One-dimensional histograms of reconstruction errors y_diff and χ² scores for each reconstructed solar wind parameter. In each panel, the x-axis is constrained to contain 100 bins from −2 to 1. Each of the top panels depicts the reconstruction accuracy of one of the five reconstructed solar wind parameters (a–e) based on the MAPE score, and the bottom panels (f–j) show normalised densities of the χ² score for each determined parameter. Each column refers a different reconstructed parameter, from left to right: v_p (a) and (f) in blue, n_p (b) and (g) in red, T_p (c) and (h) in green, B (d) and (i) in light grey, and n_O^7+/n_O^6+ (e) and (j) in dark grey. In each histogram, an area is marked with grey hatches that contains all data from the 15.9th to the 84.1th percentiles. The respective MAPE score and are indicated in insets in the top and bottom rows. For both, the respective confidence intervals derived from the Monte Carlo runs are included. An additional black histogram outline is included in each panel, which represents the variability of the Monte Carlo simulations based on randomised input data. In each panel, the x-axis contains 30 logarithmic bins. In the bottom row only, the y-axis is also logarithmic with a lower bound of 10⁻⁸, i.e., 0.00001% of the χ² score density, and the histogram is normalised to the sum of the distribution. The inset also depicts the score. In addition, the median of the individual χ² scores is noted to show the spread of the distribution.