Документ взят из кэша поисковой машины. Адрес оригинального документа : http://site2010.sai.msu.ru/static/doc/TTravouillon_TMT_site_testing_site2010.pdf Дата изменения: Mon Oct 18 13:30:52 2010 Дата индексирования: Mon Oct 1 20:10:56 2012 Кодировка: Поисковые слова: arp 220

Lessons learned from the Thirty Meter Telescope site testing
Tony Travouillon Sebastian Els, Angel Otarola, Reed Riddle, Matthias Schoeck, Warren Skidmore

TMT.XXX.PRE.05.0XX.DRF01

1


Introduction TMT and its Site testing. Important choice before going on site. Lesson learned from on-site work. Considerations during analysis of the data.

TMT.XXX.PRE.05.0XX.DRF01

2


Some Background about the TMT Site Testing
TMT is a 30m, segmented, RitcheyChretien telescope design. Field of view of 20' (15' unvignetted) MCAO with 2 deformable mirrors 3 first light instruments

TMT.XXX.PRE.05.0XX.DRF01

3


Some Background about the TMT Site Testing
Site testing started in 2004 and was concluded by the selection of Mauna Kea 13N in 2009 Minimum of 2 years of data at each site. Comprehensive measurement of turbulence, weather conditions, PWV and more. Data available for download at: sitedata.tmt.org
TMT.XXX.PRE.05.0XX.DRF01 4


Instruments & Parameters
Weather stations DIMM ­ seeing monitors MASS ­ turbulence profilers SODAR ­ acoustic sounders IRMA ­ mid-infrared radiometers ASCA ­ Allsky cameras Particle sensors Sonic anemometers Simulations, satellite analysis ­ ­ ­ ­ ­ ­ ­ ­ temp, hum, wind, press, sol.rad, heat flux seeing, coh. time, basic photometry high-el. profiles, isopl. angle, coh. time 20 ­ 800m turb/wind profiles, coh.time PWV, atm. transparency Cloud statistics (incl. cirrus), light pollution Ground level dust particle count 7m wind, temperature, turbulence

­ Turbulence, weather, long baseline

Other considerations: ­ Location, elevation, geology, access, cost of construction and operation, operation model, ...


TMT's candidate sites

Cerro Tolar (Coastal site, Chile) Cerro Armazones (Coastal site, Chile) Cerro Tolonchar (Isolated mountain, Chile) San Pedro Martir (Coastal site, Baja California) Mauna Kea 13N (Volcanic Island, Hawaii)

TMT.XXX.PRE.05.0XX.DRF01

6


Chile Sites - Locations

Tolar
La Silla Las Campanas AURA

ALMA
Santiago

100 km
Paranal
7

Tolonchar

Armazones


Tolar


Armazones


Tolonchar


San Pedro MАrtir


13 North Site

Mauna Kea 13N


Early choices 1: Strong pre-selection
Satellite study commissioned to narrow down the study to 5 sites All sites confirmed by in-situ study to be good site After confirmation from the ground, data useful for longer time scale study.
Clear (photometric) fraction for the observing night

A Satellite Survey of Cloud Cover and Water Vapor in the Southwesthern U.S.A. and Northern Mexico D Andre Erasmus, Ph.D.
Certified Consulting Meteorologist

TMT.XXX.PRE.05.0XX.DRF01

13


Early choices 2: Not an Engineering project.

"Buy, don't build" philosophy Focus on gathering data and build up statistics Verification of instrument specifications done in the field.

TMT.XXX.PRE.05.0XX.DRF01

14


Early choices 3: Standardization

TMT.XXX.PRE.05.0XX.DRF01

15


Early choices 4: Redundancy
Spares of all instrumentation: Allows for quick replacement and minimizes down time. Instrumentation with overlapping use: Opportunity to cross check data and complete knowledge gaps. Applies to people too: 2 people with expertise of each instrument and analysis.

TMT.XXX.PRE.05.0XX.DRF01

16


Spend time on a cross-calibration campaign

TMT.XXX.PRE.05.0XX.DRF01

17


Cross calibrate everything

TMT.XXX.PRE.05.0XX.DRF01

18


Simultaneity of measurements

Multi-year campaign to avoid measuring in atypical weather conditions (el nino vs la nina, Bolivian winter, etc..) Maximize overlapping time of campaigns between sites (understand local effects and removes potential trends)

TMT.XXX.PRE.05.0XX.DRF01

19


Telescope choice

Open dome Design: minimize local turbulence Open tube: No tube seeing, smaller footprint Robust telescope: Observe in most wind conditions Stiff tower: Keep vibrations to a minimum Active focus: Allows better and remote alignment checks
TMT.XXX.PRE.05.0XX.DRF01 20


Hardware: Telescope choice

Open dome Design: minimize local turbulence Open tube: No tube seeing, smaller footprint Robust telescope: Observe in most wind conditions Stiff tower: Keep vibrations to a minimum Active focus: Allows better and remote alignment checks
TMT.XXX.PRE.05.0XX.DRF01 21


Hardware: Telescope choice

Open dome Design: minimize local turbulence Open tube: No tube seeing, smaller footprint Robust telescope: Observe in most wind conditions Stiff tower: Keep vibrations to a minimum Active focus: Allows better and remote alignment checks
TMT.XXX.PRE.05.0XX.DRF01 22


Hardware: Wind speed sensors

TMT.XXX.PRE.05.0XX.DRF01

23


Software
Fully automatic, robotic system, but constantly connected via Internet. Full manual remote operation and a lot of remote troubleshooting and control are possible

Example of the Site Status Overview web page


Software management and data access
Software updated on the central server and pushed to all sites to avoid discrepancies between sites Daily backup of data and logs to central server and backup server Database accessible online, formatted for quick re- analysis

Site 1 Central server Backup server
TMT.XXX.PRE.05.0XX.DRF01

Site 2 Site 3 Etc...
25


Analysis coherence and verification

Analysis routine started early in the program. Each analysis done by two people. Allows verification of code and results but also highlighted different methods for the analysis and filtering of the data. Reports written with Matlab...

TMT.XXX.PRE.05.0XX.DRF01

26


Cloud Cover Definition
First idea: Use geometrical cloud coverage, averaged in both space and time. Problem: Not a useful quantity for TMT as illustrated using the following examples (see also the images on the next slides) ­ Sky entirely covered first half of the night, entirely clear second half: 50% coverage ­ Bands of clouds moving through every 15 minutes, with equally long clear gaps in between: 50% coverage ­ Small broken clouds covering half of the sky at any time during the night: 50% coverage All these cases give the same geometrical coverage, but are very different in their usefulness for TMT. This is not sufficient. Need other definition of cloud cover.


Cloud Cover Definition
Geometrical coverage is not a useful definition. Other possibilities: Cut night into one (half, two, ...?) hour segments. Class segment as: ­ Photometric: no clouds within 65o of zenith during entire hour ­ Spectroscopic: thin or broken clouds present ­ Unusable: continuous thick clouds Define a "cloud coherence time": ­ Average time during which conditions remain constant ­ Problem: obtaining a quantitative estimate Others?


ASCA Cloud Cover Analysis
So how do we do it in practice? Options: Visual analysis of ASCA images/movies to get cloud cover ­ Pros: "Simple" method; easy to implement ­ Cons: How quantitative is it? Somebody has to watch all the movies Automated photometry software package ­ Pros: Quantitative analysis of cloud cover, time constant, transparency High time and spatial resolution opacity measurements ­ Cons: Huge software development effort needed (not possible for us and LSST effort will probably not be useful for TMT)


Fin...

TMT.XXX.PRE.05.0XX.DRF01

30


Acknowledgement
The TMT Project gratefully acknowledges the support of the TMT partner institutions. They are the Association of Canadian Universities for Research in Astronomy (ACURA), the California Institute of Technology and the University of California. This work was supported as well by the Gordon and Betty Moore Foundation, the Canada Foundation for Innovation, the Ontario Ministry of Research and Innovation, the National Research Council of Canada, the Natural Sciences and Engineering Research Council of Canada, the British Columbia Knowledge Development Fund, the Association of Universities for Research in Astronomy (AURA) and the U.S. National Science Foundation.

TMT.XXX.PRE.05.0XX.DRF01

31