LargeSynopticSurveyTelescop

CharacterizationMetric

PipelinesVersion11.0(Summer

FrossieEconomou,JohnSwinbank,JimBosch,

DMTR-11

LatestRevision:2015-10-01

Abstract

StartingfromSummer2015,administrative(”cycle”)

measurementsreportcharacterisingthecurrentperformance.

thesereportsareexpectedtoincreaseinnumberandsophistication

releases.Thisbriefreportdescribemeasurementsof

LARGESYNOPTICSURVEYTELESCOPE

---

LARGESYNOPTICSURVEYTELESCOPE

Summer2015CharacterizationReportDMTR-11LatestRevision2015-10-01

Contents

1SummaryofPhotometricRepeatabilityMeasurements1

2SummaryofAlgorithmicPerformanceMeasurements1

3SummaryofComputationalPerformanceMeasurements2

ii

---

LARGESYNOPTICSURVEYTELESCOPE

Summer2015CharacterizationReportDMTR-11LatestRevision2015-10-01

CharacterizationMetricReport:Science

11.0(Summer2015)

1SummaryofPhotometricRepeatabilityMeasurements

SubmittedbyJimBosch

Thisdatasetisaselectionof?-bandHyperSuprime-Camengineeringdat

Stripe82region.Thisdatasetconsistsof30sexposures,

LSSTdataindepth.Ourcurrentcalibrationapproachhas

weultimatelyplantoimplementforLSST:

•There’scurrentlynorelativecalibrationbeingrunat

•Wehaveonlylimitedcorrectionforchromaticeffects.

•There’s’scurrentlynoallowanceforzeropointvariations

•WealsouseamuchsimplersampleselectionthanthatproposedSRD.

Annotatedcodetocomputethemetricscanbefoundat

https://github.com/lsst/afw/blob/

tickets/DM-3896/examples/repeatability.ipynb

Metric CharacterisedMetric

Target

Measured Measurement

Ref

Value Method

Photometric

repeatability

DLP-307

≤

(procCalRep)

13 mmag

10.6DM-3338

mmag (using

?

-band data)

Photometric

repeatability

DLP-315

≤

(PA1gri)

13 mmag

10.6DM-3338

mmag

Photometric

repeatability

DLP-316

≤

(PA1uzy)

13 mmag

10.6DM-3338

mmag (using

?

-band data)

2 Summary of Algorithmic Performance Measurements

Submitted by John Swinbank

1

---

LARGE SYNOPTIC SURVEY TELESCOPE

Summer 2015 Characterization Report

DMTR-11

Latest Revision 2015-10-01

The

?

-band HSC engineering data (described above) was used

caveats apply. Consult the tickets in the Measurement Method

Metric Characterised Metric

Target

Measured Measurement

Ref

Value Method

Residual PSF

Ellipticity

DLP-290

Correlations

≤ 5e-3

6e-5

(TE1)

DM-3040

Residual PSF

Ellipticity

DLP-290

Correlations

≤ 5e-3

2e-5

(TE2)

DM-3047

Relative Astrometry

(AM1)

DLP-310

< 60 mas

12.49

DM-3057

mas

Relative Astrometry

(AM2)

DLP-311

< 60 mas

12.19

DM-3064

mas

3 Summary of Computational Performance Measurements

Submitted by John Swinbank and Simon Krughoff

At this point of Construction, the computational performance

tion of precursor data processing and extrapolation from

DECam/HITS data was used for the OTT1 estimate and for the

surement of the Alert Production Computational Budget in

3rd Data Challenge

Document-26217

[].

For the Data Release Production of the computational budget,

estimating diffim performance, HSC-I for assembling and measuring

measurement, estimates from FDR for multifit, and data from

SDQA. Calculations for the DRP

computational

this iPython

notebook

budget used

Metric Chracterised

Metric

Target

Measured

Measurement

Ref

Value

Method

OTT1

DLP-328

≤ 240 sec200-250

sec

DM-3724

AP

Computational

DLP-329

Budget

≤ 231 TFLOPS

34-39

DM-3267

TFLOPS

DRP

Computational

DLP-314

Budget

≤ 645 TFLOPS 318

TFLOPS

DM-3083

2

---

LARGE SYNOPTIC SURVEY TELESCOPE

Summer 2015 Characterization Report

DMTR-11

Latest Revision 2015-10-01

References

[1]

[LPM-17]

, Ivezić, Ž., The LSST Science

LSST

Collaboration,

Science Requirements

2011,

,LPM-17,URLhttps://ls.st/LPM-17

[2][Document-26217],Kantor,J.,2010,DataChallenge3bPerformanceTest,Document-

26217,URLhttps://ls.st/Document-26217

3