In this report we refer to the proton event that affected the SWS observations in revolution 722, since a previous proton event, starting on November 4th, had no impact on the SWS observations. The increase of the proton flux at energies greater than 10 MeV and 100 MeV as observed by the GOES-9 satellite (geosynchronous orbit) started at 06/1305z and 06/1245z, respectively. As can be seen in Figure 1, this is the most intense event occurred in 1996 and 1997, that is, most of the ISO mission. The proton flux was almost three orders of magnitude higher than average for protons with E > 10 MeV and more than one order of magnitude for protons E > 100 MeV. Moreover, an interplanetary shock arrival and a Sudden Storm Commencement took place at 06/2252z, caused by the solar event observed on the 4th. Figures 2a and 2b, taken from the Space Weather Web page of the Space Environment Center (NOAA), show the evolution of the proton and electron flux and of the Planetary K index during the period of intense solar activity under study.
The following six SWS observations were scheduled during revolution 722:
72200302 06/1927
72200805 06/2228
72201405 07/0350
72201607 07/0435
72201702 07/0540
72201901 07/0711
The most affected observation was the first one in the revolution, when the proton flux was maximum. Therefore it appears that the proton flux originated at the solar flare on the 6th is the prime responsible for the observed effects on the detectors, and not the Sudden Storm Commencement that started three hours and a half later than the first observation. The impact of the high electron flux has not been analyzed separately because it would not be possible to disentangle its effects from the proton ones.
The trend analysis indicates that the last observations in the revolution were the least affected ones, and that all the parameters were back to nominal conditions for subsequent revolutions. In the following, a description of the behaviour of different parameters related to the quality of the observations is given.
2. 1 Execution of the observations
Neither real time problems in the execution of the observations were reported nor messages indicating error conditions were registered in the RTA Kernel Log. A memory dump comparison of the instrument carried out in the end of the revolution showed no differences with the expected image. These factors, together with the analysis of the observational data lead to the conclusion that the proton event did not affect the correct execution of the SWS observations.
2.2 Dark currents
An increase in the dark current level was observed in all detector bands, except for band 2 (see upper part of Figure 3). This increase ranged from 25% in band 3 to almost 200% in bands 1 and 4. The dark currents noise increased about 200% in all bands (see lower part of Figure 3). The impact on the quality of the observation is mainly due to the higher noise, which is extensive to the signal data. The higher level in the dark currents is a problem because they decrease quickly with time in the first observations, but not necessarily linearly, implying a less accurate dark current subtraction.
With the current data, it seems that the proton event has not damaged particularly band 3 detectors, that is, the trend of the dark currents increase has not worsen. 2.3 Photometric checks
The response of the detectors as reflected by the internal photometric checks was reduced during the proton event, by 30%, 10% and 15% in bands 2, 3, 4, respectively (see Figure 4). The change in responsivity in band 1 cannot be assessed because monitoring with internal calibrators is not done on routine basis. The response variation with time during this revolution may lead to a greater uncertainty in the flux determination, because in a long observation this variation cannot be accounted by the photometric check.
2.4 Glitch rate 2
The glitch rate increased dramatically in all bands reaching values between five and ten times the nominal rate, decreasing with time after the maximum during the first observation. In Figure 5a, detectors from bands 1 to 3 are shown and Figure 5b displays the glitch rate for all detectors in band 4 (which is normally the one most affected by glitches). The higher glitch rate also results in a higher number of ramps with multiple glitches and out of limits conditions, reducing the amount of accurate data.
In order to check the proton energies to which the SWS detectors are more sensitive, the glitch rate versus proton flux has been plotted in Figure 6 for a detector in each detector band and three proton energies (E > 1 MeV, E> 10 MeV, and E > 30 MeV). The stars in the plots correspond to two observations in revolution 722 in which the gain was set to 1 or 4. They are marked differently because observations with gain 1 or 4 show a lower glitch rate than those with gain 16, which can mask the trend. As can be seen, the correlation between proton flux and glitch rate is better for higher energies, becoming more clear for protons with energies greater than 30 MeV.
2.5 SWS HK parameters
No deviation from nominal conditions has been observed in any analogue HK parameter.
The proton event occurred in revolution 722 had an important impact on the quality of the SWS observations, although they were executed correctly. Due to the lower S/N ratio (three times smaller than expected), uncertainties in dark current subtraction and photometric calibration, and being the number of ramps affected by glitches between five and ten times greater, we have recommended to reschedule all the SWS observations in this revolution.
Since all detector parameters were back to normal in revolution 723 and subsequent ones, and no problems were observed in the instrument commanding, no special real time procedure needs to be applied in case such an event occurs again (TBC by Home institute).
