Open Access
Research Article

Fig. 5

image

Download original image

. Analysis of fit residuals for the annual Ap reconstructions shown in Figure 4a. In each panel, histograms of the fractional error are shown − i.e., the deviation predicted value, Xp, from the observed value, Xo, as a ratio of Xp. Any years giving Xp = 0 or with data availability f < 0.5 for either Xo or Xp are omitted. (a) Xo is the annual mean of the observed factor sin4(θ/2) for 1996–2016 (evaluated from 5-minute data and then averaged) and Xp is the constant value of [sin4(θ/2)]all = 0.355 that must be used in the reconstructions. The grey histogram gives the probability distribution function (pdf) in bins 0.2 wide and the red line the best fit normal distribution to the histogram. The vertical red dot-dash lines mark the upper and lower 1−σ points. The histogram and fitted Gaussian (the green line) in (b) is for Xo = Ap and Xp is the Ap predicted using Pα measured by interplanetary craft (over 1964–2016 but only including years with Pα availability f > 0.5) and the linear regression given in Figure 3a. The red line distribution from part (a) is also reproduced in this panel for comparison purposes. The vertical green dot-dash lines mark the upper and lower 1−σ points of the fitted distribution. The histogram and fitted Gaussian (the blue line) in (c) is also for Xo = Ap and Xp is the Ap predicted using the modelled Pα again using the linear regression given in Figure 3a. The red and green distributions from the upper panels are both included for comparison. The blue dot-dash lines are at the upper and lower 1−σ points of the fitted distribution.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.