Detection and (Linear) Data Model Jrn Wilms Remeis-Sternwarte & - - PowerPoint PPT Presentation

Detection and (Linear) Data Model

Jörn Wilms Remeis-Sternwarte & ECAP Universität Erlangen-Nürnberg http://pulsar.sternwarte.uni-erlangen.de/wilms/ joern.wilms@sternwarte.uni-erlangen.de

Part 1: Why is X-ray and Gamma-Ray Astronomy Interesting?

Part 2: Tools of the Trade: Satellites

Earth’s Atmosphere

CXC

Earth’s atmosphere is

• paque for all types of EM

radiation except for opti- cal light and radio. Major contributer at high energies: photoabsorption (∝ E−3), esp. from Oxygen (edge at ∼500 eV). = ⇒If one wants to look at the sky in other wave- bands, one has to go to space!

The Present

XMM-Newton (ESA): launched 1999 Dec 10 Chandra (NASA): launched 1999 Jul 23

Currently Active Missions: X-ray Multiple-Mirror Mission (XMM-Newton; ESA), Chandra (USA), Suzaku (Japan, USA), Swift (USA), International Gamma-Ray Laboratory (INTEGRAL; ESA), Fermi (USA), AGILE (Italy), MAXI (Japan), ASTROSAT (India), NICER (USA), Spectrum-X-Gamma (RU/D) We are living in the “golden age” of X-ray and Gamma-Ray Astronomy

The Present

IceCube/PINGU KAGRA Advanced LIGO/Virgo H.E.S.S. VLT APEX ALMA IRAM/NOEMA FAST MEERKAT ASKAP/MWA LOFAR Lomonosov/MVL-300 DAMPE Fermi AGILE INTEGRAL Spectrum-X-Gamma Kanazawa-Sat3 HaloSat MVN Insight-HXMT NICER XPNAV-1 POLAR ASTROSAT NuSTAR MAXI Swift Chandra XMM-Newton Solar Orbiter TESS Gaia Kepler/K2 HST Sofia Spitzer Queqiao/NCLE Longjiang-1/2 RadioAstron

2032 2030 2028 2026 2024 2022 2020 2018 Year

The Present

KM3NeT IceCube/PINGU Einstein Telescope LIGO-India KAGRA Advanced LIGO/Virgo CTA H.E.S.S. E-ELT LSST VLT APEX ALMA SKA IRAM/NOEMA FAST MEERKAT ASKAP/MWA LOFAR M5 Lomonosov/MVL-300 Glowbug DAMPE Fermi AGILE INTEGRAL Athena Einstein Probe XRISM SVOM IXPE ASO-S Spectrum-X-Gamma Kanazawa-Sat3 HaloSat MVN Insight-HXMT NICER XPNAV-1 POLAR ASTROSAT NuSTAR MAXI Swift Chandra XMM-Newton Ariel PLATO Euclid WSO-UV CHEOPS Solar Orbiter TESS Gaia Kepler/K2 HST WFIRST SPHEREx JWST Sofia Spitzer Queqiao/NCLE Longjiang-1/2 RadioAstron

2032 2030 2028 2026 2024 2022 2020 2018 Year

Part 3: Tools of the Trade: Mirrors and Detectors

Introduction

How is X-ray astronomy done? Detection process: Imaging Detection Data reduction Data analysis

Introduction

How is X-ray astronomy done? Imaging:

• Wolter telescopes (soft X-rays up to ∼15 keV)
• Coded Mask telescopes (above that)
• Collimators

Introduction

How is X-ray astronomy done? Detectors:

• Non-imaging detectors

Detectors capable of detecting photons from a source, but without any spatial resolution = ⇒ Require, e.g., collimators to limit field of view. Example: Proportional Counters, Scintillators

• Imaging detectors

Detectors with a spatial resolution, typically used in the IR, optical, UV or for soft X-rays. Generally behind some type of focusing optics. Example: Charge coupled devices (CCDs), Position Sensitive Proportional Counters (PSPCs)

X-ray Imaging

Cassegrain telescope, after Wikipedia

Reminder: Optical telescopes are usually reflectors: primary mirror → secondary mirror → detector Main characteristics of a telescope:

• collecting area (i.e., open area of telescope, ∼ πd2/4, where d: telescope diameter)
• angular resolution,

θ = 1.22 λ d (1)

if surface roughness and alignment can be ignored

X-ray Imaging

Optical telescopes are based on principle that reflection “just works” with metallic surfaces. For X-rays, things are more complicated...

1

α

2

α

1

θ n2 n1 n1 <

Snell’s law of refraction: sin α1 sin α2 = n2 n1 = n (2) where n index of refraction, and α1,2 angle wrt. surface normal. If n ≫ 1: Total internal reflec- tion Total reflection occurs for α2 = 90◦, i.e. for sin α1,c = n ⇐ ⇒ cos θc = n (3) with the critical angle θc = π/2 − α1,c. Clearly, total reflection is only possible for n < 1

Light in glass at glass/air interface: n = 1/1.6 = ⇒ θc ∼ 50◦ = ⇒ principle behind optical fibers.

X-ray Imaging

In general, the index of refraction is given by Maxwell’s relation, n = √ǫµ (4) where ǫ: dielectricity constant, µ ∼ 1: permeability of the material. For free electrons (e.g., in a metal), (Jackson, 1981, eqs. 7.59, 7.60) shows that ǫ = 1 − ωp ω 2 with ω2

p = 4πnZe2

me (5) where ωp: plasma frequency, n: number density of atoms, Z: nuclear charge.

(i.e., nZ: number density of electrons)

With ω = 2πν = 2πc/λ, Eq. (5) becomes ǫ = 1 − nZe2 πmec2λ2 = 1 − nZre π λ2 (6)

re = e2/mec2 ∼ 2.8 × 10−13 cm is the classical electron radius.

X-ray Imaging

n =

• 1 − nZre

π λ2 ∼ 1 − nZre 2π λ2 = 1 − ρ (A/Z)mu re 2π λ2 =: 1 − δ (7)

Z: atomic number, A: atomic weight (Z/A ∼ 0.5), ρ: density, mu = 1 amu = 1.66 × 10−24 g

Critical angle for X-ray reflection: cos θc = n = 1 − δ (8) Since δ ≪ 1, Taylor (cos x ∼ 1 − x2/2): θc = √ 2δ = 5.6′

• ρ

1 g cm−3 1/2 λ 1 nm (9) So for λ ∼ 1 nm: θc ∼ 1◦.

X-ray Imaging

Typical parameters for selected elements Z ρ nZ g cm−3 e− Å

−3

C 6 2.26 0.680 Si 14 2.33 0.699 Ag 47 10.50 2.755 W 74 19.30 4.678 Au 79 19.32 4.666

After Als-Nielsen & McMorrow (2004, Tab. 3.1)

To increase θc: need material with high ρ = ⇒ gold (XMM-Newton) or iridium (Chandra).

For more information on mirrors etc., see, e.g., Aschenbach (1985), Als-Nielsen & McMorrow (2004),

• r Gorenstein (2012)

X-ray Imaging

5 10 15 20 Photon Energy [keV] 0.0 0.2 0.4 0.6 0.8 1.0 Reflectivity

0.5deg 0.4deg 0.2deg 1deg

Reflectivity for Gold

X-rays: Total reflec- tion only works in the soft X-rays and

• nly under grazing

incidence = ⇒ grazing inci- dence optics.

Wolter Telescopes

Incident paraxial radiation Hyperboloid Paraboloid Hyperboloid Focus

after ESA

To obtain manageable focal lengths (∼10 m), use two reflections on a parabolic and a hyperboloidal mirror (“Wolter type I”)

(Wolter 1952 for X-ray microscopes, Giacconi & Rossi 1960 for UV- and X-rays).

But: small collecting area (A ∼ πr2l/f where f: focal length)

Wolter Telescopes

− +

Ni Electroforming Cleaning Gold Deposition Mirror Au Ni Separation (cooled)

Integration Production

Metrology Integration

• n Spider

Hole Drilling Handling Mandrel Super Polished Mandrel Recycle Mandrel

(after ESA)

Recipe for making an X-ray mirror:

• 1. Produce mirror negative (“Mandrels”): Al coated with Kanigen nickel (Ni+10% phos-

phorus), super-polished [0.4 nm roughness]).

• 2. Deposit ∼50 nm Au onto Mandrel
• 3. Deposit 0.2 mm–0.6 mm Ni onto mandrel (“electro-forming”, 10 µm/h)
• 4. Cool Mandrel with liquid N. Au sticks to Nickel
• 5. Verify mirror on optical bench.

numbers for eROSITA (Arcangeli et al., 2017)

Wolter Telescopes

Wolter Telescopes

Characterization of mir- ror quality: Half Energy Width, i.e., circle within 50% of the detected en- ergy are found. Note: energy dependent!

for XMM-Newton: 20′′ at 1.5 keV, 40′′ at 8 keV. for eROSITA: 16′′ at 1.5 keV, 15.5′′ at 8 keV Ground calibration, e.g., at PANTER

Detection of X-rays

Space Energy

EFermi

Semiconductors: separa- tion of valence band and conduction band ∼1 eV (=energy of visible light). Absorption of photon in Si: Energy of photon released

photo electron(s) + scattering off e− + phonons...

Number of electron-hole pairs produced: Problem: normal semiconductor: e−-hole pairs recombine immediately

Detection of X-rays

Space Energy

EFermi

Acceptors Donors

n-type p-type “Doping”: moves valence- and conduction bands. Connecting “n-type” and a “p-type” semiconductor: pn-junction. In pn junction: electron- hole pairs created by ab- sorption of an X-ray are separated by field gradient = ⇒electrons can then be collected in potential well away from the junc- tion and read out.

Detection of X-rays

Material Z Band gap E/pair (eV) (eV) Si 14 1.12 3.61 Ge 32 0.74 2.98 CdTe 48–52 1.47 4.43 HgI2 80–53 2.13 6.5 GaAs 31–33 1.43 5.2 Number of electron-hole pairs produced determined by band gap + “dirt effects”

(“dirt effects”: e.g., energy loss going into bulk motion of the detector crystal [“phonons”])

Npair ∼ Ephoton Epair (10)

• optical photons (E: few eV): ∼1 e−-hole-pair per absorption event
• X-ray photons: ∼1000 e−-hole-pairs per photon

But: Since band gap small: thermal noise = ⇒ need cooling

(ground based: liquid nitrogen, −200◦C, in space: more complicated...)

Detection of X-rays

e _ e _ e _ e _ photoelectrons Atom Photoelectron track p−type (undepleted) Potential energy for an electron p−type (−) silicon (depleted) n−type (+) silicon (depleted) SiO insulator 0.1 m deep Polysilicon electrodes 1 1 1 2 3 1 pixel (~15 m) ~2 m ~10 m ~250 m 3 1 2 3 1 2 3 2 3 µ µ µ µ µ Photon µ conductors; ~0.5 m deep 2

After Bradt

Two-Dimensional imaging is possible with more complicated semiconduc- tor structures: Charge Coupled Devices (CCDs).

CCDs

−Pulse Φ

1

Φ

2

Φ

3

Φ − − − − − −

3

Φ

2

Φ

1

Φ Transfer

− − − − − − − − − 0V 0V 0V 0V 0V 0V 0V 0V +V +V +V +V (after McLean, 1997, Fig. 6.9)

Principle of the readout of a CCD with Φ-pulses.

CCDs

Gate strips p−stops Read−out electronics

combine several readout stripes gives a two dimen- sional detector.

Separation of individual columns with p-stops (highly doped Si) to prevent charge diffusion between columns.

Read out:

• move charge to corner
• preamplify
• digitize in Analog-

Digital-Converter (ADC) Fast CCDs: one read out electronics per column

Expensive, consumes more power = ⇒ only done in the fastest X-ray CCDs (< µs resolution).

CCDs

+ + + + + + +

− − − − − −

preamp

• n−chip

n−

p p p p p p p

285 µm 12 µm Phi−Pulses to transfer electrons Depletion voltage Potential for electrons

Anode

Transfer direction

Schematic structure of the XMM-Newton EPIC pn CCD.

Problem: Infalling photons have to pass through structure on CCD surface = ⇒ loss of low energy response, also danger through destruction of CCD structure by cosmic rays... Solution: Irradiate back side of chip. Deplete whole CCD-volume, transport electrons to pixels via adequate electric field (“backside illuminated CCDs”)

Note: solution works mainly for X-rays

Grades

Charge cloud in Si has roughly 2D Gaussian distribution = ⇒ Distribution of clouds on adjacent pixels: “event grades” Total E:

pixels Ei

ESA: single events, double events,...; US: grade 0 for singles, grade 1–4 for doubles, etc.; other grading schemes are possible.

• C. Schmid

Background

• M. Wille

Background in CCDs:

• cosmic rays (e.g., muons), leaving long tracks on detector
• low E (<MeV) protons, focused via mirrors onto CCDs. Especially during solar flares

In principle also electrons: charged “net” over mirror deflects them; this not possible for protons due to higher mass; back-illuminated CCDs can usually cope with this and are not damaged.

Typically background reduction on board through thresholding events, e.g., >15 keV where mirror is non-reflective

Optical Loading

• F. Krauss

Optical loading: like all CCDs, X-ray sensitive CCDs are sensitive to optical light Mitigation: optical filter, either on chip or via filter wheel Blocks out light to 8 mag or so = ⇒ brighter stars are problem

Part 5: Analyzing Data – Theory

Formal Data Analysis

In order to analyze X-ray data, we need to understand how the measured signal is produced:

• 1. Sensitivity of the detector: “how much signal do we have?”

= ⇒ modeled as energy-dependent collecting area

• ften called the “ARF” (ancilliary response function)
• 2. Energy resolution of the detector: “where is the signal detected?”

= ⇒ modeled as convolution of signal with energy resolution

• ften called the “detector response”

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

count rate in channel c (counts s−1)

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

count rate in channel c (counts s−1) photon flux density (ph cm2 s−1 keV−1),

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

count rate in channel c (counts s−1) photon flux density (ph cm2 s−1 keV−1),

We measure this

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

count rate in channel c (counts s−1) photon flux density (ph cm2 s−1 keV−1),

We measure this Astrophysics is here

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

count rate in channel c (counts s−1) detector response (∝ probability to detect photon of energy E in channel c). photon flux density (ph cm2 s−1 keV−1),

We measure this Astrophysics is here

Linear Model

Summarizing the previous information mathematically: nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (11)

count rate in channel c (counts s−1) detector response (∝ probability to detect photon of energy E in channel c). effective area (cm2) photon flux density (ph cm2 s−1 keV−1),

We measure this Astrophysics is here Calibration (“response” / “rsp”)

Effective Area

Mirror

Ir M4,5 Ir M3 Ir M2 Ir M1 Ir L3 Ir L1

10 1 0.1 10+4 10+3 10+2 Energy [keV] Area [cm2]

Effective area for an Athena-like mission with a simplified Si-based detector

Effective Area

Mirror +45nm polyamide+70nm Al

Al K O K N K C K Ir M4,5 Ir M3 Ir M2 Ir M1 Ir L3 Ir L1

10 1 0.1 10+4 10+3 10+2 Energy [keV] Area [cm2]

Effective area for an Athena-like mission with a simplified Si-based detector

Effective Area

Mirror +45nm polyamide+70nm Al +Si QE

Al K O K N K C K Si K Ir M4,5 Ir M3 Ir M2 Ir M1 Ir L3 Ir L1

10 1 0.1 10+4 10+3 10+2 Energy [keV] Area [cm2]

Effective area for an Athena-like mission with a simplified Si-based detector

Effective areas of the most important current X-ray satellites

(Hanke, 2011)

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

= ⇒ imaging spectroscopy!

requires very fast readout (≫ arrival rate of photons) bright sources: several 1000 photons per second = ⇒ readout in µs!

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

= ⇒ imaging spectroscopy!

requires very fast readout (≫ arrival rate of photons) bright sources: several 1000 photons per second = ⇒ readout in µs!

Poisson statistics of relative collected charge: ∆Npair Npair

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

= ⇒ imaging spectroscopy!

requires very fast readout (≫ arrival rate of photons) bright sources: several 1000 photons per second = ⇒ readout in µs!

Poisson statistics of relative collected charge: ∆Npair Npair =

• Npair

Npair

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

= ⇒ imaging spectroscopy!

requires very fast readout (≫ arrival rate of photons) bright sources: several 1000 photons per second = ⇒ readout in µs!

Poisson statistics of relative collected charge: ∆Npair Npair =

• Npair

Npair = 1

• Npair

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

= ⇒ imaging spectroscopy!

requires very fast readout (≫ arrival rate of photons) bright sources: several 1000 photons per second = ⇒ readout in µs!

Poisson statistics of relative collected charge: ∆Npair Npair =

• Npair

Npair = 1

• Npair

∼ ∆E E

(∆E/E ∝ E−1/2 because of Poisson!)

Response Matrix

• ptical:

1 pair/photon = ⇒ collected charge ∝ intensity X-ray: many pairs/photon = ⇒ collected charge ∝ energy

Npairs ∝ Ephoton

= ⇒ imaging spectroscopy!

requires very fast readout (≫ arrival rate of photons) bright sources: several 1000 photons per second = ⇒ readout in µs!

Poisson statistics of relative collected charge: ∆Npair Npair =

• Npair

Npair = 1

• Npair

∼ ∆E E =

exact eq.

= 2.355

• 3.65 eV · F

E

(∆E/E ∝ E−1/2 because of Poisson!)

F ∼ 0.1 = ⇒ ∼2% at 5.9 keV

F is called the Fano factor

Response Matrix

(Fürst, 2011)

Monoenergetic photons get “smeared” in energy by the detector response. Width of peak given by Eq. on previous slide

note: RMF is not required to be normalized to unity! Some missions include some quantum efficiency terms in it.

Response Matrix

(Treis et al., 2009)

Mn Kα (5.9 keV), Mn Kβ (6.5 keV), Si escape (4.2 keV), Al fluorescence (1.5 keV) from housing [background]

Response Matrix

Response Matrix of the RXTE-PCA, log scale

Secondary “escape” peaks: response caused by Xe Kβ and Xe Lα photons escaping the de- tector.

Pile Up

some invalid patterns Pile up: Arrival rate of photons > readout timescale of CCD = ⇒ get wrong energy assignment

• energy pileup: multiple photons in

same pixel

• pattern pileup: multiple photons in

adjacent pixels some produce invalid patterns, but some mimic normal photons nonlinear effect! figures by C. Schmid (PhD thesis Remeis-Observatory & ECAP, 2012)

Pile Up

Pile up: Arrival rate of photons > readout timescale of CCD = ⇒ get wrong energy assignment

• energy pileup: multiple photons in

same pixel

• pattern pileup: multiple photons in

adjacent pixels some produce invalid patterns, but some mimic normal photons nonlinear effect! figures by C. Schmid (PhD thesis Remeis-Observatory & ECAP, 2012)

χ2-minimization

We had nph(c) = ∞ R(c, E) · A(E) · F(E) dE + nbackground(c) (12) where

• nph(c): source count rate in channel c (counts s−1),
• F(E): photon flux density (ph cm2 s−1 keV−1),
• A(E): effective area (units: cm2),
• R(c, E): detector response (probability to detect photon of energy E in

channel c). We measure nph(c), but the astrophysics is contained in F(E). Inver- sion of Eq. (12) is not possible! = ⇒ Data analysis:

• 1. guess F(E) from astrophysics
• 2. predict nph(c) from Eq. (12)
• 3. compare prediction and measurement
• 4. modify guess...

χ2-minimization

To analyze data: discretize Eq. (12): Sph(c) = ∆T ·

nch

• i=0

A(Ei) · R(c, i) · F(Ei) · ∆Ei ∀c ∈ {1, 2, . . . , nen} (13) where Sph(c): total source counts in channel c, ∆T: exposure time (s), A(Ei): effective area in energy band i (“ancilliary response file”, ARF), R(c, i): response matrix (RMF), F(Ei): source flux in band (Ei, Ei+i), ∆Ei: width of energy band. Because of background B(c) (counts), what is measured is Nph(c) = Sph(c) + B(c) (14) So estimated source count rate is ˜ Sph(c) = Nph(c) − B(c) (15) with uncertainty (Poisson!) σ˜ Sph(c) =

• σNph(c)2 + σB(c)2 =
• Nph(c) + B(c)

(16)

χ2-minimization

To get physics out of measurement, need to find F(Ei). Big problem: In general, Eq. (13) is not invertible. = ⇒ χ2-minimization approach Use a model for the source spectrum, F(E; x), where x vector of parameters (e.g., source flux, power law index, absorbing column,...), and calculate predicted model counts, M(c; x), using Eq. 13). Then form χ2-sum: χ2(x) =

• c

˜ Sph(c) − M(c; x) 2 σ˜ Sph(c)2 (17) Then vary x until χ2 is minimal and perform statistical test based on χ2 whether model F(E; x) describes data.

Programs used: XSPEC, ISIS, SPEX In practice, background is not subtracted from measurement, but added on model prediction.

Part 4: Further Reading

Literature

LONGAIR, M.S., 1992, High Energy Astrophysics, Vol. 1: Particles, Photons, and their Detection, Cambridge: Cambridge Univ. Press, ∼50e

Good introduction to high energy astrophysics, the 1st volume deals extensively with high energy procsses, the 2nd with stars and the Galaxy. The announced 3rd volume has never appeared. Unfortunately, everything is in SI units.

TRÜMPER, J., HASINGER, G. (eds.), 2007, The Universe in X-rays, Heidel- berg: Springer, 96.25e

Book giving an overview of X-ray astronomy written by a group of experts (mainly) from Max Planck Institut für extraterrestrische Physik, the central institute in this area in Ger- many.

BRADT, H., 2004, Astronomy Methods: A Physical Approach to Astronomi- cal Observations, Cambridge: Cambridge Univ. Press, $50

Good general overview book on astronomical observations at all wavelengths.

Literature

CHARLES, P., SEWARD, F., 1995, Exploring the X-ray Universe, Cambridge: Cambridge Univ. Press, out of print

Summary of X-ray astronomy, roughly presenting the state of the early 1990s.

SCHLEGEL, E.M., 2002, The restless universe, Oxford: Oxford Univ. Press, 32e

Popular X-ray astronomy book summarizing results from XMM-Newton and Chandra.

ASCHENBACH, B. et al., 1998, The invisible sky, New York: Copernicus

Popular “table top” book summarizing the results of the ROSAT satellite, with many beautiful pictures.

KNOLL, G.F., 2000, Radiation Detection and Measurement, 3rd edition, New York: Wiley, 126e

The bible on radiation detection. If you want one book on detectors, this is it.

WWW-Pages

• http://pulsar.sternwarte.uni-erlangen.de/wilms/teach

My lectures. Including a 2 semester long course on X-ray astronomy (with exercises) and a 1 semester long course on radiation processes. Currently offline, but soon to be online again, but I provide slides upon request.

• http://heasrc.gsfc.nasa.gov

Main www page of NASA’s satellite missions, including the data archive interface at http://heasarc.gsfc.nasa.gov/db-perl/W3Browse/w3browse.pl

• http://cxc.harvard.edu

Chandra data center (and analysis info)

• http://xmm.esac.esa.int

XMM-Newton data center (and analysis info)

• http://ledas-www.star.le.ac.uk
• Univ. Leicester data archive (many missions)
• http://www.isdc.unige.ch/heavens/

Interface to prereduced INTEGRAL and RXTE data

X-ray Data Analysis 65a Bibliography Als-Nielsen, J., & McMorrow, D. 2004, Elements of Modern X-ray Physics, (New York: Wiley) Arcangeli, L., Borghi, G., Bräuninger, H., et al. 2017, in Internat. Conf. on Space Optics, ed. E. Armandillo, B. Cugny, N. Karafolas, Vol. 10565, SPIE Conf. Ser., 1056558 Aschenbach, B., 1985, Rep. Prog. Phys., 48, 579 Fürst, F., 2011, Ph.D. thesis, Universität Erlangen-Nürnberg, Erlangen Giacconi, R., & Rossi, B. 1960, J. Geophys. Res., 65, 773 Gorenstein, P., 2012, Opt. Eng., 51, 011010 Hanke, M., 2011, Ph.D. thesis, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen Jackson, J. D., 1981, Klassische Elektrodynamik, (Berlin, New York: de Gruyter), 2 edition McLean, I., 1997, Electronic imaging in astronomy: detectors and instrumentation, Wiley) Treis, J., Andritschke, R., Hartmann, R., et al. 2009, J. Instrumentation, 03, 03012 Wolter, H., 1952, Annalen der Physik, 445, 94