Solar Wind Plasma Parameter Variability Across Solar Cycles 23 and 24: From Turbulence to Extremes

Abstract

Solar wind variability spans a wide range of amplitudes and timescales, from turbulent fluctuations to the 11 year solar cycle. We apply the data quantile‐quantile (DQQ) method to NASA/Wind observations spanning solar cycles 23 and 24, to study how the uniqueness of each cycle maximum and minimum manifests in the changing statistical distribution of plasma parameters in fast and slow solar wind. The DQQ method allows us to discriminate between two distinct components of the distribution: the core region simply tracks the solar cycle in its moments but shows little sensitivity to solar wind state or the specific activity of each cycle. This would be consistent with an underlying in situ process such as turbulence driving the evolution of fluctuations up to an outer scale. In contrast, the tail component of the distribution is sensitive both to the differences between the maxima and minima of cycles 23 and 24, and the fast or slow state of the solar wind. The tail component varies over the solar cycle in such a way as to maintain a constant functional form, with only its moments varying with solar activity. Finally, after isolating the core region of the distribution, we test its lognormality over the solar cycle in each solar wind state and find the lognormal provides a more robust description of the statistics in slow wind than fast; however, in both states the goodness of fit is significantly reduced at solar maximum.

AE, DST, and Their SuperMAG Counterparts: The Effect of Improved Spatial Resolution in Geomagnetic Indices

“…For example, where two linear regions exist in a plot, one may be indicative of a core component of the distribution where the second component corresponds to the tail in the distribution. This kind of dual‐distribution QQ plot is common in space weather variables where the core of the distribution corresponds to small‐scale turbulent behavior and the tail of the distribution identifies large‐scale driven behavior (Tindale & Chapman, ). Here, we will compare the distribution of − SMR to − D ST using the QQ plot.…”

Reproducible Aspects of the Climate of Space Weather Over the Last Five Solar Cycles

“…The detailed quantification of the time variation of the statistical distribution of key space weather relevant variables is thus a topical question. The distributions of geomagnetic activity (Tanskanen et al, ) and geomagnetic indices (Campbell, ; Lockwood et al, , ) vary both within and between solar cycles (Tindale & Chapman, , ). The distributions of solar wind variables can be non‐Gaussian (Veselovsky et al, , and references therein).…”

“…However, as Lockwood et al () note, the lognormal does not necessarily provide the best fit across the full distribution for all variables of interest. Indeed, Tindale and Chapman () found that the distribution of solar wind magnetic field can be far from lognormal. Here we will show how to test whether the functional form of the distribution of a given variable is invariant against solar cycle dependent changes, without assuming any specific functional form, lognormal or otherwise, for the underlying distribution.…”

“…Quantile‐quantile (QQ) plots (Gilchrist, ) compare the statistical distribution of one data sample against that of another. Using a QQ plot analysis over intervals of the maxima and preceding minima of solar cycles 23 and 24, we recently found (Tindale & Chapman, , ) that samples of the largest events change in a different manner to the bulk of the distribution as we compared one solar maximum interval to the other. The distribution is multicomponent: It has (at least) two components—a core or bulk of smaller amplitude, more frequently occurring values, and a long tail of larger, less frequently occurring values.…”

Comparison of SuperDARN peak electron density estimates based on elevation angle measurements to ionosonde and incoherent scatter radar measurements

“…To alternatively assess the quality of the N m F 2 estimates from RKN measurements, we performed the additional analysis by producing the so-called quantile-quantile (Q-Q) plots of the RKN and CADI/ RISR data. We adopted an approach similar to that of Tindale and Chapman (2017). The data for each of the instruments were first ranked according to reported values of the electron density and then the quantiles were selected between 5 and 95% of the total number of points with a step of 5%.…”