CISL, Bernadett Belucz, Yuhong Fan, Scott McIntosh and Lisa Upton of HAO) as well as colleagues from several other institutes and universities (Paul Cally of Monash, Aimee Norton of Stanford, Marty Snow of CU/LASP, Doug Biesecker of NOAA, Orkan Umurhan of NASA/AMES, Eyal Heifetz of Tel Aviv) to make that connection, by generalizing their model using modern data assimilation methods, so successfully used for weather forecasting. Assimilating solar magnetogram observations and helioseismic far-side images, the team-effort for predicting future solar bursts is underway.