Analysis of High Cadence in Situ Solar Wind Ionic Composition Data Using Wavelet Power Spectra Confidence Levels

Abstract

The variability inherent in solar wind composition has implications for the variability of the physical conditions in its coronal source regions, providing constraints on models of coronal heating and solar wind generation. We present a generalized prescription for constructing a wavelet power significance measure (confidence level) for the purpose of characterizing the effects of missing data in high cadence solar wind ionic composition measurements. We describe the data gaps present in the 12-minute ACE/SWICS observations of O 7+ /O 6+ during 2008. The decomposition of the in-situ observations into a 'good measurement' and a 'no measurement' signal allows us to evaluate the performance of a filler signal, i.e., various prescriptions for filling the data gaps. We construct Monte Carlo simulations of synthetic O 7+ /O 6+ composition data and impose the actual data gaps that exist in the observations in order to investigate two different filler signals: one, a linear interpolation between neighboring good data points, and two, the constant mean value of the measured data. Applied to these synthetic data plus filler signal combinations, we quantify the ability of the power spectra significance level procedure to reproduce the ensemble-averaged time-integrated wavelet power per scale of an ideal case, i.e. the synthetic data without imposed data gaps. Finally, we present the wavelet power spectra for the O 7+ /O 6+ data using the confidence levels derived from both the Mean Value and Linear Interpolation data gap filling signals and discuss the results.

Formation and Reconnection of Three-dimensional Current Sheets with a Guide Field in the Solar Corona

“…Wavelet methods are commonly used to characterize structure coherency scales in non-steady signals (e.g., see Edmondson et al 2013, and references therein). In this analysis, we employ the Morlet wavelet transform of the normalized mass density, ρ(z)/ρ 0 , given by,…”

COHERENT STRUCTURE IN SOLAR WIND C6 +/C4 +IONIC COMPOSITION DATA DURING THE QUIET-SUN CONDITIONS OF 2008

“…Table 1 summarizes the distribution of the data gaps in the C 6+ /C 4+ 12-min and 2-hour data. In general, the distributions of gap duration and maximum gap lengths are similar across the respective carbon data sets (e.g., see Edmondson et al 2013). The 2-hour averaged data are robust with 97.5% valid measurements.…”

“…The wavelet power significance level procedure allows useful information to be derived from even sparsely populated data sets. Here we apply the novel procedure described by Edmondson et al (2013) to the 2-hour and 12-minute averages of the C 6+ /C 4+ ionic composition ratio data.…”

“…To obtain information about the range of coherency scales of transient structures (see Edmondson et al 2013, Appendix A for discussion) in the charge state ratios, we construct a probability distribution (PDF) from the histogram of the local 2D maxima in each of the wavelet power spectra as a function of correlation scale. Here, we identify the local 2D maximum in the wavelet power within the 80% significance contours (to illustrate, note the local power maximum -bright yellow -in Figure 2 occurring at the approximate time and scale coordinates, t ∼ 40 days and s ∼ 9 days, respectively).…”

“…In order to underscore the physical origins of structure within the C 6+ /C 4+ ionic composition data, we have examined the wavelet properties of a Monte Carlo ensemble of simulations synthetic C 6+ /C 4+ data using the method outlined in Edmondson et al (2013) based on the first-order Markov chain model used by Zurbuchen et al (2000). We have calculated 100 realizations of the model carbon time series at both the 2 hour and 12 minute cadences.…”