Artifacts in RHESSI Light Curves

The RHESSI detector counting rates are made up of several components in addition to the contributions from solar X-ray emissions. Consequently, the interpretations of variations in the counting rates can sometimes be confusing. This list is intended to describe the various effects that can result in artifacts in the light curves of detector counting rates versus time. They include effects caused by cosmic rays and trapped charged particles passing through the detectors and surrounding material, instrumental effects, and variations that are only seen during solar events. These descriptions are intended to aid the user in differentiating between these various effects and in identifying those variations that can be used to provide information on the solar flare X-ray and gamma-ray emissions.

Artifacts from Charged Particles

• Pseudo-sinusoidal variations with a period of about 45 minutes or half an orbit.
  These variations result from the variations in the cosmic ray flux of charged particles as the magnetic cutoff rigidity changes with geomagnetic latitude. The cutoff rigidity is lowest at the highest north and south geomagnetic latitudes reached by the spacecraft on each orbit. Thus, the detector background counting rate is lowest over the equator and highest at high north and south latitudes. The amplitude of this effect is about a factor of two, varying somewhat with energy, and the spectrum is very hard.
   

• Gradually varying increases in rate at intermediate southern latitudes.
  These are caused by high particle fluxes when the spacecraft gets close to the South Atlantic Anomaly (SAA). The spectrum of these increases is very hard.
   

• More rapid increases in counting rate at intermediate to high north and south latitudes.
  These sometimes occur at magnetic conjugate points in the spacecraft orbit that are on the same magnetic shell. They are caused by precipitating electrons that produce a very hard spectrum. They can last for as short as a minute but are usually quite symmetric in time.

Instrumental Artifacts

• Shutter In/Out
  There are two thin disks of aluminum for each detector that can be moved between each of the nine lower grids and the corresponding detectors to attenuate the flux of lower energy photons. One of the disks is thinner than the other and it attenuates the flux at energies below about 25 keV. The other thicker shutters attenuate the flux to energies as high as 70 keV. The attenuators are generally under onboard software control. During periods of low solar activity, both shutters are moved out of the fields of view of all detectors. When the rate increases during a flare and the dead time reaches a predetermined level like 15%, the thin shutters are moved into place in less than a second. If the rate continues to increase to a second predetermined level, then the second thicker set of shutters is moved into place, further attenuating the flux. Once in place, the shutters cannot be removed for a fixed length of time of the order of 5 minutes. The shutters are removed in reverse order as the flare rate decreases.
   

• Data dropout.
  This is manifest as periods as long as a second or more in which no counts are recorded from a particular detector. This is believed to be caused by charged particles passing through the detectors and producing large pulses that saturate the electronics. The pulse height analysis is designed to provide the highest possible energy resolution and hence shuts down when this happens. It does not allow any further photon to be analyzed until it is fully recovered and can perform the analysis with the required energy resolution. During this dead time, no photons can be handled and so the detector is dead for that length of time. A software correction is applied by default to allow for this dead time in calculating the incident photon flux, but when short time intervals are used the data dropouts will show up as zero fluxes for individual detectors.
   

• Decimation
  When the rate becomes too high or the onboard solid-state recorder (SSR) becomes too full, the photon counts are 'decimated'.  The front and rear detector segments are decimated separately.  For each, there is a decimation level - the fraction of counts that will be removed, and a decimation energy - the energy threshold below which counts will be removed. In practice, a given default decimation level is applied to prevent the SSR from becoming too full and the decimation is increased in steps if a flare occurs or the SSR begins to fill at an unexpectedly high rate. Since decimation is a digital process (unlike the rate reduction produced by the mechanical attenuators), the effect can be accurately accounted for in the software and the user should not have to worry about it. The only noticeable affect should be the reduction in the number of counts recorded and the consequently larger statistical uncertainties in the counting rates.
   

• Sudden increases in detector 8 counting rate lasting for several minutes.
  These are caused by electrical interference from the aft antenna that is situated very close to the detector 8 electronics.
   

Artifacts in Solar Flare Light Curves

• Fast Modulation
  This modulation is produced by the two grids above each detector as the source appears to move in the rotating spacecraft coordinate system. It is the modulation that allows images to be reconstructed. It can be clearly seen with a moderate M-class flare in detectors 5 – 9 by plotting the counting rate with 0.1-s time bins.
   

• Modulation in counting rate with half the spin period.
  This arises during solar flares when the source is away from the spin axis and is the result of the finite field of view through the grids above each detector. All collimators except for those above detectors 7 and 8 have about a 1-degree field of view or smaller in a direction perpendicular to the slits but much greater than this parallel to the slits. Thus, as the source appears to move in the rotating spacecraft coordinate system, the transmission through the grids is modulated with twice the spin period. The phase of this modulation is different for each detector and depends on the orientation of the slits. The amplitude of the modulation depends on the offset of the source from the spin axis.
   

• Modulation in counting rate with a period of ~75 s.
  This can be particularly noticeable during the steady decay in counting rate after a large flare. It is caused by the offset of the instrument imaging axis from the spacecraft spin axis. The amplitude of this modulation is largest for detector #5 light curves because of the relatively large offset of the axis of the bi-grid subcollimator in front of this detector from the imaging axis. This whole issue of recovering the true flare light curve from RHESSI count-rate measurements is discussed in a paper by Zimovets et al. (2010).
   

• Earth Albedo
  This manifests itself during a big flare with a reasonably hard spectrum as a modulation in the counting rate during a flare with a period equal to the spin period. It results from the Compton scattering of the flare X-ray and gamma-ray photons in the Earth’s atmosphere.
   

• Day/Night Transitions
  The Sun is readily detectable at the lowest energies covered by RHESSI down to 3 keV. A significant increase in rate is observed in the lower channels when the spacecraft comes into sunlight each orbit. The magnitude of the increase depends on the soft X-ray flux from all the active regions on the Sun at that particular time. The spectrum of this flux is very soft corresponding to thermal emission from plasma at a few million Kelvin.
   

Non-solar Events

Some increases in the RHESSI counting rates are caused by X-ray photons of non-solar origin. These include transient events such as terrestrial gamma-ray flashes related to thunderstorms, gamma-ray bursts, or outbursts on cosmic sources such as magnatars and Cygnus X1. The strongest steady cosmic X-ray sources, such as the Crab Nebula and sources near the galactic center, can also be seen from the step in the rate when they come out of or go into eclipse behind the Earth.
 

Last updated 15 December, 2010 by Kim Tolbert, 301-286-3965.  Content by Brian Dennis.