2011ATel.3168....1R

After the report of renewed activity of IGR J17091-3624 as seen with Swift (ATel #3144, #3148, #3167), with the confirmation of the nearby

source IGR J17098-3628 being dormant, we observed this region with RXTE.

4 pointings have been performed on Feb 9, 10, 12, and 14, 2011.

Inspection of the PCA spectra show some contamination by the nearby (0.67 deg away) X-ray binary GX 349+2. We roughly estimate the contribution of this source

to be about 287 cts/s over the ∼2-60 keV PCA range. We assume a constant rate over all our observations.

The net ∼2-60 keV count rate of IGR J17091-3624 increased from ∼82cts/s to ∼278cts/s, but then decreased to ∼121 cts/s from the first to the third obs. In the last observation it increased again to ∼308 cts/s.

We analysed the Good Xenon data and extracted ∼2-60 keV light curves with time resolution of ∼1ms (2^(-10) s). In the first observation the power density spectrum (PDS) is consistent with Poisson noise above ∼1Hz. It shows large variability below this frequency. A single

zero-centred Lorentzian is sufficient to account for the continuum. The total amplitude is estimated to be 26% although the uncertainties here

are large.

In the 2nd observation the PDS is consistent with Poisson noise above 10 Hz. The 0.01-10 Hz PDS is well represented (red. chi square=1.21 for 159 dof)

by the sum of 3 Lorentzians, all zero centred, with a total power of about 15% RMS over this frequency range. Large residuals are, however, present at around 0.1 Hz. Adding another Lorentzian improves the fit to 1.02 for 156 dof (chance probability 1.7e-06 according to an F-test). The last feature has a frequency of 9.4±0.3 e-2 Hz, a quality factor Q∼9.1, and an RMS amplitude of 3.2±0.2%.

In the third observation, the PDS is consistent with Poisson noise above typically 5 Hz. Two broad Lorentzians are sufficient

to represent the continuum. The total RMS amplitude is ∼15%. A thin feature is also present in the residuals. A Lorentzian with freq=0.105 -0.003 +0.004 Hz, Q=7.4 ,A=3.4 (-0.3 +0.2)%

improves the fit (chance probability of 2.3e-4 according to an F-test).

The PDS of the 4th observation is consistent with Poisson noise above

1 Hz (the observation is much shorter). Below 1 Hz the continuum is well

represented by the sum of 2 zero-centred Lorentzians with a total RMS amplitude 12.3%. The fit also requires the addition of a QPO for a

good fit to be achieved (chance probability of 2.2e-4 according to an F-test).

The QPO has the following parameters: freq=0.13±0.01, Q=3.5, and

We note that the uncertainties quoted do not take into account the uncertainties

in the estimate of the flux from GX349+2. A ∼0.1 Hz feature has never been reported for GX 349+2 despite the huge number of observations dedicated to this source. We, therefore, attribute the QPO to IGR J17091-3624.

This is the first report of a low frequency QPO in this source. This and

the increasing frequency in correlation with the increase of flux, is consistent

with a black hole binary in its hard state. More observations of this source in its outburst are encouraged.