Antenna development and calibration for measurements of radio emission from extensive air showers at the Pierre Auger Observatory

In the atmosphere, high-energy cosmic rays interact with air molecules and create air showers of secondary particles. The cosmic-ray properties, e.g., its energy, can be determined by observing such air showers by stochastic measurements of secondary particles at the ground as well as directly by measuring fluorescence light emitted from air showers. Both detection techniques are established at the Pierre Auger Observatory. The measurement of radio emission is a complementary detection technique to investigate cosmic rays. The Auger Engineering Radio Array (AERA) is the radio detector of the Pierre Auger Observatory. Here, cosmic rays are observed with a duty cycle close to 100 %. The atmosphere is transparent to radio waves which enables measuring the cosmic-ray energy potentially with reduced uncertainties compared to those made with the fluorescence detection technique. The radio antennas used in the measurements and an accurate description of their frequency and directional-dependent sensitivities are of central importance for observing radio emission of air showers and for reconstructing cosmic-ray properties accurately. This so-called antenna response pattern is described by the formalism of the vector effective length (VEL) and is decomposed in a horizontal and in a meridional component .The antenna VEL is obtained through an absolute antenna calibration. The uncertainty of the antenna calibration has been identified as the dominant uncertainty in reconstructing the cosmic-ray energy and thus, a reduction of the uncertainties is most desirable. In this thesis, an absolute antenna calibration of the radio stations equipped with logarithmic-periodic dipole antennas (LPDAs) has been performed. Challenges of measuring cosmic-ray induced air showers at AERA are the variations in local weather conditions. During the day, temperature fluctuations of more than 20°C are not uncommon. These fluctuations have a direct influence on the signal processing of the radio station electronics, e.g., the low-noise amplifier (LNA) connected to the antenna. The LNA and the antenna form a technical unit and thus, it is included in the antenna response pattern. An accurate description of the LNA and its dependency on the temperature is therefore of great interest concerning the antenna calibration as well as data reconstruction. To measure the temperature dependency of the LNAs implemented in LPDA-equipped radio stations, an LNA was positioned in a temperature-adjustable and PC-controlled fridge. The LNA power gain has been measured as function of the frequency for several different temperatures. A temperature dependency of the LNA power gain of -0.017 dB/K in the median with nearly no dependency on the frequency has been found. These results have been reproduced in a further measurement of a second LNA. The ground conditions in the field described by the ground permittivity and ground conductivity and their variations, e.g., due to rainfall are further important environmental conditions influencing the antenna response pattern. Both, ground permittivity and ground conductivity, have a direct influence on the signal reflection on ground and thus impact the antenna response pattern. It was found that the signal reflection on ground with typical ground conditions at AERA is mostly dependent on the ground permittivity. A measurement setup has been developed to measure the frequency-dependent permittivity of soil from the AERA site using a capacitor. Two measurements campaigns have been performed to obtain typical values of the ground permittivity at AERA. In the first campaign, soil samples from the AERA site have been collected and the ground permittivity has been measured in the laboratory with respect to the soil humidity. It was found that the ground permittivity is typically larger than 2 and increases with the soil humidity. In the second campaign, the ground permittivity has been measured directly at different positions within AERA to investigate the ground homogeneity at the site. For the majority of the positions, the measurements exhibit values from 6 to 7. From the measurement campaigns, typical values of the permittivity between 2 and 10 have been found depending on the soil humidity. Using the NEC-2 simulation code, an overall LPDA response pattern in terms of the respective VEL component with standard environmental conditions for all arrival directions and frequencies ranging from 30 MHz to 80 MHz has been simulated. From multiple simulations the influence of changing environmental conditions such as changing ground conditions, conductive elements close to the antenna and an antenna misalignment on the LPDA response pattern have been investigated. In the case of an applied ground permittivity of 2 and of 10, which is based on the permittivity-measurements campaigns, the influence on the horizontal (meridional) LPDA VEL component is at the level of 1.3 % (0.5 %) in the median and impacts the whole phase space defined by the frequency and zenith angle. Simulations of an electronics box beneath the LPDA exhibit an influence at the level of 0.3 % (1 %) on the horizontal (meridional) VEL component of the LPDA. In the case of the meridional component, especially the regions with large zenith angles are affected where the LPDA is less sensitive. The simulations yield that a misalignment of the LPDA of 1° as well as different ground conductivities do not change the LPDA response pattern neither in the horizontal component nor in the meridional component. In a calibration campaign, the LPDA response pattern has been measured at the AERA site using a remotely piloted octocopter. The octocopter was used to place a calibrated source at different positions around an LPDA. The calibrated LPDA is representative for all LPDAs as they are mechanically and electrically identical on the percent level. All devices used in the measurements have been quantified concerning their uncertainties. Special focus was put on the accuracy of the position reconstruction of the calibrated source. The accuracy has been increased by combining the position information measured by the octocopter build-in sensors such as GPS, barometer, and inclination sensors with a newly developed optical method consisting of two cameras. To ensure reproducibility, the measurements have been repeated at different days with different environmental conditions. The measurements have been finally combined according to their uncertainties resulting in an overall uncertainty for the horizontal component of the LPDA VEL of 7.4 %, and for the meridional component of 10.3 %. These uncertainties are to compared with the uncertainty of a previously performed balloon-based calibration in which a smaller phase space was measured and an uncertainty of 12.5 % for the horizontal component was achieved. The accurate measurements have been compared with NEC-2 simulations considering the antenna, the LNA as well as the transmission antenna. Correction factors, which are defined as the ratio between measured and simulated patterns for each frequency and arrival direction, have been calculated and the median of all correction factors has been found to be close to unity varying about 0.12 (0.26) for the horizontal (meridional) component. The influence of the modified LPDA response pattern on the data reconstruction has been demonstrated on one example event with a zenith angle of 30° and an azimuth angle of 14° south of east. The trace of the modified LPDA pattern exhibits an up to 7 % larger amplitude. The total energy fluence of all polarizations increase at the level of 9 %. The reconstructed radiation energy of the full event changes from 7.96 MeV to 8.54 MeV. Finally, the antenna calibration uncertainties of the horizontal and meridional antenna VEL as well as the uncertainties due to different ground permittivites are propagated to the square root of the energy fluence. The square root of the energy fluence is taken because the energy fluence scales quadratically with the electric-field amplitude and the cosmic-ray energy. Hence, the uncertainty of the square root is the relevant uncertainty in most analyses of radio data. It is found that for arrival directions with zenith angles smaller than 60°, the uncertainty in the square root of the energy fluence due to the calibration uncertainties and due to changing ground conditions is 8.8 % in the median. Typically, radio stations in air shower experiments consist of two perpendicular polarized antennas which are usually aligned horizontally. The third component, the vertical polarization of the radio signal, is not measured directly but is important for the data analysis of radio signals from cosmic ray induced air showers. So, 3D radio stations are motivated as the simultaneous measurement of all three components potentially reduces the uncertainties of the electric-field reconstruction. Furthermore, such radio stations have an increased field-of-view towards the horizon. For AERA, a new 3D radio station has been developed. The mechanical and electrical properties of the antenna are presented. In 2015, a subset of five Butterfly radio stations was extended with a vertical antenna called Wifivert antenna forming 3D radio stations. Within three month of data taking, more than 2800 events have been observed using the 3D radio stations in coincidence with the surface detector of the Pierre Auger Observatory. From these events, 45 events are identified as cosmic-ray candidates with a signal-to-noise ratio (SNR) of 10 or more measured in the horizontally aligned antennas of at least one of the 3D stations. If additionally an SNR > 5 in the vertical antenna is measured, the event is classified as a golden 3D-event. In total, 13 events are identified as golden 3D-events in the given data set. Using the Wifivert antenna, radio emission of air showers in the vertical polarization has been measured. It is shown that the antenna is especially sensitive on air showers at large zenith angles which is a focus topic of current research. Future work will profit from the results presented in this thesis. The accurate measurements of the calibration campaign will improve the reconstruction of cosmic-ray properties using data measured with AERA, e.g., the primary cosmic-ray energy. With the reduced uncertainties of the calibration, the energy scale for air shower measurements from radio detectors is directly impacted. Using 3D radio stations, the vertical polarization of the radio signal is additionally measured compared to conventional 2D radio stations which reduces the uncertainties of the electric-field reconstruction. Furthermore, the field-of-view of these radio stations is extended towards the horizon and thus, 3D radio stations improve the research of horizontal air showers