A comparison of the FWHM of standard stars observed with VISIR, the mid-IR imager and spectrometer at ESO's VLT, with expectations for the achieved mid-IR Image Quality based on the optical seeing and the wavelength-dependence of atmospheric turbulence, shows that for N-band data (7{12μm), VISIR realizes an image quality about 0.1" worse than expected based on the optical seeing. This difference is large compared to the median N-band image quality of 0.3-0.4" achieved by VISIR. We also note that other mid-IR groundbased imagers show similar image quality in the N-band. We attribute this difference to an under-estimate of the effect of the atmosphere in the mid-IR in the parameters adopted so far for the extrapolation of optical to mid-IR seeing. Adopting an average outer length-scale of the atmospheric turbulence above Paranal L0 = 46 m (instead of the previously used L0 = 23 m) improves the agreement between predicted and achieved image quality in the mid-IR while only having a modest effect on the predicted image quality at shorter wavelengths (although a significant amount of scatter remains, suggesting that l0 may not be constant in time). We therefore advocate adopting L0 = 46 m for the average outer length scale of atmospheric turbulence above Cerro Paranal for real-time scheduling of observations on VLT UT3 (Melipal).
[1]
Marc S. Sarazin,et al.
On the Difference between Seeing and Image Quality: When the Turbulence Outer Scale Enters the Game
,
2010
.
[2]
Andre Wong,et al.
Observing conditions and mid-IR data quality
,
2008,
Astronomical Telescopes + Instrumentation.
[3]
David R. Silva,et al.
Observatory Operations: Strategies, Processes, and Systems II
,
2017
.
[4]
Dimitri Mawet,et al.
VISIR upgrade overview and status
,
2012,
Astronomical Telescopes and Instrumentation.