Lessons learned from SPIDER’s

first flight implications for future ballooning

and satellite missions

1

Cosmology on Safari February, 2017

Jón E. GuðmundssonThe Oskar Klein Centre for

Cosmoparticle Physics,Stockholm University

on behalf of the SPIDER Collaboration

SPIDER

2

• A balloon-borne polarimeter• Mapped about 10% of the sky• Six telescopes, 3/3 at 94/150 GHz

• Approx. 2000 transition edge sensors cooled to 300 mK (85% yield)

• Sensitivity near photon-noise limit• About 0.5 deg resolution

• ℓ ≈ 10–300

• Science goals:• Set limits on primordial

gravitational wave amplitude of r < 0.03 @ 99% confidence (no foregrounds)

• Characterize polarized foregrounds• Launched on Jan 1, 2015• Flew for 16.5 days• See Benjamin Racine’s talk

following this one!

Long Duration Ballooning

• Circumpolar winds ~10 days/rev• On average 20 day flights at 36 km• Why Ballooning?• Space like loading (NET)• Access to larger angular scales• Wider frequency windows• Preparation for SPB promised land• Why Antarctica?• Continuous solar power• Long flight times• At what price?• Narrow launch windows• Recovery difficulties• Mass, power, and automation

3

Data from CSBF

4

December 2014 ca. Ross Island, Antarctica

The observational challengeTo clearly separate a primordial signal from more local sources we must

Constrain spectral energy

distribution

The observational challengeTo clearly separate a primordial signal from more local sources we must

Constrain spectral energy

distribution

Verify statistical isotropy

—

Commander dust intensity, 150 GHz

The observational challengeTo clearly separate a primordial signal from more local sources we must

Constrain spectral energy

distribution

Verify statistical isotropy

Probe all angular scales

— —

TT

EE

BBLensing B-modes

Large scale polarizationReionization

Primordial gravitational waves

8

Primary characteristicsSky coverage About 10 %

Scan rate (az) 3.6 deg/s at peak

Polarization modulation Stepped cryogenic HWP

Detector type Antenna-coupled TES

Multipole range 10 < ℓ < 300

Observation time 16 days

Limits on r† 0.03† Assuming no foregrounds, at 99% confidence

Frequency [GHz]

95 150

Telescopes 3 3

Bandwidth [GHz] 22 36

Optical efficiency 30-45% 30-50%

Angular resolution* [arcmin] 41.1 28.2

Number of detectors† 678 973

Detector loading‡ [pW] ≤ 0.25 ≤ 0.35

NET per detector [μK∙rts] 163 165*FWHM. †Current channel cuts‡Including atmosphere, sleeve, window, and baffle

Pivot

Aperture

Sun shield

Top dome

Hermetic feedthroughGondola

Reaction wheel

SIP

Vacuumvessel

8.1 m

9

VCS1

VCS2 10

Main tank

Vacuum vessel

Refrigerator

FPU

Objective

Eyepiece

Our trajectory

11

Feb 5, 2015 — data recovery

Fig. BAS

Nov 17, 2015 — payload recovery

Fig. Ed Young

Fig. Ed Young

Oct 13, 2016 — recycled aluminum

Observation regions

15

Fig. Sasha Rahlin

Polarization amplitude

16

● SPIDER scanned approximately 10% of the sky

Fig. Sasha Rahlin

SPIDER hits map

17

● About 12.3% covered, about 6.3% if you apply the following hits weight

First flight performance summary

18

Quantity 94 GHz 150 GHz

Good channels 678 973

Percentage flagged [%] 30.4 33.4

Integ. time [ch-yr] 16.2 22.2

Median NET [μKrts] 165 163

Total NET [μKrts] 5.4 6.7

Map depth [uK arcmin] 22.9 18.9

Beam FWHM [arcmin] 41.1 28.2

￭ Most instantaneously sensitive experiment to have deployed!○ Currently flagging heavily to avoid RFI and thermal variations

￭ Map depth currently around 20 μK arcmin ○ Planck HFI map depth around 77 and 33 μK arcmin

at 100 and 143 GHz, respectively￭ Total data volume around 1.5 TB (very similar to Planck total)￭ Remarkably low detector loading￭ Detect non-CMB polarization with high significance!

SPIDER 150 GHz band

19

Map depth

currently about

20 μK arcmin

PRELIMINARY

Comparison to Planck HFI

20

Map depth about

33 μK arcmin

PRELIMINARY

Planck PR2 Maps, nominal Full Misison map

First science results

21

See preliminary results in Ben’s talk following

this one

22JPL Microdevices Lab

Cosmic rays

are insignificant...

23

Planck 143 GHz timeline

SPIDER 150 GHz timeline

Our TES detectors are

great for space!

Radio Frequency Interference (RFI)

24

...but radio interference from payload antennas is significant

Currently flagging about

30% of data.

Expect this to

come down.

Plot from A. Gambrel

25

• The most instantaneously sensitive experiment to date!

• Exceptionally low photon load• Phonon noise limited

• Colder focal plane desirable

In-band loading Instrument NET

94 GHz ≤ 0.25 pW 6.5 μK√s (652)

150 GHz ≤ 0.35 pW 5.1 μK√s (1030)

Boomerang (150 GHz) 0.5 pW ~ 60 μK√s

BICEP2(150 GHz) 4.7 pW ~ 16 μK√s

Planck (143 GHz) 0.8 pW ~ 16 μK√s

Overall detector performance

SPIDER-2 development

26

● Receivers operating at 285 GHz are built and undergoing testing● Project about 335 uKrts sensitivity per detector and 17 arcmin beam

New cryostat built

and leak tight!

NIST OMTs and silicon

platelete feedhorns

SPIDER-2 development

27

• Third frequency band (285 GHz) complementary to Planck

Detector back-of-the-envelope

28

• Absolute detector sensitivity set by:• Geometry, materials, and temperature• Instrument transmission and emissivities• Photon loading incl. atmosphere and the CMB• Readout electronics

• Altitude and water set atmospheric (photon) loading

• What’s the typical bolometer performance as a function of altitude?

am Atmospheric Model — Scott PaineThanks to S. Bergman!!

Atacama1 mm PWV45 deg elev.

Photo: JPL

Performance as a function of altitude

29

Performance as a function of altitude

30

Of course, integration time

and filtering are

equally important!

Third flight scheduled for the spring of 2017

Wanaka, New Zealand44°42′S 169°09′E Second mid-latitude flight:

May 2016, 46 days in air

Wanaka, New Zealand44°42′S 169°09′E First mid-latitude flight:

March 2015, 33 days in air

Wanaka, New Zealand44°42′S 169°09′E

SPB Ballooning

31

• Constant volume balloons• Stable altitude

• First science flight in 2016• Potentially offers ~100 day flights

• Launch base in New-Zealand• Intermediate latitudes• Full diurnal cycles• Payload mass ~1000 kg

Data from CSBF

Nordita program – July 2017

32

Peter AdsheadAndreas AlbrechtFrank Avignone Nicholas BattagliaDaniel BaumannAnthony ChallinorCynthia ChiangJens ChlubaJoanne CohnCora DvorkinHans K. Eriksen Silvia GalliJohn T. GiblinAriel GoobarLavinia HeisenbergRenée HlozekDaniel HolzBhuvnesh JainBill Jones Renata Kallosh

Will KinneyEiichiro Komatsu Massimiliano Lattanzi Andrei LindeMarilena LoVerdePavel NaselskyPaolo NatoliHans Peter Nilles Lyman Page Enrico Pajer Marco Peloso Graca Rocha Subir Sarkar Uroš Seljak Ravi Sheth Jon SieversScott Watson Martin White

Invited and confirmed participants

OrganizersKatherine Freese, Martina Gerbino, Shirley Ho, Jon Gudmundsson, Ingunn Wehus

including conference on inflation and the CMB

SPIDER summary

• SPIDER is a completely autonomous payload • Electrical power, pointing control and reconstruction,

redundant data systems (100 GB/day), cryogenic single-crystal sapphire polarization modulators

• Most sensitive microwave receiver to date• Weighs a bit over 3.5 tons

• Over 500 kg cooled to 4 K• About 30 kg cooled to 250 mK

• About 15 km of cryogenic wiring• Hand-made with love (most of the time)• Subsequent flight planned for 2018

Photo: Z. Kermish

34

Stay tuned!