1 Conclusions

However odd it may seem to start with conclusions, let me do it anyway since (1) that is where most of the interesting information resides and (2) that is what most people will actually read.

I have carefully monitored, and sometimes (see later) analysed in details, ISOCAM data taken on revolutions affected by the solar flare (revolutions 720 to 724). Based on this analysis I can draw the following conclusions:

• Only Revolution 722 is seriously affected by the flare. Revolution 721 is normal from all points of view (glitch rate, glitch type, dark current and sensitivity) as I could see from a detailed analysis of the last CAM observation taken during that orbit (AOMONT.ISOGAL 3P68M7). Revolution 720, during which CAM mostly did 5.04s darks is also unaffected by the flare. On revolution 723, the glitch activity is gone back to a more normal rate, though it appears that we still see an increase in the number of glitches leaving an imprint on the detector.
• CAM observations on revolution 722 (MROWANRO.ISOHDFS HDF-4 LW3 and HDF-2 LW3) are useless. In fact, during such an event, it is pointless to do CAM observations: only bright sources can be observed, with a very short integration time, but nothing more than detection can be achieved as the gain of any pixel strongly varies on timescales of a few tens of seconds due to the impacts. Considering that, the decision was taken to re-schedule all observations from revolution 722.
• CAM recovers quite nicely from these particle showers: No impact is seen either on the sensitivity or on the dark current after revolution 722. This implies that there is no need to switch it off during such events. Yet, as it is impossible to observe during such showers, if one can predict their occurence with a reasonnable accuracy, one might as well turn off the instrument. Given the doubts we have on the predictibility of the events however, we do not recommend switching off of the instrument.