Slide1 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Detection of Ultra High-Energy Cosmic Particles with the Use of Radio and Radar Methods December 7, 2005
Oscar Stål Physics in Space Programme Swedish Institute of Space Physics, Uppsala
Dept. of Astronomy and Space Physics, Uppsala University
Supervisor: Bo Thidé
Slide2 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Outline Presentation of objectives
Introductory high-energy cosmic particle physics
Radar studies of EAS ionisation columns
Lunar satellite detection of Askar’yan radio pulses
Discussion and outlook
Slide3 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Background and framework LOIS radio sensor network and deep space radar project Initiated by Bo Thidé, collaboration with LOFAR (Netherlands)
Provides new methods to study fundamental physics in space Can this facility or these methods be of use in astroparticle physics?
If not usable, at least it is necessary to quantify what radio/radar background to expect from UHE cosmic particles
Ultimate goal is low-frequency array on the Moon – LIFE (LURBO)
Slide4 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Objectives Our objectives for this diploma work have been two-fold:
To determine, by approximate analytical methods, the radar cross section of the ionisation columns created by Extensive Air Showers
To investigate the feasibility of using a lunar satellite for in situ detection of Askar’yan radio pulses from cosmic particle interactions with the lunar regolith
Slide5 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Outline Presentation of objectives
Introductory high-energy cosmic particle physics
Radar studies of EAS ionisation columns
Lunar satellite detection of Askar’yan radio pulses
Discussion and outlook
Slide6 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Cosmic rays Charged particles: p,
Steep decrease of flux with energy: ~ E-2.7 E < 1015 eV ~ E-3.1 E > 1015 eV
Flux extends to the highest energy ever observed (UHE)
Isotropic flux, B obfuscation
Sources of UHE particles not yet determined
Slide7 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Extensive Air Showers (EAS) Cosmic particle interacts with constituents of the atmosphere
Shower of secondary particles generated
Hadronic, muonic and EM components
Transverse scale given by Molière radius, 90% of the energy contained within rM
For air, rM = 70 m at sea level and increasing with altitude to several hundred metres
Slide8 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Detection of UHE cosmic rays Air shower arrays Haverah Park, AGASA, KASCADE, Pierre Auger Detects shower particles at ground level
Fluorescence telescopes Fly's eye, HiRes Fly's Eye Excellent energy resolution, sensitivity Low duty cycle
Radio methods Pioneered in the 60's, now LOPES Coherent geosynchrotron emission Highly polarised short radio pulses Falcke et al., Nature 435 (2005)
Slide9 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 The Greisen-Zatsepin-Kuzmin (GZK) cutoff For E > 1019.5 eV, cosmic rays interact with CMB photons
Pion photoproduction: p + p0 + p p + + + n etc.
Intergalactic medium no longer transparent over Mpc scales
Still, cosmic rays have been observed beyond this “cutoff”
A most interesting question in astroparticle physics is why?
Slide10 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Ultra high-energy neutrinos UHE neutrinos should be produced in the GZK process: + + + e+ +e + +
More sources of UHE neutrinos suggested, but none confirmed
No magnetic field influence
Detection using optical, radio or acoustic methods
Best limits on UHE neutrino flux obtained using radio methods Gorham et al., Phys. Rev. Lett. 93 (2004)
Slide11 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Outline Presentation of objectives
Introductory high-energy cosmic particle physics
Radar studies of EAS ionisation columns
Lunar satellite detection of Askar’yan radio pulses
Discussion and outlook
Slide12 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Nishimura-Kamata-Greisen (NKG) model for EAS Longitudinal development parametrised by shower age:
“slant depth”
Total number of particles in the shower:
Transverse distribution of shower particles:
Slide13 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Ionisation columns Atmospheric ionisation yield calculated from transverse particle density and ionisation parameters for air
Ionisation over long distances, > 10 km
Evaporation time scale uncertain: 20 s – 20 ms
Slide14 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Ionisation columns Ionisation volume modelled as collisionless, cold, non-magnetised plasma with inhomogeneous density
Slide15 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Radar cross section and scattering width The radar cross section [m2] is the projected area of a perfectly reflecting sphere giving a reflected power equivalent to that of the real target
In a two-dimensional problem, the cross section is replaced by the scattering width [m] Cross section (RCS):
Scattering width (SW):
Slide16 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Inhomogeneous wave equations To determine the radar cross section of the EAS ionisation columns, the scattered E and B fields are required
Wave equations for the spatially inhomogeneous medium derived from first principles (Maxwell):
Slide17 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Separation of the wave equations For radially inhomogeneous n(x) it is possible to separate the wave equations in cylindrical coordinates
The Ez and Bz component equations decouple for infinite cylinder Shower aging is effectively neglected, we treat only maximum
TM and TE mode scattering for normally incident wave m is the azimuthal “quantum number”
Slide18 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Scattering theory in cylindrical geometry The Sommerfeld radiation condition for n = 2 allows the total field to be written as incoming plane wave + scattered cylindrical wave:
Asymptotic dependence of scattered wave is chosen consistently with plane wave expansion when there is no scatterer:
The phase shifts m contain all information about the scattering:
Slide19 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Phase-integral approximations How to obtain the phase-shifts? Usually by analytic matching if the inner solution is known. We use the phase-integral (WKB) method The general second order linear ODE of a complex z
has approximate phase-integral solutions in terms of
where q(z) is generated asymptotically from an arbitrary base-function, for order 2N+1 of approximation:
Slide20 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 First order approximations to the phase-shifts Phase-shifts in phase-integral method obtained from the asymptotics of the solutions
Connection formula used to cross the turning point from q2 < 0 to q2 > 0
In first order approximation, it is possible to use a path along the real line
Slide21 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Choice of base function The base function needs to be consistent with the physics, and this choice is in general a difficult problem. No generic method exists From direct transformation of the radial equations, we obtain for the TM mode equation
By using instead the modified base function
we reproduce the zero phase-shifts in the no-scattering limit, which is desirable. Corresponds to the Langer modification in a spherical problem and was previously studied by Berry et al.
Slide22 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Scattering widths – numerical results Numerical integration of the phase-shift formula using the NKG refractive index for horisontal EAS at maximum development
Calculations of the scattering width for various shower altitudes, radar frequencies and primary particle energy
Slide23 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Longitudinal length scale and 3D cross section Shower not infinitely long – what is the longitudinal length scale?
Scattering theory is formulated in the extreme far-field?
Consider the first Fresnel zone, the longitudinal dimension is given by inverting the far-field condition:
Affects the range of the radar system, and thereby the sensitivity to UHE particle showers. No simulations performed on this
Slide24 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Outline Presentation of objectives
Introductory high-energy cosmic particle physics
Radar studies of EAS ionisation columns
Lunar satellite detection of Askar’yan radio pulses
Discussion and outlook
Slide25 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Background Chandrayaan-1 Moon mission
India's first mission to the Moon
Scheduled for launch in 2007-08
100 km polar orbit
ELVIS instrument proposal
HF/VHF radio receiver of LOIS type
Can it be used for detection of UHE cosmic particles?
Slide26 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Detection principles 1. Neutrino enters the Regolith 2. First interaction, shower initiated 3. Coherent V-C radio emission 4. Rays refracted at interface 5. Detection by satellite or surface-based aerials
Slide27 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Radio emissions from particles in dense media Showers very localised in dense media, rM ~ 10 cm
Wavelengths longer than shower dimensions means emission becomes coherent
Ouput power scales quadratically with E, dominates optical output at UHE
Radio transparent material required
Ice, very dry rock, permafrost, giant underground salt domes suggested
Slide28 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Coherent Vavilov-Cerenkov emission Particle track is not infinite, so the output is smeared around the Cerenkov angle
Coherent radiation for higher frequencies closer to the Cerenkov angle qC
Long wavelengths means the shower radiates as a single particle, hence only the net charge contributes v > c/n cos(qC)=1/(nb)
Slide29 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 The Askar`yan effect Neutral shower gives no emission
The particles in an initially neutral shower will undergo scattering processes as they traverse the material...
... and acquire a negative charge excess of 20-30%
This process is called the Askar’yan effect Compton scattering: + e-atom e- Bhaba scattering: e+ e-atom e+ e- Annihilation: e+ e-atom
Slide30 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Radiation properties Zas-Halzen-Stanev (ZHS) Monte Carlo based on simulations of showers in dense media confirms the 20-30% charge excess Radio emission at Cerenkov angle well parametrised by n0 specific decoherence frequency, 2.5 GHz for regolith The angular spread is given by a Gaussian: where the width is frequency dependent, decreasing with n E. Zas, F. Halzen and T. Stanev, Phys. Rev. D 45 (1992)
Slide31 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Radiation properties
Slide32 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Experimental confirmation Experiment at SLAC using 3.6 tons Si target, GeV photons, total shower energy up to 1019 eV
The ZHS simulation results confirmed to a factor of two, radiation coherent and linearly polarised D. Saltzberg et al., Phys. Rev. Lett. 86 (2001)
P. W. Gorham et al., Phys. Rev. D 72 (2005)
Slide33 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 The Moon as an UHE particle detector Moon opaque to neutrinos with E > 1016 eV
UHE cosmic rays of minor interest
Target (upper 10 m) is regolith: Si grains and tiny rocks
Dielectric properties depends on TiO, FeO contaminants
Radio transparency for n < 1 GHz if 5% contaminants assumed Olhoeft and Strangway, Plan. Sci. Lett. 24 (1975)
Slide34 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Threshold energy Threshold determined from the ZHS parametrisation and the minimum detectable signal
Using Chandrayaan-1 and ELVIS parameters:
Altitude h = 100 km Centre freq. n = 100 MHz Bandwidth = 50 MHz Sensitivity Pmin = -135 dBm/Hz => Threshold neutrino energy becomes 5*1019 eV
Slide35 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Detection aperture Detection rate is given by the effective aperture (“cross section”), determined through simulations
Sensitivity dependent on:
- Primary neutrino mixing and branching ratios
- Primary energy and neutrino-nucleon cross section - Dielectric properties, attenuation in the regolith
- Surface effects (refraction, reflection etc.)
- Distance from surface to observation point (geometry)
- Measurement frequency, bandwidth and minimum signal
Monte Carlo sensitivity simulation implemented in Matlab
Simulations performed for different primary neutrino energy
Slide36 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Simulation results 2pAmoonNdetected/Ntotal
Slide37 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Model dependent event rate Using model for minimum GZK neutrino flux (Engel et al., 2001)
The event rate can then be determined from => 2.2 detectable events per year =)
Slide38 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Outline Presentation of objectives
Introductory high-energy cosmic particle physics
Radar studies of EAS ionisation columns
Lunar satellite detection of Askar’yan radio pulses
Discussion and outlook
Slide39 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Discussion, part 1 We have presented how an EM scattering problem in cylindrical geometry can be conveniently treated by the phase-integral approximation
The results have been applied to scattering of radio waves from EAS ionisation columns for determination of radar cross sections
The physics of the EAS has been modelled in a simplistic manner, which might put restrictions on the applicability of our results from this aspect
Applications of radio wave scattering in cylindrical geometry also exist for meteor trails, ionospheric striations and for lightning ionisation
Slide40 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Discussion, part 2 We have considered observation of coherent radio pulses from showers induced in the lunar regolith by UHE cosmic neutrinos
A simple simulation program has been constructed for estimating the efficiency of a satellite experiment with this purpose
This program offers generous possibilities for further variation of different experimental parameters
Porting to a faster code (e.g. FORTRAN) is desirable before more extensive simulations can be performed
We believe that optimisation might further increase the 2.2 events to something really useful, although 2.2 yr-1 is still competitive (!)
Slide41 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 THE END Thank you for listening! Special thanks to everyone who has supported me during this work: Bo Thidé, IRF-U Jan Bergman, IRF-U Gunnar Ingelman, THEP, UU John A. Adam, ODU (Norfolk, VA) Fellow diploma students at IRF-U and elsewhere All members of the friendly staff at IRF-U
Slide42 : Oscar Stål os@irfu.se IRFU seminar Uppsala, 7/12/2005 Surface effects Radio waves refracted at surface
Beam becomes wider Better opening angle, but lower field strength
Total internal reflection, qTIR = /2 – c
TIR more important at high frequencies
CR detection supressed since all rays are down-going
Smooth surface or detailed topographic map ' Regolith n ≃ 1.7 Vacuum n = 1