Astroclimatology of ESO sites

Atmospheric Parameters Relevant for Adaptive Optics

Download the presentations (pdf) of the Mini Workshop: Atmosphere Knowledge and Adaptive Optics for 8 to 100 m telescopes (Oct12, 2003)

Isoplanatic angle

The statistics for the years 1999-2003 are obtained from 1mn resolution scintillation measurements built from the statistics on several hundreds individual 5ms DIMM exposures taken in photometric sky. the scintillation data is then passed through a 10mn gaussian filter and converted into isoplanatic angle following the method proposed by Loos and Hogge (Appl. Optics 18, 15, 1979). A correction for finite exposure time is applied following the algorithm computed by A. Tokovinin (Sarazin & Tokovinin, 2001). The horizontal velocity of the scintillation pattern is taken as half the wind speed at 200mB (hypothesis supported by an analysis of dual exposure GSM measurements made at Paranal in December 1998). The average yearly exposure time bias correction was about 30% in 2000 at Paranal. The accuracy of this method is of the order of 10% rms on the isoplanatic angle. Check for last night measurements at Paranal and La Silla.

Coherence Time

The statistics for the years 1999-2003 are obtained from the 10mn averaged DIMM seeing measurements made in photometric sky along tau0=0.31 R0/V0 where V0 is taken as 2/5th of the wind speed at 200mB (supported by balloon flight results at Paranal in 1992 and Pachon in 1998, Sarazin & Tokovinin, 2001). The accuracy of this method is estimated to about 20% rms. Check for last night measurements at Paranal and La Silla. Real time measurements are available at Paranal and at La Silla (internal access only) based on 1mn DIMM seeing measurements.

Atmospheric Time Constants at Paranal during VLTI VINCI & Siderostats Commissioning (pdf int. report). Tau0 time history since VLTI first light up to August 2002, and the slowest wavefront of this period. First comparisons with atmospheric phase fluctuation power spectra obtained with VLTI show a very good correlation (see discussion in the paper SPIE48-35 by Di Folco et al.)

Wind at 200mB

The statistics for years 1999, 2000, 2001, 2002, 2003, 2004 and 2005 are computed from the 00hUT ECMWF analysis and subsequent 06, 12, and 18hUT forecasts after linear rebinning. The ECMWF analyses and forecasts from 1989 to 1993 were compared to Antofagasta radiosoundings at CRS4. They were found out to be accurate within +/-20% (analyses) and +/-40% (forecasts) in more than 70% of the cases in the 100-350mB altitude range. Check for the latest forecast.

Dependency of Isoplanatic Angle and Seeing

As illustrated in this zoom of the Night Aug31 to Sep01, the seeing and the isoplanatic angle are not directly correlated because of the variations in the vertical profile of the turbulence. In this example, the mean altitude (hbar) of the turbulence shows a wavelike structure which strongly influences the isoplanatic angle, while being poorly correlated with the integral seeing.
On the long term however (1999-2003) a relation (one curve per year) appears between seeing and hbar because the very bad seeeing comes generally from the first kilometer and the very good occurs often when only the high layers are turbulent. The global statistics of hbar 1999-2003 show an average optimal conjugation altitude of about 3.5km above Paranal, similar to what has been found above Mauna Kea by Flicker and Rigaut (PASP 114: 1006-1015,2002 Sep.)

Dependency of Isoplanatic Angle and Coherence Time

The isoplanatic angle is relatively dependent on the 200mB wind speed which governs high altitude turbulence. As a consequence, the largest isoplanatic angle values are obtained for the smallest 200mB wind speed; they are however not systematically linked to large coherence times which can be very small during bad seeing, regardless of the velocity. On the other hand, the smallest isoplanatic angles occurr more often with high 200mB wind speeds, which invariably correspond to low coherence times (although very often with good seeing!).

Campaigns

-PARSCA (Paranal Seeing Campaigns, 1992 and 1993) results: Balloon borne microthermal measurements led to estimate the average vertical profile of the turbulenceas well as isoplanatic angle and coherence time above Paranal.
-GSM in 1997 at La Silla (outer scale), 1998 at Paranal (outer scale and ground layer turbulence) and 2007 at Paranal (outer scale and surface layer).
-Balloons launching in 1998 at Cerro Pachon, the site of Gemini South.
-Remote Optical Profilers:
MASS@Tololo started in April 2002
SCIDAR@LaSilla started in July 2002
MASS@Mauna Kea in October 2002
MASS@Pachon in January 2003
MASS@Paranal in March and September 2003.
MASS-LITE@LaSilla started in December 2003.

Modelling

Using PARSCA data along with the output of global models such as ECMWF (European Center for Medium-Range Weather Forecasts) 1993 vertical wind profiles Velocity and Direction at Paranal, we may estimate the average wavefront temporal characteristics (ps file).

More ambitious is the 3D modelling of the turbulence in the atmosphere, using mesoscale non-hydrostatic models Meso-Nh or MM5