ESO REVIEW 1997

STEPS TOWARDS PREDICTION OF OBSERVING CONDITIONS

M. Sarazin

• Steps Towards Prediction : several feasibility studies are conducted simultaneously to prepare the operational VLT astroclimate prediction scheme.
• La Silla/Paranal Monitoring : long term monitoring is essential to build a reference database, this requires remotely controllable hardware and a high level of automation. The current Paranal monitoring system is currently being upgraded to the VLT standards to become: the VLT Astronomical Site Monitor.
• Seeing Monitoring: monthly averages Seeing does not present clear climatic cycles but rather accidental highs and lows, with a much larger amplitude at La Silla.
• Cloud Monitoring: monthly averages (filtered by a 5 month moving window) show clear seasonal trends at La Silla and a long term climatic cycle at both sites.
• Cloud Monitoring: differential Paranal-La Silla. Although the global figures did not change compared to 10 years ago, a shorter site survey conducted at a non-representative period (1988-1990) would have underestimated the superiority of Paranal.
• ECMWF Database: Introduction . The database purchased by ESO extends over 5 years (1989-1994) with a resolution of 1/2 degree (60 km) and 15 vertical pressure levels. The area covered is 1800x1800 km (-19 to -35 degree latitude, -65 to -80 longitude) on which analyses and HH+24, HH+48 forecasts of 6 main parameters (Geopotential, Temperature, Vertical Velocity, u-v components of wind, relative humidity) are available. In an operational mode, real time predictions are provided everyday at 12UT for HH+3, +6, +9, +12, +18, +24, etc by further steps of 6h.
• Cloud Cover/ Water Vapour Prediction: digital Meteosat images are combined with ECMWF wind forecasts to estimate next night's cloudiness and precipitable water vapour above the observatory (A. Erasmus, Univ. of Northern Colorado).
• ECMWF Vertical Wind Profiles Prediction. The feasibility study (CRS4, Cagliari, Sardinia) shows that high altitude wind fields forecast by ECMWF are accurate enough for direct use at the observatory: calm nights should be dedicated to observing modes having a high sensitivity to the temporal coherence of wavefront perturbations.
• ECMWF Horizontal Wind Vector: bad seeing case. Several test cases are currently studied for the application of a mesoscale non hydrostatic model in view of seeing predictioni (Nice University+Meteofrance Toulouse)
• ECMWF Vertical Slice of Horizontal Wind: bad seeing case
• ECMWF Horizontal Wind Vector: good seeing case
• ECMWF Vertical Slice of Horizontal Wind: good seeing case

Friday Jan 31 1997