Subject: Some plots on position accuracy Dear all, motivated by Niels Joergen's visit, I have taken another look at position uncertainties for the Crab and Vela X-1. Attached you can find: * CrabPos.pdf Offset in Crab position (using ADS pos. as baseline) as function of off-axis angle (JMX1 = blue diamonds, JMX2 = green boxes). While JMX1 can be almost spot-on, JMX2 has a basic offset of ~10" but from a glance less of an angle dependence - curious. It is not clear to me what caused the two 'outliers' in JMX1 results. * VelaPos.pdf A similar plot based on the Vela X-1 observations last summer (JMX2 only). While the encompassing range rises with off-axis angle, the majority of points remains between 10" and 20" off the true source position. * Offset_vs_Detsig.pdf Crab (JMX1+JMX2 with same symbols as above) and Vela X-1 data thrown together (JMX2, red triangles). This plot of the dependence of the position error on DETSIG shows even clearer the basic offset found for JMX2. => NJW you should consider scaling your position errors on DETSIG. * CrabPosHisto.pdf Histograms of the position error distribution as function of angle (left JMX1, right JMX2). The sampling is far from even across angles but one sees the trends better, in my opinion. * VelaPosHisto.pdf Dito for the Vela X-1 data, where the sampling is very uneven. I could redo this with other sampling but it's a bit of work so I would prefer to have good guidelines first. Should someone want the data or the PS files, just let me know. We should also see if some of these plots replacee some existing plots in the SVR. Cheers, Peter Delivered-To: oxborrow@dsri.dk Date: Thu, 17 Jun 2004 13:50:26 +0200 (MEST) From: Niels Joergen Westergaard To: Peter Kretschmar Cc: Active SDAST -- Carol Anne Oxborrow , Jérôme Chenevez , "Niels J. Westergaard" , Peter Kretschmar , Sami Maisala , Silvia Martínez Núñez , Soeren Brandt , Stefan Larsson Subject: Re: Some plots on position accuracy MIME-Version: 1.0 Hello Peter, I'm back. Thanks for your nice plots. The Crab offset for high DETSIG for JMX2 is that a misalignment matrix issue ? I actually do scale the position error with DETSIG. I'll check if how good my formula is compared with these results and I'll see if the off-axis angle also should be a parameter. Regards Niels Joergen From: Peter Kretschmar Subject: Re: Some plots on position accuracy To: Niels Joergen Westergaard Cc: Active SDAST -- Carol Anne Oxborrow , Jérôme Chenevez , Sami Maisala , Silvia Martínez Núñez , Soeren Brandt , Stefan Larsson MIME-version: 1.0 Content-transfer-encoding: 7BIT X-Accept-Language: en-us, en User-Agent: Mozilla/5.0 (X11; U; SunOS sun4u; en-US; rv:1.0.1) Gecko/20020920 Netscape/7.0 X-Comment: This message was scanned against viruses by mbx.unige.ch. Hi Niels Joergen - hope you had a good trip back. > The Crab offset for high DETSIG for JMX2 is that a misalignment matrix > issue ? > I would guess so. But the JMX1 results also make me wonder a bit since the really close matches (<5" off) are all from revolutions 39-42. Well, I juts checked and besides that knot of close values I don't see a real distribution difference between early and late revolutions. > I actually do scale the position error with DETSIG. I'll check if how good > my formula is compared with these results and I'll see if the off-axis > angle also should be a parameter. Yes, but you take evidently 60" as lower limit and I would say that is a bit too conservative. Though it's not evident to me what we should use. The formal stddev for Vela X-1 at 4-5 degrees (DETSIG not much >20) is ~30" but by eye I would rather put 60" as an estimate for that range. Btw, you can find copies of the PS files and the FITS files I've compiled up to now (jmx[12]_crab.fits, vela_summer.fits) in ~pkretsch/tmp/SVR. Cheers, Peter