Subject: JEM-X future performance Hallo alle: Jeg vedhaefter et udkast om den forventede JEM-X performance aarene frem til 2013 som indkaldt af Arvind Parmar. Jeg skal sende notatet senest i morgen, saa I bedes kikke det igennem hurtigt. Det ville vaere godt hvis vi kunne udbygge afsnittet om "Scientific Performance, har I nogen forslag der illustrerer JEM-X betydning for INTEGRAL science? X-ray bursters, obscured sources, transients, Earth observations, Galakser? niels **** Hello Peter and Arvind: I attach a short note discussing the expected performance of JEM-X under the assumption that the INTEGRAL mission is extended until the next Solar maximum (2013 ?). The conclusion is that we believe that JEM-X will continue to deliver good science, and no dramatic performance degradations are expected based on the data from the first four years in orbit. I also attach the plot (oulu_64_06.gif) showing the Oulu neutron monitor rates for the last 40 years. From this plot it is clear that forecasting of the Solar activity for the next 10 years must be quite uncertain when it comes to the details. With best regards Niels ******************************************************************* Expected performance evolution of JEM-X over the next 6 years. Niels Lund, Danish National Space Center JEM-X on INTEGRAL consist of two redundant detector units wieving the sky through two coaligned, but independent masks. Normally, only one JEM-X unit is active, the other unit is kept in reserve. The exception from this rule is the Crab calibration operations, typically performed during one orbit every 6 month. 1) JEM-X technical performance. In the four first years of operation JEM-X unit 1 has been used for 34 months and unit 2 for 17 months. During this time there has been only one computer upset in one of the two DPE units and none such upsets in the DFEE units. The DPE problem was quickly cured through power cycling. None of the analog circuitry in the two units or any of the high voltage supplies have showed any noticeable degradation. We conclude that all the JEM-X electronics appears to function in a stable manner and no lifetime limiting elements can be identified at this time. The JEM-X onboard software is also stable. The JEM-X detectors have shown some performance evolution: 1.1) Loss of sensitive area caused by broken microstrip anodes. The JEM-X detectors are based on a fine pattern of gold electrodes deposited on a glass plate (the microstrip plate). Originally there were 256 anode strips in each unit. A few anode strips were damaged during the plate manufacture and detector assembly (4 strips in JEM-X1 and 9 in JEM-X2). In the first weeks after launch anodes were lost at an alarming rate (one per day in each unit), apparently due to discharges initiated by the passage of heavy cosmic rays. This rapid erosion was halted through a reduction by a factor three of the gas gain of the detectors. This was achieved through a reduction of the operating voltage from about 900 to 800 V. After this gain reduction the anode loss rate have been reduced to about 5 anodes per year of detector usage (12 anodes lost over 900 days). At the present time the "dead-anode" count stand at 28 for JEM-X1 and 29 for JEM-X2. We have therefore lost about 11 % of the sensitive area in each unit. At the present rate of new anode losses we expect this loss to evolve as follows (assuming an even use of the two JEM-X units): Remaining fraction of effective area: Current: 89 % 2009: 87 % 2011: 85 % 2013: 83 % We conclude that this loss will not be significant within the time horizon discussed here - this will be true even if we loose one unit completely and therefore have to use the other unit continously. 1.2) Gain evolution and evolution of characteristic time for gain stabilization. The JEM-X microstrip detectors exhibits a gradual increase of the gas gain (at constant voltage) when used. The effect is ascribed to ion- diffusion in the glass substrate leading to a change in the electric field configuration around the anodes. The effect is linear in time, about 1 % increase relative to the initial gain der 2.5 days. For JEM-X1 one the gain has now changed by more than a factor three - assuming that we still operated at the same voltage. However, to guard against the anode erosion problem we have chosen to keep the gain roughly constant by reducing the detector voltage by 10 V at about three month intervals. We see no problems in continuing this procedure over the coming decade, at present the operating voltage is just over 700 V, by 2013 we may be operating around 600 V. When switching on and off the voltage on the microstrip plate during the passage through the radiation belts the detector gain is seen to vary with several timescales: - a '10-minute' timescale - presumably related to charge redistribution on the glass surface, - a '1-hour' timescale where the ion-diffusion pattern in the bulk of the glass substrate changes between a 'virgin'-state and the state to which the plate has evolved during its lifetime of use. The surface charge effect is associated vith a drop in the gain by about 30 %, and the ion-diffusion effect with a gain increase which grows as the plate accumulates more and more lifetime. The first effect was initially the dominant one, and is still very noticeable when the high voltage accidentally is switched off in the middle of an orbit and quickly reestablished by ground intervention. In this case the ion diffusion effect has not had sufficient time to modify the plate gain and as a result we see a strong overshoot of the plate gain if the voltage is returned quickly to its former value. Over time the second effect has evolved to become dominant for switch-off periods of several hours, the typical durations of the radiation belt passages. Following such extended periods of detector inactivity we therefore see a gain undershoot, fully recovering only after several hours. The '1-hour' timescale of the second effect also shows some evolution with time, the detector becomes slower in its recovery to the nominal gain. If this evolution continues at the present rate we can foresee a somewhat longer period in the beginning of each orbit where the JEM-X response in the energy range below 6 keV is sub nominal. We anticipate the following evolution: Duration of sub-nominal low energy response Current: 1 hour 2009: ? 2011: ? 2013: ? We conclude that these effects should only result in a 3% loss of nominal observation time by 2013. 1.3) Detector resolution The gain evolution is not completely uniform across the detector surface so we see a slight degradation of the detectro resolution over time. This effect can be compensated for by deriving a new energy calibration map from the observed position of the Xenon background line at 29.7 keV which is visible in the detector spectrum. At present the resolution of the Xenon line in JEM-X1 is about 10 % compared to 8.5 % immediately after launch. We expect that we can maintain a detector resolution better than 11 % at least until 2013. 1.4) Background count rate evolution. As the Solar activity is reduced the cosmic ray intensity goes up and this leads to an increased trigger rate (hardware triggers) in the INTEGRAL instruments. For JEM-X this has turned out to be a minor effect. Although we have seen an increase in the hardware triggers by 70 % since the beginning of the mission, the rate of triggers accepted by our on-board for transmission to the ground has increased only very slightly (how much?). The rate of anode loss events might also have been expected to increase in proportion to the cosmic ray intensity, on the basis of the available flight experience we can only say that the cosmic ray background has remained near its peak value (seen over the last four solar minima) during the last 900 days where we have experienced 12 anode losses in JEM-X1. We have therefore used this rate for our predictions of the future evolution discussed in section 1.1. We conclude that there is no reason to expect significant changes in the JEM-X background rates over the next decade. 2) JEM-X scientific performance As discussed above the technical performance of JEM-X is not expected to change significantly over the next decade. JEM-X will therefore continue to complement the observations done by the primary INTEGRAL instruments, IBIS and SPI.