1. (a) (b) (c) DETX DETY DETX DETY XIS2 DETX DETY XIS0 XIS2 Counts keV -1 ACTY N. Tawa, M. Nagai, K. Hayasida, H. Nakamoto, M. Namiki (Osaka-U), H. Yamaguchi (Kyoto-U), E. Miller (MIT), Y. Takei (ISAS/JAXA), T. Mizuno (Hiroshima-U), H. Katayama (JAXA) and the Suzaku Team XIS Non-X-ray BGD Database and Application to Detecting Faint Diffuse X-ray Emissions Abstract One of the advantages the Suzaku XIS(X-ray Imaging Spectrometers) is their low background level. To maximize this advantage, we study the Non-X-ray Background (NXB) in XIS, using the night Earth data. Many fluorescent lines, Al, Si, Au, Ni are observed in the NXB spectra. Positional dependence of the NXB level with ACTY is found. Variations of factor larger than 2 is primarily correlates with the Cut-Off-Rigidity (COR) in the satellite orbit. The NXB database of about 800ks is open to observers. The database can be used as any kind of NXB modeling and subtraction, but we first employ the simplest model using the COR as a parameter. The reproducibility of this NXB model is examined with the same NXB database, dividing for each 5ks exposure chunk. We find large residual to this model in some occasions. Those deviation is removed by imposing a filter with latitude and altitude. The reproducibility of the current NXB model is 6.8-11.6% in 5-12keV band for 5ks exposure. Further improvement is expected. We also show the application of this NXB model to an analysis of cluster outskirt. １． NXB characteristic 5.0-14.9keV 0.1-5.0keV Count rate COR Counting rate of the XIS1 during a blank sky observation on 2005 September 2nd. Reproducibility of the NXB model with COR c/s in 5-12 keV We made the 1st version of the NXB database consisting of 1) large screened event file for each sensor, 2) a EHK file that has orbital and other useful information as a function of time. The following data selection criteria was used. The data base with associated tools are accessible via Suzaku web page at ISAS/JAXA and GSFC/NASA. ACTY dependence 7.2-7.8keV (Ni-K  ) 3.0-7.0keV (Continuum) ACTY dependence Mn-K  band (5.7-6.0keV) image 55Fe source area Cut-off rigidity dependence Count rate in 5-10keV （ XIS0,1,2,3 ） vs COR XIS0 XIS1 XIS2 XIS3 If we apply the same procedure to a portion of the NXB database, the reproducibility of the NXB model (the simplest version) can be examined. Here we divide the database for each 5ks. Note that 5ks NXB data spans for typically one day or a few days’ duration. ３． NXB database reproducibility The (COR sort) NXB model reproduce the NXB level usually, but it is found that large deviations in some 5ks exposures. We examine various filters (filtering is done by each 32s not by each 5ks) using orbital parameters. It is found that filtering by the satellite latitude and altitude is effective. Al-K  Si-K  Au-M  Mn-K  Mn-K  Ni-K  Ni-K  Au-L  Au-L  XIS0 XIS2 XIS1 XIS3 NXB spectra T total ： total exposure time of target observation 5ks exposure no. in time order 0.04 0.02 0 -0.02 XIS0 ２． NXB database Standard deviation of the counting rate Systematic error of the NXB model with COR Average statistical error for 5ks data No filterOrbit filter XIS019.5%11.2%7.0%5.8 % (6.0%) XIS122.1%8.3%6.8%2.5% (2.6%) XIS218.6%11.6% 5.9% (6.1%) XIS316.9%8.4%7.6%6.3% (6.5%) ４． Application to Cluster outskirt The NXB database is a simple event file and flexible use is encouraged. However, for the 1st step, the COR dependence is good to be considered. Thus we prepare a tool to 1) sort the NXB events by the COR value, 2) make the NXB spectra for each COR value, 3) and then make appropriate NXB spectra for a target observation by summing up the CPR sorted spectra with appropriate weights. Schematically the procedure is expressed with the follow equation, where the COR is sorted into n bins. Sw is a kind of NXB models, for which we examine reproducibility. T i ： exposure time of the target observation with ith COR bin S i ： spectrum made of the NXB database in the ith COR bin S w ： appropriately weighted NXB spectrum for this target XIS0 0.04 0.02 0 -0.02 5ks exposure no. in time order 0.04 0.02 0 -0.02 XIS0 Latitude. ≤ -23 (deg) & Altitude. ≥ 576.5 (km) and Latitude. ≥ 29 (deg) & Altitude. ≥ 577.5 (km) Source : A3376_WEST_RELIC Sensor : XIS-0 + XIS-2 + XIS-3 Night Earth observation data [ELV 100°] Total exposure time : 799ks (786ks) Rev0.7 (Rev0.6) ＜ Extraction ＞ Total observation data 2005/09/01 ~ 2006/05/21 (2006/03/the last day) NXB database （ In Night Earth, neither ② nor ③ contribute. ） 0.5 1 2 5 Energy (keV) Distance from center of A3376: 1.02Mpc (b) Energy (keV) : 1.37Mpc 0.5 1 2 5 (c) Energy (keV) : 1.72Mpc 0.5 1 2 5 Model : wabs * ( apec + powerlaw )+apec CXB Galactic&local data model A3376 CXB Galactic 10 -5 10 -4 10 -3 10 -2 0.1 Normalized counts s -1 keV -1 Region → (a) 〃 0.3+/-0.83.8+/-1.5(c) 〃 0.18+/-0.163.5+/-0.3(b) 0.259+/-0.0120.24+/-0.073.91+/-0.13(a) Temperature (keV)AbundanceTemperature (keV) Galactic&localA3376 Low background level in XIS Scattered Mn K in XIS0 Background (NXB+CXB in this plot) level of X- ray CCDs in various mission is compared in terms of the N/S ratio to diffuse emissions. NXB spectra of XIS show fluorescece lines of various elements used in the camera body. The NXB level is higher at larger ACTY, for which time spent on the frame store region is longer. The scattered component of Mn K from calibration sources is not negligible in XIS0. The NXB level varies by factor of more than 2. The primary parameter is the Cut- Off-Rigidity (COR) in the satellite orbit. Other Components in the sky XIS background The XIS sky background consists of the NXB, the extragalactic Cosmic X-ray Background (NXB), and local & galactic diffuse X-rays depending on the sky position. In addition, when the telescope FOV is close to the limb of the shining Earth, atmospheric fluorescence lines of N-K and O-K are heavily contaminated. (Left) Counting rate averaged for each 5ks (right) Residual to the NXB model described above c/s in 5-12 keV NXB model with COR longitude latitude Altitude(km) Excluding the data with the left condition (we here call “orbit filter”) makes the residual clean. The NXB model with COR clearly reduce the reproducibility from ~20% to 8-12% (of the average NXB level in the same band). The orbital filter is effective not only to exclude extraordinary deviation but also to reduce the reproducibility (except for XIS2). The reproducibility of the NXB model with COR is evaluated to be 6.8-11.6% in 5-12keV for 5ks exposure data. For larger exposure, we get smaller systematic errors. The NXB model with COR is the simplest one, and should be improved further employing more parameters. The total exposure of NXB database slightly decreases when orbit filter is applied(800ks→725ks). *) The values are relative fraction to average NXB level in the same band.