Suzaku Help Jp: Question #0057

Question ID = 0057

Question (質問)

I am comparing this dataset with others taken from Swift, Chandra and XMM within few days from the Suzaku pointing . I have a problem concerning the absolute timing accuracy.

I rely upon a coherent timing solution (P and Pdot) among Swift/Chandra/XMM. However, the Suzaku data are the only ones which stand apart form the coherent solution of about 0.4 in phase, or approx 4.2s ....

I was wandering whether there is any problem in the absolute timing accuracy of this Suzaku pointing or more in general with Suzaku. If so, do you have any recipe to account for this problem.

Answer from(回答の文責)

Y.Ishisaki(Suzaku timing), T.Dotani (XIS timing), Y.Terada (HXD timing) and the HelpDesk members.

Answer (答え)

(Q1) see above.

If you are on the XIS data, plase visit this web page. If you are analyzing HXD data, please read following paper. Fig.38 and 39 are related on the topic. M. Kokubun et al., Publ. Astron. Soc. Japan, 59, S53-76 (2007)

(Q2) I was wandering whether the XIS correction to apply is exactly 7 and 6s , or approximately 7 and 6s. In other words, how much accurate is the correction ? Moreover, is the time-shift constant through all the SUZAKU datasets.

The timing is exactly 7 and 6 seconds. Accuracy of the timing is about 1 milliseconds in design.
Yes, it is. If the process version is the same, you do not take care of the time-shift constant through the all datasets.

(Q3) The second question is referring to the way of correctly compare the SUZAKU times with those of Chandra and XMM. If I understood well SUZAKU uses a UTC times, while Chandra and XMM uses the TT times. Do you know which is the time shift I need to apply to the SUZAKU times in order to "correct" and to compare them with XMM/Chandra ? Do you have a tool for that ?

Suzaku uses the TT time, too. You can use the "aebarycen" tool in HEASOFT when you apply barycenter correction to the Suzaku data. For detail about the tool, please type:

(Q4) All the datasets are in TT. However, when I look at the difference between TT and UTC of the XMM and SUZAKU reference MJD (see below) I found one missing second. Does this affect the comparison ?
SUZAKU
51544.0007428703703703700 TT = 2000-01-01 00:00:00 UTC
51544.0007428703703703700 UTC = 2000-01-01 00:01:04 UTC
64s difference
XMM
50814.0000000000000000000 TT = 1997-12-31 23:58:57 UTC
50814.0000000000000000000 UTC = 1998-01-01 00:00:00 UTC
63s difference

There was a leap second at 1999-01-01 00:00:00. That is why you had missing one second. There were other leap seconds at 1997-07-01 00:00:00 and 2006-01-01 00:00:00. You can fine the data in:
${CALDB}/data/gen/bcf/leapsec_010905.fits

(Q5) Moreover, should I worry by the fact that SUZAKU is starting from TT 00:00:00 while XMM starts 63s in advance with respect to the TT 00:00:00 ?

Please apply the time shift "before" the barycentric correction, because the shift is due to some inconsistency on the early stage of time assignment between the design of on board electronics and the analysis software.

Status （詳細なステータス）

2007-06-12 Accept
2007-06-12 Asgn.
2007-06-12 Cont.
2007-06-26 Done.
2007-07-10 update/done.

RIKEN Makishima Cosmic Radiation Laboratory

Assigned Member (担当者)

Name and Institute (質問者/機関)

Question (質問)

Answer from(回答の文責)

Answer (答え)

Status （詳細なステータス）