A2.3.1 Getting the Electron Data

The correction factor CF = [emidd / emid60]2 · geomd-1 · dEd-1 is multiplied times the count rates for each detector d to normalize each detector to the central energy of the LEFS60 channel. In addition, the data from the DE (LEMS30) detector is multiplied by the user-supplied value deflect. The anisotropy arrays (three arrays of 8 values each) are filled with the average of the two appropriate sectors times the appropriate correction factor. Missing data (LEMS120) is set to -0.5.

A2.3.2 Getting the Proton Data

The quantities dE, emid, and geom for the proton channels are similarly extracted from the IDF.DAT file except that the P7-8 and P'7-8 channels are considered together as one channel to be compared with FP7 and FP7' .

For P1-4, LEMS channels P and P'1-4 and LEFS channels E and E'1-4 will be used. (See Armstrong [4] re LEMS and LEFS proton and electron passbands.) The user-supplied variable foil fact can be used to adjust for any differences between the LEMS and the LEFS; foil fact = 0.0 will omit the LEFS channels (see section A2.5).

The ions are normalized to the central energy of the LEMS120 channel. Thus the ion correction factor is

CF = [emidd / emid120]2 · geomd-1 · dEd-1 · foilfact

assuming that foil is 1.0 for the LEMS channels.

A2.3.3 Getting the WART Data

The quantities dE, emid, and geom for the WART channels are similarly extracted from the IDF.DAT file. Since the data for these plots comes only from the CA detector, no normalization needs to be done, and the ion correction factor is

CF = geomd-1 · dEd-1

A2.3.4 Scaling

Once the eight normalized flux values are stored for each panel for a given time, the anisotropy can be calculated. The first step is to calculate the average rate of the positive flux values for each panel. An average for the entire sphere is also calculated. (This represents the average omnidirectional flux at this time.) Each of the eight values for each panel is divided by either the panel average or by the sphere average, as requested by the user. The values are then limited to the range (0-scale) for linear scales, or [1/scale - scale] if a log scale was requested. If a log color scale was requested, the log of the anisotropy values is taken. A log scale means that the mapping to the color bar is done with log values.

A2.3.5 Smoothing

If smoothing is to be done, it is performed on the anisotropy values, before conversion to a color bar. There are 160 vertical pixels in each panel and anisotropy values in eight directions. Thus if smoothing is requested, linear interpolation is done between adjacent points to give a total of 160 anisotropy values. If data is missing (e.g., the LEMS120 channel for electron plots, or a LEFS channel when foilfact = 0 for low energy protons), smoothing will not be done across the missing data. (See Figures A2-5a and A2-5b for an example of the same data plotted without and with smoothing.)

A2.3.6 Linear or Log Scaling

For linear scaling each value is divided by the value scale. For log scaling, the quantity

{avalue + log10 (scale) / 2log10 (scale)}

is calculated. This value is between 0 and 1 and is used to select an integer color index between 2 and 242 from the color bar (which has been loaded from the file lansmooth.col as in LANSPECT). A sample calculation of a color index for linear and log scaling is included in Section A2.7.

Next: Appendix 2.3 Anisotropy Calculation

Return to Appendix 2 Main Table of Contents Page

Return to Ulysses HISCALE Data Analysis Handbook Table of Contents