Calibration Uncertainties in Data Analysis

hea-www.harvard.edu/AstroStat/calerr

| Introduction | Bibliography | Acknowledgements |

Last Updated: 2008sep09

Introduction

The importance of statistical errors has been well established in science: a measurement is of little value without an estimate of its credible range. However, without the transformation from measurement signals to physically interesting units afforded by the instrument calibration, the observational results cannot be understood in a meaningful way. Thus a good knowledge of the instrument characteristics, described by the instrument calibration data, is a crucial aspect of data analysis. However, it is well known that the measurements of the instrument's properties (e.g., quantum efficiency of a CCD detector, point spread function of a telescope, etc.) have associated measurement uncertainties. Currently, during the normal course of data analysis, these uncertainties are not properly taken into account. It is generally ignored entirely, or in some cases, it is assumed that the calibration error is uniform across an energy band or an image area. Such a process can often cause an erroneous interpretation of the data.

Incorporating calibration uncertainties is crucial on the grandest and the smallest scales. An obvious example is the comparison of results across many instruments and many wavebands, such as comparing results from standard sources like the Crab (Kuiper et al. 2001; esp. Appendix A), comparing multi-wavelength modeling of Gamma-ray Bursts (Hanlon et al. 1995), and comparing diffuse gamma-rays (Strong, Moskalenko, & Reimer 2000). Calibration uncertainties are also critical in the analysis of high-resolution data such as solar/stellar coronal spectra (Drake et al. 2003, Chung et al. 2004). In image analyses, issues such as the effect of a bright nearby source -- or a diffuse background emission -- on the measured flux of point sources, and the extent of the wings of the Point Spread Function (PSF), are critically important, but are usually not well-calibrated (Karovska et al. 2001, Thompson et al. 2001, Karovska et al. 2005). In general, a compact representation of the changes in instrument response -- whether representing systematic uncertainties or variations across an instrument field of view -- can be key in being able to handle these known but complicated instrument response details.

The main goal of this project is to provide a standard procedure for (1) describing the calibration uncertainties and (2) incorporating the calibration uncertainties into the data analysis.

Bibliography

Lee, H., Kashyap, V., Drake, J., Ratzlaff, P., Siemiginowska, A., Zezas, A., Connors, A., van Dyk, D., Park, T., & Izem, R., 2009, A Semiparametric Approach to Incorporating Systematic Uncertainties into Bayesian X-Ray Spectral Fitting, at the Workshop on Semiparametric Methodology, Jan. 8-10, 2009, U of Florida, FL

[ascii]

[.pdf]

Kashyap, V., Lee, H., Siemiginowska, A., MacDowell, J., Rots, A., Drake, J., Ratzlaff, P., Zezas, A., Izem, R., Connors, A., van Dyk, D., \& Park, T., 2008, SPIE 7016, 7016OP

manuscript: [.pdf]

presentation: [.ppt] ; [.pdf]

Kashyap, V., 2008, AISR PI meeting

Presentation [.ppt]

effective area uncertainty [animated .gif]

response matrix uncertainty [.mpg]

effect of ARF uncertainty on parameter pdf [.mov]

Lee, H., Kashyap, V., Drake, J., Ratzlaff, P., Siemiginowska, A., Zezas, A., Connors, A., van Dyk, D., Park, T., & Izem, R., 2008, Incorporating Effective Area Uncertainties Into Spectral Fitting, AAS-HEAD 2007, Los Angeles, #41.15

Abstract

[.pdf]

Kashyap, V., Lee, H., Drake, J., Ratzlaff, P., Siemiginowska, A., Zezas, A., Connors, A., van Dyk, D., Park, T., & Izem, R., 2007, Incorporating Effective Area Uncertainties into Spectral Fitting, Chandra Calibration Workshop, 2007.2

CCW2007.2

[.pdf]

Drake, J.J., Ratzlaff, P., Kashyap, V., Edgar, R., Izem, R., Jerius, D., Marshall, H., Siemiginowska, A., & Vikhlinin, A., 2007, Monte Carlo Processes for Including Calibration Uncertainties in Model Fitting Analyses, 2007, Chandra Calibration Workshop, 2007.1

CCW2007.1

[.pdf]

Drake, J.J., Ratzlaff, P., Kashyap, V., Edgar, R., Izem, R., Jerius, D., Siemiginowska, A., & Vikhlinin, A., 2006, Monte Carlo Processes for Including Chandra Instrument Response Uncertainties in Parameter Estimation Studies, Proc. SPIE, v6270, p627019

2006SPIE.6270E..49D

[.ps]

Drake, J.J., Ratzlaff, P., Edgar, R., Izem, R., Jerius, D., Kashyap, V., & Siemiginowska, A., 2005, Incorporating Calibration Errors in Parameter Estimation Analysis, at Chandra Calibration Workshop, Oct 31-Nov 1, Cambridge, MA

Presentation

Izem, R., 2005, Dimensionality Reduction of Functional Data, at Chandra Calibration Workshop, Oct 31-Nov 1, Cambridge, MA

Presentation

Acknowledgements

| Introduction | Bibliography | Acknowledgements |