Andreas Horneffer for the LOPES Collaboration Detecting Radio Pulses from Air Showers with LOPES

LOPES (LOFAR Prototype Station) prototype of a LOFAR station frequency range of 40 – 80 MHz set up at the KASCADE-Grande site 10 antennas in the first phase, 30 antennas in the second phase Goals: develop techniques to measure the radio emission from air showers determine the radiation mechanism of air showers calibrate the radio data with theoretical and experimental values from an existing air shower array

Hardware of LOPES

LOPES-Antenna short dipole with “inverted vee shape” beamwidth 80°-120° (parallel/ perpendicular to dipole)

Hardware of LOPES Receiver Module direct sampling of the radio signal with minimal analog parts: amplifier, filter, AD- converter sampling with 80 MSPS in the 2nd Nyquist domain of the AD-converter

Hardware of LOPES Memory Buffer aka. TIM-Module (Twin Input Module) uses PC133-type memory memory for up to 6.1 seconds per channel pre- and post-trigger capability

Hardware of LOPES10 Clock & Trigger distribution board 1 master & 3 slave boards master board generates clock and accepts trigger slave boards distribute clocks and trigger

LOPES: Setup & Status 30 antennas running at KASCADE (10 antennas in first phase: LOPES10) triggered by large event (KASCADE) trigger (10 out of 16 array clusters) offline or online correlation of KASCADE & LOPES events KASCADE provides starting points for LOPES air shower reconstruction core position of the air shower direction of the air shower size of the air shower Averages 2500 – 3000 events per full day ca. 10 GByte uncompressed data per day  ca. 3.6 TByte per year

Data Processing steps of the data processing: 1. instrumental delay correction from TV-phases 2. frequency dependent gain correction 3. filtering of narrow band interference 4. flagging of antennas 5. correction of trigger & instrumental delay 6. beam forming in the direction of the air shower 7. optimizing radius of curvature 8. quantification of peak parameters

Delay correction TV-transmitter with picture- and two sound carriers relative phases between antennas lets us correct for delay errors delay corrections residual delays

Gain calibration LNA (dark blue), cable (green), receiver Module (red), total (light blue) measured the gain of the different parts “in the lab” combined all to a frequency dependent gain curve biggest uncertainty: match of antenna to LNA

Digital Filtering raw data: filtered data: blocksize: 128 samplesblocksize: 64k samples power spectrum:

Beamforming Electric field and power after time shifting Electric field and power before time shifting

Radius of curvature plane wave doesn’t fit the data sphere with finite radius of curvature is better makes radio data sensitive to position of shower center Plane Wave Sphere

Event Discrimination criteria for “good” events: existence of a coherent pulse position in time of pulse uniform pulse height in all antennas selection currently done manually Good Event Bad Event

Example events 1 Antenna Data Formed Beam Pulse undetected by program Good Event Bad Event

LOPES10 Data LOPES10 ran from January to September thousand events total used selection for further study: KASCADE array processor didn’t fail distance of the core to the array center < 91m shower size (number of electrons) > 5e6 or truncated muon number > 2e5 → 412 events

Detected Events 228 out of 412 events considered good Fraction of “good” to “bad” events increases with increasing Muon number and increasing geomagnetic angle → fraction also increases with zenith angle

Dependencies: Geomagnetic Angle Divided pulse height by muon number Fit results to the cosine of the geomagnetic angle Fit exponential decrease to distance

Divided pulse height by the results from previous fits. Added undetected events with height 2σ Only little dependency on electron number Power law is a good fit for muon number or energy Dependencies: Size, Nµ trunc and Energy

Scaling with Energy at Different Selections slope = 0,66±0,05 slope = 0,78±0,08 slope = 0,92±0,14 compared standard selection to two (sub-)selections with no undetected Events slope of power law fit changes, but not significantly Good Events Rmax < 130m Good Events Rmax < 70m All Events

Summary LOPES measured radio pulses from air showers for the first time since 30 years with digital filtering and beam forming these radio pulses can be measured even in a radio loud environment measured radio pulse height depends on the angle to the geomagnetic field radio pulse height correlates well with Nµ and energy and not so well with Ne radio can give useful complementary information for air shower analysis additional value for energy and mass determination independent direction measurement position of the shower center

Radio Emission from Air Showers air showers are known since 1965 to emit radio pulses radiation probably due to geomagnetic emission process measuring the radio emission from air showers could give several benefits: large collecting area for low costs higher duty cycle than fluorescence telescopes effective RFI suppression allows measuring in polluted (populated) areas data integrated over the shower evolution, can be complementary to particle detectors this can be achieved by new digital radio telescopes

LOFAR digital radio interferometer for the frequency range of MHz array of 100 stations of 100 simple antennas fully digital: received waves are digitized and sent to a central computer cluster digital radio interference suppression ability to store the complete radio data for a short amount of time this allows to form beams after a transient event has been detected, combining the advantages of low gain and high gain antennas LOFAR will be a good tool to measure the radio emission from air showers