Upload File

an atmospheric modeling tool for site characterization and telescope calibration using ... · 2017....

  • Home
  • Documents
  • An Atmospheric Modeling Tool for Site Characterization and Telescope Calibration using ... · 2017. 8. 24. · An Atmospheric Modeling Tool for Site Characterization and Telescope
1
An Atmospheric Modeling Tool for Site Characterization and Telescope Calibration using MERRA2 Andrew Wang, Denis Barkats, Scott Paine, Simon Radford, John Kovac Abstract Using NASA’s MERRA2 atmospheric reanalysis and the am radia7ve transfer model, we have developed a new tool that predicts the sky brightness temperature and opacity anywhere on Earth for any date from 1970 to the present. We validate MERRA2 by comparing our predic7ons with long- baseline opacity measurement campaigns and quan7fy the accuracy of our predic7on tool. Finally, we have created a command line tool and web interface for researchers to generate predic7ons independently. Introduction Method Results 1 M. M. Rienecker, et al. 2011 “MERRA, NASA's Modern-Era Retrospective Analysis for Research and Applications.” J. Climate 24:3624. , https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/ 2 Paine, Scott. (2017, March 29). The am atmospheric model (v. 9.2). Zenodo. http://doi.org/10.5281/zenodo.438726 3 Radford & Peterson, 2016, arXiv:1602.08795 4 Matsushita et al., 2017, PASP, 129, 025001 5 C-L. Kuo, arXiv: 1707.08400 6 NOAA/NCEP Global Forecast System (GFS) Atmospheric Model Relevance to CMB-S4 • Site Characteriza1on: Our tool can provide predic7ons of any poten7al site’s opacity or brightness temperature to determine viability as a future observing loca7on. • Recent work has been done to compare exis7ng CMB observing loca7ons with new sites, including Summit Sta7on 4 in Greenland, Ali 5 Observatory in Tibet, and Dome A in Antarc7ca. • Telescope Calibra1on: Zenith brightness temperature 7mestream can be used to calibrate or es7mate real-7me sensi7vity at current CMB observatories (Figure 5). References How accurate are the MERRA2-based predic1ons? • We compare our MERRA2 sky brightness temperature predic7ons to 850 GHz 7pper measurements taken from 1998-2017 at South Pole (Fig 2), Chajnantor Plateau, Cerro Chajnantor, and Mauna Kea. • The median 7pper uncertainty is 7-10 K (5-6%) at all sites, and remains below 10% over 93% of the 7me at Mauna Kea and South Pole and over 88% of the 7me at the Chajnantor sites. • We assign the MERRA2 predic7on an uncertainty rela7ve to the 7pper’s known uncertainty. Modeling the residual’s variance as the sum of the 7pper σ t 2 and the MERRA2 σ m 2 variances, calcula7ng chi-squared values of the residuals allows us to solve for the unknown σ m (Fig 3): χ 2 = (7pper measurement - MERRA2 model predic7on) 2 / (σ t 2 +σ m 2 ) • If MERRA2 were a perfect model, we would expect χ 2 = 1 (i.e. σ m = 0). For all sites, we find χ 2 =2–6 which suggests the MERRA2 model has between 1 and 2.3 7mes the 7pper uncertainty (Fig 4). • At South Pole and Mauna Kea, the MERRA2 predic7ons are remarkably close to the measured brightness temperature. Predic7ons at the Chajnantor sites are slightly less accurate, but s7ll within 10-15% of the measured value. This is expected given that the complex local weather pajern in the mountainous Chajnantor region is harder to spa7ally resolve accurately on MERRA2’s 50x50 km grid. • Ground-based CMB telescopes observe through the atmosphere, which absorbs a significant frac7on of the mm/ sub-mm cosmic signal and adds thermal noise. • Atmospheric turbulence also adds noise fluctua7ons to incoming radia7on. • We determine the atmospheric state using MERRA2. MERRA2’s focus on the atmospheric water vapor is ideal for this applica7on. • We validate MERRA2’s predic7ons by quan7fying its uncertain7es against long-baseline historical 850 GHz and 225 GHz opacity 7pper 3 measurements. • Interpolate 3-hour MERRA2 1 data products (e.g. ver7cal profiles of pressure, temperature, water mixing ra7o, and other relevant quan77es) to any posi7on on Earth. • Use the 2D MERRA2 products to obtain ground values of the atmospheric profiles. • Run the am radia7ve transfer model 2 on these profiles to generate frequency-dependent atmospheric transmission spectra, along with PWV value. • Integrate these spectra over an instrument bandpass to predict the sky brightness temperature for that instrument. Publicly Available Tool • Web Interface (see below) and command line tool currently in development; beta version at h$ps://goo.gl/czB24x • User inputs a loca7on, date range, and bandpass, and receives an output file and plots with the brightness temperature and PWV predic7ons at 3-hour intervals. • Future Work: We are adding the capability to make predic7ons for future dates using GFS 6 forecasts, to upload custom bandpass files. We are also working on a paper describing the detailed methods and valida7on of this tool. Figure 1: Transmission spectra of the atmosphere with varying levels of precipitable water vapor (PWV). Transmission drops significantly as atmospheric PWV increases, which is why we need to iden7fy low-PWV sites. Figure 5: Brightness temperature predic7ons at South Pole in 2015 for the BICEP/Keck Array 100/150/220/270GHz bands. Figure 2 Le#: Time-series comparison of MERRA2 brightness temperature predic7ons to real 7pper data every 3 hours from February to April 2015 at South Pole. Right: One-to-one comparison and best-fit line for a full year. Figure 3: Residuals for Fig 2-len at South Pole from February to April 2015, with a χ 2 of 3. Figure 4 (below):χ 2 values for the above residuals at each of the sites. A median χ 2 of 2.9 (as in SP and MK) means the standard devia7on of the MERRA2 predic7on is 1.4 7mes that of the 7pper. A χ 2 of 6 (as in Chajnantor) implies a predic7on uncertainty 2.3 7mes the 7pper’s uncertainty.

Upload: others

Post on 20-Feb-2021

5 views

Category:

Documents


0 download

Report

TRANSCRIPT

  • An Atmospheric Modeling Tool for Site Characterization and Telescope Calibration using MERRA2

    Andrew Wang, Denis Barkats, Scott Paine, Simon Radford, John Kovac

    Abstract

    UsingNASA’sMERRA2atmosphericreanalysisandtheamradia7vetransfermodel,wehavedevelopedanewtoolthatpredictstheskybrightnesstemperatureandopacityanywhereonEarthforanydatefrom1970tothepresent.WevalidateMERRA2bycomparingourpredic7onswithlong-baselineopacitymeasurementcampaignsandquan7fytheaccuracyofourpredic7ontool.Finally,wehavecreatedacommandlinetoolandwebinterfaceforresearcherstogeneratepredic7onsindependently.

    Introduction

    Method

    Results

    1  M. M. Rienecker, et al. 2011 “MERRA, NASA's Modern-Era Retrospective Analysis for Research and Applications.” J. Climate 24:3624. , https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/

    2  Paine, Scott. (2017, March 29). The am atmospheric model (v. 9.2). Zenodo. http://doi.org/10.5281/zenodo.438726

    3  Radford & Peterson, 2016, arXiv:1602.08795 4  Matsushita et al., 2017, PASP, 129, 025001 5  C-L. Kuo, arXiv: 1707.08400 6  NOAA/NCEP Global Forecast System (GFS) Atmospheric Model

    Relevance to CMB-S4

    •  SiteCharacteriza1on:Ourtoolcanprovidepredic7onsofanypoten7alsite’sopacityorbrightnesstemperaturetodetermineviabilityasafutureobservingloca7on.•  Recentworkhasbeendonetocompareexis7ngCMB

    observingloca7onswithnewsites,includingSummitSta7on4inGreenland,Ali5ObservatoryinTibet,andDomeAinAntarc7ca.

    •  TelescopeCalibra1on:Zenithbrightnesstemperature7mestreamcanbeusedtocalibrateores7matereal-7mesensi7vityatcurrentCMBobservatories(Figure5).

    References

    HowaccuratearetheMERRA2-basedpredic1ons?•  WecompareourMERRA2skybrightnesstemperaturepredic7onsto850GHz7ppermeasurements

    takenfrom1998-2017atSouthPole(Fig2),ChajnantorPlateau,CerroChajnantor,andMaunaKea.

    •  Themedian7pperuncertaintyis7-10K(5-6%)atallsites,andremainsbelow10%over93%ofthe7meatMaunaKeaandSouthPoleandover88%ofthe7meattheChajnantorsites.

    •  WeassigntheMERRA2predic7onanuncertaintyrela7vetothe7pper’sknownuncertainty.Modelingtheresidual’svarianceasthesumofthe7pperσt2andtheMERRA2σm2variances,calcula7ngchi-squaredvaluesoftheresidualsallowsustosolvefortheunknownσm(Fig3):χ2=(7ppermeasurement-MERRA2modelpredic7on)2/(σt2+σm2)

    •  IfMERRA2wereaperfectmodel,wewouldexpectχ2=1(i.e.σm=0).Forallsites,wefindχ2=2–6whichsuggeststheMERRA2modelhasbetween1and2.37mesthe7pperuncertainty(Fig4).

    •  AtSouthPoleandMaunaKea,theMERRA2predic7onsareremarkablyclosetothemeasuredbrightnesstemperature.Predic7onsattheChajnantorsitesareslightlylessaccurate,buts7llwithin10-15%ofthemeasuredvalue.ThisisexpectedgiventhatthecomplexlocalweatherpajerninthemountainousChajnantorregionishardertospa7allyresolveaccuratelyonMERRA2’s50x50kmgrid.

    •  Ground-basedCMBtelescopesobservethroughtheatmosphere,whichabsorbsasignificantfrac7onofthemm/sub-mmcosmicsignalandaddsthermalnoise.

    •  Atmosphericturbulencealsoaddsnoisefluctua7onstoincomingradia7on.

    •  WedeterminetheatmosphericstateusingMERRA2.MERRA2’sfocusontheatmosphericwatervaporisidealforthisapplica7on.

    •  WevalidateMERRA2’spredic7onsbyquan7fyingitsuncertain7esagainstlong-baselinehistorical850GHzand225GHzopacity7pper3measurements.

    •  Interpolate3-hourMERRA21dataproducts(e.g.ver7calprofilesofpressure,temperature,watermixingra7o,andotherrelevantquan77es)toanyposi7ononEarth.

    •  Usethe2DMERRA2productstoobtaingroundvaluesoftheatmosphericprofiles.

    •  Runtheamradia7vetransfermodel2ontheseprofilestogeneratefrequency-dependentatmospherictransmissionspectra,alongwithPWVvalue.

    •  Integratethesespectraoveraninstrumentbandpasstopredicttheskybrightnesstemperatureforthatinstrument.

    Publicly Available Tool

    •  WebInterface(seebelow)andcommandlinetoolcurrentlyindevelopment;betaversionath$ps://goo.gl/czB24x

    •  Userinputsaloca7on,daterange,andbandpass,andreceivesanoutputfileandplotswiththebrightnesstemperatureandPWVpredic7onsat3-hourintervals.

    •  FutureWork:Weareaddingthecapabilitytomakepredic7onsforfuturedatesusingGFS6forecasts,touploadcustombandpassfiles.Wearealsoworkingonapaperdescribingthedetailedmethodsandvalida7onofthistool.

    Figure1:Transmissionspectraoftheatmospherewithvaryinglevelsofprecipitablewatervapor(PWV).TransmissiondropssignificantlyasatmosphericPWVincreases,whichiswhyweneedtoiden7fylow-PWVsites.

    Figure5:Brightnesstemperaturepredic7onsatSouthPolein2015fortheBICEP/KeckArray100/150/220/270GHzbands.

    Figure2Le#:Time-seriescomparisonofMERRA2brightnesstemperaturepredic7onstoreal7pperdataevery3hoursfromFebruarytoApril2015atSouthPole.Right:One-to-onecomparisonandbest-fitlineforafullyear.

    Figure3:ResidualsforFig2-lenatSouthPolefromFebruarytoApril2015,withaχ2of3.

    Figure4(below):χ2valuesfortheaboveresidualsateachofthesites.Amedianχ2of2.9(asinSPandMK)meansthestandarddevia7onoftheMERRA2predic7onis1.47mesthatofthe7pper.Aχ2of6(asinChajnantor)impliesapredic7onuncertainty2.37mesthe7pper’suncertainty.


Exoplanet Characterization with JWST Jeff Valenti (Space Telescope Science Institute)

MARITIME ATMOSPHERIC CHARACTERIZATION SYSTEM (MACS

Characterization of the optical properties of atmospheric aerosols in

Experimental characterization of marine atmospheric ... · Experimental characterization of marine atmospheric boundary layer in the Havsul area, Norway 10th Deep Sea Offshore Wind

Some Crucial Aspects of Atmospheric Dynamics · • Atmospheric dynamics modeling REQUIRED for characterization (e.g., T-P profiles needed over the globe). Good atmospheric dynamics

HOSTED BY Atmospheric Pollution Research · Chemical characterization of atmospheric particulate matter and their source apportionment at an emerging industrial coastal city,

The atmospheric transparency measured with a …fukushim/WEB/pap/tomida_NIM_lidar.pdf · The atmospheric transparency measured with a LIDAR system at the Telescope Array experiment

The University of Tokyo Atacama 1.0-m Telescope · The University of Tokyo Atacama 1.0m telescope is a prior project of ... high-transparent atmospheric environment, several radio

Molecular characterization of atmospheric NO -responsive

atmospheric Cherenkov telescope in Romania

Aviation Safety Program Atmospheric Environment … Safety Program Atmospheric Environment Safety Technologies (AEST) Project Summary Engine Icing: Characterization and Simulation

James Webb Space Telescope : Characterization of Flight Candidate of Raytheon NIR InSb Arrays

MIRACLES: atmospheric characterization of directly imaged ...Astronomy & Astrophysics A&A 644, A13 (2020) © ESO 2020 MIRACLES: atmospheric characterization of directly imaged planets

Characterization of atmospheric trace gas and aerosol concentrations at forest sites ... · 2010-06-22 · T. 2000. Characterization of atmospheric trace gas and aerosol concentrations

Atmospheric Surface Layer Characterization: Preliminary

Generation, Characterization and Application of ...journalarticle.ukm.my/10389/1/15 R. Shrestha.pdf · Generation, Characterization and Application of Atmospheric Pressure Plasma

Characterization of the Airborne Compact Atmospheric

atmospheric pressure RF plasma jet characterization of ... · Atmospheric pressure RF plasma jet: characterization of flow and O2 chemistry Atmospheric pressure plasma jets (APPJs)

Atmospheric Turbulence Characterization and …...Atmospheric Turbulence Characterization and Wavefr ont Sensing by Means of the Moiré De ectometry 3 (Tay et al., 2004), acceleration

The MonRAt telescope for atmospheric radiation€¦ · parabolic mirror mounted in a newtonian telescope setup. In front of the MAPMT photocathode, lters can be posi-tioned to select

Optical Intensity Interferometry with Atmospheric Cherenkov Telescope Arrays

Characterization of the atmospheric component of the

The VLT Survey Telescope ATLAS - arXiv · Figure 2. ATLAS and SDSS lter + telescope + atmospheric transmissions compared. All are calculated at airmass 1.3. No At-mospheric Dispersion

Estimated atmospheric emissions from biodiesel and ... · Estimated atmospheric emissions from biodiesel and characterization of pollutants in the ... Novo Hamburgo, Parobé ... ESTIMATED

Atmospheric Intensity Scintillation of Stars - III - Effects for Different Telescope Apertures

Atmospheric Characterization

Multi-element Analysis and Characterization of Atmospheric Particulate Pollution … · Multi-element Analysis and Characterization of Atmospheric Particulate Pollution in Dhaka Bilkis

Particle characterization at the Cape Verde atmospheric observatory

Characterization of particulate matter from atmospheric fluidized bed

MIRACLES: atmospheric characterization of directly imaged

Characterization of atmospheric pressure plasma plume

Beam Pattern Characterization of the 6-M TIGO Telescope · Beam Pattern Characterization of the 6-M TIGO Telescope Felipe Pedreros 1, Hugo Pacheco 2, Hayo Hase 3 1) Transportable

Characterization of atmospheric bioaerosols along the ...atmos.lzu.edu.cn/upload/news/N20200904103044.pdf · K. Tang et al.: Characterization of atmospheric bioaerosols along the

Atmospheric Characterization of Transiting Extrasolar Planets

Atmospheric Pressure Plasma as a Method for Improving .... Atmospheric... · Atmospheric Pressure Plasma as a Method for Improving Adhesive Bonding ... Characterization of the composite