Preamble

Astronomical Techniques and Instruments, Vol. 2, November 2025, 375–387 Article Open Access Probing the cosmological 21 cm global signal from the Antarc- tic ice sheet Shijie Sun , Jiaqin Xu , Minquan Zhou , Shenzhe Xu , Fengquan Wu , Haoran Zhang Juyong Zhang , Bin Ma , Zhaohui Shang , Xuelei Chen 1,5,6* 1 National Astronomical Observatories Chinese Academy of Sciences Beijing 100101, China

2 School of Mechanical Engineering , Hangzhou Dianzi University , Hangzhou 310018, China

3 College of Physics and Electronic Engineering , Shanxi University , Taiyuan 030006, China

4 School of Physics and Astronomy , Sun Yat-sen University , Zhuhai 519082, China

5 School of Astronomy and Space Science University of Chinese Academy of Sciences Beijing 100049, China

6 Center of High Energy Physics , Peking University , Beijing 100871, China

*Correspondence:

INTRODUCTION

The redshifted 21 cm line, arising from neutral hydro- gen, is potentially observable at low radio frequencies (50– 200 MHz) and carries much information about the evolu- tionary history of the cosmic dawn and the epoch of reion- ization . The global (all-sky averaged) spectrum of the redshifted 21 cm brightness temperature can be measured with a single antenna and a wide-band spectrometer, and numerous experimental efforts are devoted to this, includ- ing the Experiment to Detect the Global Epoch of reioniza- tion Signature (EDGES ), the Broadband Instrument for Global HydrOgen ReioNization Signal (BIGHORNS the Shaped Antenna measurement of the background RAdio Spectrum (SARAS ), the Probing Radio Inten- sity at high-Z from Marion (PRI ), Radio Experi- ment for the Analysis of Cosmic Hydrogen (REACH the Large-aperture Experiment to Detect the Dark Age (LEDA ), Sonda Cosmologica de las Islas para la Detec- cion de Hidrogeno Neutro (SCI-HI ), Cosmic Twilight Polarimeter (CTP ), and Mapper of the IGM Spin Tem- perature (MIST ). There are also experiments using short-spaced interferometers, such as the Short-Spaced Inter- ferometer Telescope probing cosmic dAwn and epoch of ReionizAtion (SITARA ). Considering the measure- ment difficulty from most ground sites at low frequencies, caused by ionospheric refraction and reflection of broad- band radio-frequency interference (RFI), there are also pro- posals to make the measurement in the lunar orbit, such as the Dark Ages Radio Explorer project (DARE ), the Dark Ages Polarimetry PathfindER in low lunar orbit (DAP- ), and the Discovering the Sky at Longest wave- length project (DSL Depending on the cosmological model, the depth of the global cosmic dawn signal is ~100 mK. Measuring © 2025 Editorial Office of Astronomical Techniques and Instruments, Yunnan Observatories, Chinese Academy of Sciences. This is an open access article under the CC BY 4.0 license ( Citation: Sun, S. J., Xu, J. Q., Zhou, M. Q., et al. 2025. Probing the cosmological 21 cm global signal from the Antarctic ice sheet.

Astronomical Techniques and Instruments (6): 375−387.

Abstract

The redshifted 21 cm line, arising from neutral hydrogen, offers a unique probe into the intergalactic medium and the first stars and galaxies formed in the early universe. However, detecting this signal is a challenging task because of artificial radio-frequency interference (RFI) and systematic errors such as ground effects. The interior of the Antarctic continent provides an excellent location to make such observations, with minimal RFI and relatively stable foreground signals. Moreover, a flat plateau in central Antarctica, with an ice cap over 2 000 m deep, will show less ground reflection of radio waves, reducing the signal complexity in the area around the probing antenna. It may be advantageous to perform cosmological 21 cm experiments in Antarctica, and a 21 cm Antarctic global spectrum experiment can potentially be deployed on the Antarctic ice cap. We have performed preliminary instrumental design, system calibration, and implementation of such an instrument optimized for extreme cold and capable of long-term autonomous operation. This system shows the ability to effectively detect the 21 cm signal, confirming Antarctica as an excellent observational site for radio cosmology.

Keywords

Radio astronomy; Low frequency; Global spectrum; Antarctica

this faint signal is extremely challenging because the galac- tic foreground, which dominates the radio sky, is more than four orders of magnitude brighter than the signal to be measured. Any small system effect can either swamp the signal or produce a false signal. Extremely high sensitiv- ity and large dynamic ranges are required to discern the small 21 cm signature in the spectrum, and the 21 cm sig- nal from the cosmic dawn has an unknown but generally broad shape. Its detection requires a good understanding of the instrument response, which must be determined by a calibration procedure. Any frequency dependence in instrumental gain, noise spectrum, antenna beam shape, ionospheric effect, and ground reflection or other effects, if not properly accounted for, can affect the measurement result.

The EDGES experiment has reported the detection of a 500 mK deep absorption feature centered at 78 MHz which may be associated with the signature of the cos- mic dawn. This absorption is much stronger than the predic- tion made by the standard model, so its cosmological inter- pretation may require new physics or astrophysics mecha- nisms, e.g., a cooling mechanism with exotic dark matter particles or extra radio background . However, it has also been questioned whether this feature is real or arises from systematic errors . Measurements taken by the SARAS-3 experiment have not detected such an absorption feature The ground effect is one of the most significant system- atic factors in a ground-based experiment. The proximity of the antenna to the ground surface causes antenna- ground coupling, which affects performance by altering the impedance, radiation pattern, and frequency response of the antenna. In addition, below the antenna, the ground generally has multiple layers of soil or rock, and can reflect incoming radio waves at the interfaces of these lay- ers. A reflected wave can interfere with the direct signals from the sky, forming complicated standing waves and causing a multi-path effect that distorts the data. The ground also emits thermal radiation, which can add noise to the measurements. All of these effects also depend on the temperature of the ground, which changes over time.

Experiments have taken various measures to address these issues, such as precise modeling of the ground in the data analysis or simplifying the antenna-ground cou- pling by applying a large conducting ground screen under the antenna . Taking advantage of any special condi- tion inherent to the terrain is also a route worth exploring.

For example, the SARAS experiment team placed their antenna above a water surface , and it may also be possi- ble to hang the measurement instrument in a deep canyon.

The continental interior of Antarctica may offer an ideal site for low-frequency radio astronomy requiring extreme detection accuracy, such as a 21 cm global spec- trum experiment, on account of its remote, dry, and sta- ble environment, with the Antarctic ice helping to reduce ground reflections. Here, we investigate this possibility. In

Section 2 we describe the advantages of the Antarctic inland for low-frequency radio astronomy, with an empha- sis on global spectrum experiments. In Section 3 we present the design of a global spectrum measurement instru- ment for use in Antarctica, including the antenna, receiver, calibration mechanism, receiver case, and power supply. In Section 4 we present the electromagnetic environ- ment and ground-penetrating radar measurement results in inland Antarctica, and describe the implementation of our instrument. Section 5 gives our conclusions and dis- cusses future work.

ANTARCTICA FOR ASTRONOMY The Advantages of Antarctica for Astronomical Observations Antarctica, especially its inland plateaus (Dome A, Dome C, and the South Pole) has long been recognized as offering a number of unique advantages for astronomi- cal observations . For example, the air over inland Antarc- tica is cold, stable, and very dry, and at the high eleva- tions of the plateaus the air is also thinner; the distortion, absorption and scattering of light by the air are all signifi- cantly reduced. The absence of water vapor is particu- larly beneficial for infrared and submillimeter wave- lengths. The long polar nights which last for months allow for long, uninterrupted darkness, which is ideal for time domain astronomy projects, which require continu- ous monitoring of celestial objects. During night Antarc- tica is one of the darkest places on Earth, with effec- tively no artificial light to interfere with observations. A number of astronomical observatories have been set up there, such as the South Pole Telescope , the Back- ground Imaging of Cosmic Extragalactic Polarization (BICEP) experiment for studying the cosmic microwave

background

, the Antarctic Search for Transiting ExoPlan- ets (ASTEP ), a 40 cm optical telescope at Dome C, and the Antarctic Survey Telescope (AST3), which is a series of optical telescopes designed for wide-field sky sur- veys, at Dome A , which has the highest elevation in Antarctica. This has been investigated as one of the best sites on Earth for optical observations , as well as for far-infrared observations Here, we investigate the potential advantages of the Antarctic inland plateaus for low-frequency radio astron- omy. First, in the low-frequency band, RFI is generally severe for astronomical observations, because a variety of human technological activities generate low-frequency emissions. The remote and sparsely populated nature of Antarctica, where such activities are minimal or nonexis- tent, consequently makes it an ideal location for observ- ing faint cosmic signals. In addition, the interior of Antarc- tica is covered with a thick ice sheet, with a depth of approximately 2 000 m. This may provide relatively sta- ble and uniform ground for the 21 cm global spectrum experiment, allowing antenna-ground coupling and stand-

ing wave effects can be reduced. Finally, toward the geo- graphic pole, most sidereal motion is apparent as rotation of the sky, with fewer celestial objects rising above or set- ting below the horizon, thus providing a more stable sky for global spectrum measurements. These potential bene- fits may make the Antarctic inland plateaus one of the best terrestrial sites for 21 cm global spectrum measure- ment.

Here, we look into the stable sky signal (the RFI envi- ronment and the ice sheet ground will be discussed in fur- ther detail in Section 3). Antennas used for measuring radio signals have some chromatic effect, because the antenna beam profile changes with frequency. The anisotropy of the sky foreground can cause spectral varia- tions that affect global spectrum measurements and need to be corrected according to the sky model and antenna beam, which is a major source of error in the cosmic dawn 21 cm experiment. One can attempt to minimize the chromatic effect through antenna design, but in practice it is almost impossible to completely eliminate. In the vicin- ity of the South Pole, the South Celestial Pole is close to the zenith, so that most of the visible sky area is located within the circle of perpetual apparition, and there is no obvious ascent or descent of most celestial bodies. When an antenna of wide beam is used, especially if the center of the field of view is pointing at the zenith, there is lit- tle change in antenna temperature with time. shows the beam-averaged sky signal as a function of time at sev- eral latitudes. At 44°N (e.g., a site in Xinjiang, China), the temperature variation curve (black line) is compli- cated, and the reduction of such a foreground would require an accurate model of the sky and a highly pre- cise antenna beam, which is difficult to achieve in prac- tice. At 75°S (blue line), the variation of the beam-aver- aged temperature is much simpler and its amplitude is only about 1 dB. At the South Pole, there is no variation in the signal, as apparent motion is solely rotation for all objects except the Sun and planets, with targets never ris- ing or setting. As a result, provided that the system has cylindrical symmetry around its vertical axis, it is almost immune to chromatic error.

Expedition, Site Selection and Implementation Despite the many advantages of the Antarctic sites, there are many challenges when doing astronomy in Antarc- tica. Notably, it is extremely cold in Antarctica, with tem- peratures falling below –80°C, which makes equipment maintenance and human operation difficult, and puts high requirements on instrument reliability. Antarctic sites are also very remote and isolated, meaning that instruments must rely heavily on automated systems, and have a self- sufficient energy supply. Finally, and perhaps most impor- tantly, inland Antarctica has extremely limited physical accessibility, and it is not always possible to transport peo- ple and equipment to such sites.

The Chinese National Antarctica Research Expedi- tion program offers a precious opportunity to conduct a

12-02 12:00 12-03 12:00 12-01 12:00 12-02 00:00 12-04 00:00 12-03 00:00 Datetime low-frequency radio astronomical experiment in inland Antarctica. Several scientific research stations have been set up on the Antarctic continent to provide facilities for scientific research. Kunlun Station, first established in 2009, is located in Dome A (80°25'01"S, 77°06'58"E), the highest plateau in Antarctica (see ). Many scientific instruments for astronomy and other scientific disciplines are installed at the site, and an inland expedition visits this site each year, during the southern hemisphere sum- mer, to install new instruments and maintain or upgrade the old ones. Equipment for a 21 cm global spectrum exper- iment on the Antarctic inland ice cap can be installed dur- ing such an expedition.

Zhongshan Station (green dot), Kunlun Station (red dot), and the site of our 21 cm global spectrum experiment (blue dot) at (78°32'11"S, 77°0'50"E).

Our site selection is guided by three primary princi- ples: the chosen location should be situated on the Antarc- tic inland ice cap, with an ice thickness exceeding 100 m; the site should be characterized by flat and open terrain; and it should not be in close proximity to pre-existing installations on the Antarctic ice cap, such as meteorologi- cal stations and geomagnetic instruments, as these may introduce electromagnetic interference.

In December 2023, the 40 Chinese Antarctic research team inland expedition departed from Zhong- shan Station (69°22'24.76"S, 76°22'14.28"E) toward Kun- lun Station (80°25'01"S, 77°06'58"E), advancing approxi- mately 100 km per day. The research team conducted site selection for our global spectrum experiment along the route, measuring RFI at flat potential sites, and surveying the ice layers using ground-penetrating radar (GPR). We have selected a site at (78°32'11"S, 77°0'50"E), approxi- mately 1 050 km from Zhongshan Station, for the installa- tion of the Antarctic global spectrum observation instru- ment (see At the selected site, the expedition team excavated a pit approximately 1 m deep with a level base. They then embedded the supporting bracket of the antenna of the global spectrum experiment into the pit and buried the bracket with ice after the antenna and the receiver were installed. The installed antenna is shown in Antarctic global spectrum measurement instrument.

To supply power for the operation of the instrument, we installed a set of solar panels, which are placed approxi- mately 500 m from the antenna itself, to avoid affecting any measurements. There are 12 solar panels, with sets of three connected in series to provide 100 V each, making four groups connected by shunt. Given the prevalence of strong winds on the Antarctic ice cap, these solar panels were strategically aligned parallel to the wind direction to mitigate snow accumulation. The configuration of the solar panel array is shown in We selected spiral lead-acid batteries for energy stor- age, with a relatively wide operating temperature range.

Each has a voltage of 12 V and capacity of 100 Ah. We connected eight batteries in series to achieve a total out-

put of 96 V. This battery configuration provides a cumula- tive energy capacity of 9 600 Wh. During polar day, the observation equipment can operate continuously, using solar energy, while during polar night, the batteries can sus- tain operation for approximately 5 days, based on the expected power consumption of 60 W.

Complete installation of the equipment required approx- imately 8 h of dedicated work on-site, with a few addi- tional hours spent testing the installed instrument. The sys- tem was successfully powered up, and internal calibra- tion procedures were completed before the expedition departed for Kunlun Station. Following 20 days of scien- tific work, the research team returned from Kunlun Sta- tion to Zhongshan Station, traversing over 100 km daily.

On the second day of the return route, the team arrived at the designated experiment site for a scheduled opera- tional window. Within scheduling constraints, field activi- ties were limited to a critical 4-hour time frame for instru- ment status verification.

In the following two sections, we describe the design of the Antarctic global spectrum instrument and the results of the experiment.

INSTRUMENT DESIGN

The Antarctic global spectrum measurement instru- ment consists of an antenna, a wide-band receiver includ- ing the analog front end and the digital backend, and auxil- iary systems which supply power. Aside from the usual amplifying channels, the receiver also needs a calibration subsystem, and a thermal control subsystem. The antenna uses an elliptical dipole configuration, with a smooth response needed for 21 cm global spectrum measure- ments, and also provides a natural enclosure for the receiver. The receiver is housed inside the elliptical cylin- der dipole antenna. The received signals are amplified, digi- tized, and processed to produce the spectral data, which is stored on a hard drive for further analysis. To account for system drift over time, the instrument incorporates a self- calibrating mechanism that periodically alternates between calibrators and antenna according to a preset schedule.

Unlike other global spectrum experiments, an instrument setup in Antarctica requires a special design for fully auto- matic operation, extreme cold environment, and the absence of maintenance for an entire year.

Antenna The 21 cm global spectrum measurement generally requires a wide beam antenna with low chromaticity.

According to the standard cosmological model, the 21 cm signal from the cosmic dawn era is probably redshifted to the frequency range of 50–100 MHz, which is the observ- ing band. Correspondingly, the physical size is about half the wavelength, i.e., 2 m.

The instrument is intended for deployment in the harsh environment of high plateaus in Antarctica, where the average altitude is 2 350 m above sea level, and up to

4 000 m near Dome A. The temperature during the polar day is around –30°C, and during polar night can drop to –80°C. The transportation and installation of a large, wide-band antenna in such conditions present significant logistical challenges. Therefore, the antenna design must incorporate a lightweight and modular structure for ease of transportation and installation.

Based on these considerations, the following design principles were established. (1) Low chromaticity. The antenna should maintain an almost invariant beam response across different frequen- cies within the observing band, which is crucial for detect- ing the faint 21 cm signal in a wide-band system. (2) Lightweight and integrated design. Given the harsh environment and difficulty of working conditions, the antenna should be lightweight and easy to transport and assemble. With this in mind, we designed the antenna to also house the receiver electronics, so that it is unneces- sary to have another separate electronics box; this also avoids a complicated balun structure.

(3) Electromagnetic self-shielding. As the antenna also serves as the enclosure of the electronics, it must have robust electromagnetic self-shielding to prevent inter- ference from the receiver and power supply, where numeri- cal digital circuits are necessitated.

An elliptical dipole antenna satisfies these require- ments. A planar dipole antenna was positioned horizon- tally above ground at a height of approximately one-quar- ter the central wavelength, with its maximum gain directed toward the zenith. Instead of a planar blade dipole, we adopted a pair of elliptical cylinders with finite thickness, the inside of which can also house the receiver.

The top panel of shows the elliptical dipole antenna model. The elliptical shape was chosen due to its smooth edge transitions, which allows for the uniform flow of current along the sides of the antenna. This results in a beam that points toward the zenith, and varies very little across the observing frequency band, as well as ensuring good impedance-matching (shown in the bottom panel of For optimal performance, this antenna is supported by a non-conducting frame above the ground. Given that wind speeds during Antarctic blizzards can reach up to , a supporting frame with exceptional strength and rigidity is required to maintain stability without affect- ing the electrical properties of the antenna. To meet these requirements, the support frame is constructed primarily from glass fiber reinforced plastic (GFRP), which offers a tensile strength of approximately 3

500 MPa, while also

exhibiting lower electrical conductivity, thereby minimiz- ing any interference with electrical performance of the antenna.

The overall structure of the support frame (shown in ) is divided into two symmetrical sections, left and right, connected at the base by large wooden planks. The top surface of the frame serves to support the antenna.

Since the left section of the frame houses the receiver sys- Far field

Frequency = 70.5 MHz Frequency = 80.4 MHz Frequency = 59.7 MHz Frequency = 100.2 MHz Frequency = 49.8 MHz Frequency = 90.3 MHz 0°

70.980 1° 70.704 4° 68.012 8° 67.376 1° 66.481° 66.234 1° Bottom: The simulated far-field beam of the antenna for 6 frequency points in the range of 50–100 MHz, and their corresponding 3 dB beamwidths. 840 mm 1 000 mm tem chassis, additional square GFRP tubes are added to pre- vent deformation. The total height of the support frame is 1840 mm, with 840 mm of the upper section exposed above the ice surface to ensure optimal antenna perfor- mance. The lower portion, 1 000 mm in height, is buried in the ice to provide stability against wind. This design pre- vents the antenna, which has a large surface area, from tip- ping over in extreme weather conditions. Additionally, the wooden planks at the base of the support frame prevent the entire antenna structure from sinking into the snow or

ice over time. This structural design ensures both mechani- cal stability and long-term functionality of the antenna sys- tem in the harsh Antarctic environment.

Receiver The overall structure of the receiving system is shown in . The signal from the receiving antenna is sent to the analog receiver system. The analog or calibra- tion signal is sent to the digital spectrometer, where it is digitized and processed, then saved in the data storage sys- tem. A more detailed description of our global spectrum Filter Amplitude detector Phase detector Filter Power splitter VNA subsystem Analog receiver system Digital spectrometer LNA: The custom-made LNA is optimized for low noise, and low reflection coefficients for both input and out- , which effectively reduce the reflected noise between antenna and receiver. The , and the transmission coefficient of the LNA ( ) are plotted as a function of frequency. Measurements of the LNA show an average of –45 dB, and the average is –30 dB. The LNA offers a typical gain of 37 dB with a flatness of 0.2 dB.

Magnitude/dB Magnitude/dB

Phase/(°) Phase of Frequency/MHz , and the bottom panel shows the transmission coefficient, receiving system can be found in the reference [ Analog receiver The analog receiver comprises several subsystems and components. The low-noise amplifier (LNA), and asso- ciated filters, to amplify the signal; the calibration subsys- tem and vector network analyzer (VNA) subsystem, which are used for calibration; the radio frequency (RF) switch matrix, which switches the input between antenna, calibrator, and VNA, and; the thermostatic subsystem, which maintains the temperature of the receiver.

Variable temperature load Load calibrator Short calibrator Open calibrator Shorted long cable Open long cable receiver Calibration subsystem Thermostatic subsystem (1) Calibration subsystem. The calibration subsystem uses a variable temperature load to calibrate the tempera- ture drift, a noise diode, an open long cable, a shorted long cable and the VNA calibrators. The RF switch matrix is employed to control signal paths, ensuring effi- cient transitions between calibration and measurement modes.

This receiver channel can be switched to one of six inputs: the antenna for observation of the sky signal; a noise diode to calibrate the relative change of the system gain; the variable temperature load, which can be heated to 400 K, for absolute calibration, with an accuracy exceed- ing 0.1 K; the calibrators for calibration of VNA character- istics; an open long cable; and a shorted long cable used to solve the noise wave parameters, , and of the system. (2) VNA subsystem. A VNA is used for precise impedance calibration of both the antenna and receiver.

Impedance mismatch between the antenna and receiver can lead to signal reflections and standing waves, which introduce additional noise. During self-calibration, the VNA subsystem facilitates real-time measurement of the antenna and the receiver impedance. This real-time mea- surement enables a more precise determination of the sys- tem response.

Antenna RF switch matrix storage system

(3) Thermostatic subsystem. A 21 cm global spec- trum measurement requires high stability. To achieve this, the analog front end is integrated with the thermostatic sub- system. Active components, including amplifiers, are housed within small, precision-controlled enclosures that use a 4 × 4 cm thermoelectric cooler to maintain the tem- perature to within ±0.1°C. The thermostatic subsystem uses a purely analog circuit to avoid possible electromag- netic interference (EMI). It is controlled by an analog Pro- portional-Integral-Differential circuit. With components exhibiting minimal temperature drift, the front-end gain vari- ation is maintained within 0.03 dB. We set the equilib- rium temperature at –30°C, since the daytime tempera- ture in polar day is below –20°C, and the “nighttime” tem- perature is about –30°C. When the thermostatic subsys- tem is cooling, the peak power consumption is less than 30 W, and in a stable state, the power consumption is 3 W.

Digital spectrometer The digital spectrometer is designed for high-preci- sion, wide-bandwidth sampling, characterized by low power consumption (approximately 20 W) and robust ther- mal stability. The digital board employs a Xilinx Kintex7 high-performance Field-Programmable Gate Array and a 12-bit, 500 MSPS TI ADS5407 chip to provide high- dynamic range sampling. The system achieves a fre- quency resolution of 8 192 channels, thereby ensuring ade- quate frequency resolution and dynamic range. It has been tested to function effectively at temperatures as low as –40°C.

The receiver system integrates self-calibration mecha- nisms to ensure high sensitivity and accuracy. shows the calibration result when an antenna simulator (load) is connected. After an integration time of 2 h, it can measure the load physical temperature (in this case equal to ambient temperature) to within 0.1 K. We esti- mate that our instrument has a sensitivity of 0.1 K across the entire observation band. The instrument specification is shown in Recovered temperature Smoothed Temperature/K Frequency/MHz Receiver Chassis

Both the analog front end and digital backend of the receiver, along with the power supply unit, are housed within a chassis integrated into the body of the elliptic dipole antenna. The components of the analog section include variable temperature loads, RF switch, RF switch

Specifications Value Frequency range/MHz Gain/dB Gain stability/dB Out-of-band rejection/dB Sensitivity@30–200 MHz

0.1 K

Sensitivity@50–100 MHz

0.04 K

Sampling rate/MSPS Sampling precision/bit Frequency resolution/kHz logic control board, solid state relays, and a compact ther- mostat enclosure for the LNA, noise source, and the sec- ond-stage amplifier. The digital section comprises digital acquisition boards, frequency oscillator source, tempera- ture measurement unit, USB hub, serial hub, VNA, tempera- ture controllers, and a microcomputer. The overall layout for the receiver system is shown in . The digital section is denoted in green, the analog section in blue, and the power supply section in red.

The receiver chassis requires careful electromagnetic compatibility design to ensure effective shielding for each section. Furthermore, the RF devices are sensitive to tem- perature variations, necessitating a dedicated thermostatic design. The chassis contain three separate compartments, housing the digital, analog, and power supply sections, as shown in . This compartmentalized design miti- gates EMI by the digital and power supply component on the sensitive analog component. The arrangement of each component within the chassis is optimized based on size and functionality. The physical dimensions of the receiver case are 700 mm × 700 mm × 180 mm. There are four cav- ities between the switching power supply and the elec- tronic equipment, and multistage low-pass EMI filters are inserted into these cavities. The external connections of the whole chassis are limited to three cables. One coaxial line is for the antenna feed, which transmits the signals coaxial line is for external data transmission, allowing for external communication with the microcomputer of the sys- tem. Lastly, there is a cable to the power supply section from the battery.

The antenna body must be rigid, to maintain its shape and avoid deformation, while also needing to be lightweight to facilitate transportation and installation in the harsh Antarctic environment. Our design is based on a thin shell with internal reinforcement. The overall struc- ture of the antenna is shown in . The receiver case is placed in one of the dipoles, further reinforced with crossbars.

Power Supply Laboratory tests show that the power consumption of the Antarctic global spectrum measurement instrument is approximately 60 W in common observation mode. We also deployed an instrument of similar design in Xinjiang,

1 A

Side view. The digital section is shown in the green shaded area. 1: Oscillator; 2: Digitizer; 3: USB hub; 4: Serial port hub; 5:

Temperature controller; 6: VNA; 7: Fan; 8: Microcomputer; 9:

Temperature controller. The power supply section is denoted in the red shaded area. 10: DC-DC converter; The analog section is denoted in the blue shaded area. 11: Variable temperature load; 12: 40-pin connectors; 13: Solid state relay; 14: 1P12T RF switch; 15: 1P2T RF switch; 16: Noise source; 17: Low-noise amplifier; 18: Second-stage amplifier.

China, with operating results indicating that the power con- sumption is nearly 60 W in either winter or summer (when the average temperatures are –20°C and 30°C, respectively). To provide energy, the solar panels charge the batteries during polar day, via a solar controller, which subsequently provides power to the device. During polar night, the device enters a dormant state upon bat- tery depletion and is programmed for automatic activa- tion at the onset of polar day to resume observations. The proposed power supply configuration for the instrument is illustrated in . The solar controller features three inputs/outputs, corresponding to the solar panel, battery, and load.

At the site, which has a latitude of ~80°S, the Sun does not rise more than 35° above the horizon, so we installed the solar panels in a vertical orientation. The stan- dard output voltage for each panel is 50 V with the pan- els connected in series, in groups of three, yielding a com- bined output of 150 V. The selected solar controller oper-

4 Support

frame

6 Glass fiber

support Solar panel

96 VDC battery

EM shielding box Filter Filter Filter Solar controller 500 m cable 96–12 VDC transformer Filter Receiver power supply Ripple attenuation moudule 12 VDC output to receiver Solar controller enclosure. ates at a mean charge voltage of 113.6 V (maximum 230 V) to charge the batteries.

A significant challenge associated with this energy sup-

ply scheme is that both the solar controller and the DC- DC unit for voltage conversion produce low-frequency EMI, which could severely affect our experiment. To miti- gate this issue, electromagnetic shielding enclosures have been designed to house any components prone to generat- ing electromagnetic radiation, as shown in . Addi- tionally, the energy supply unit is positioned 500 m away from the antenna, to minimize the potential impact of Within the power supply section of the system chas- sis, the current first passes through a filter, after which a DC-DC module converts the high voltage of 96 V down to the 12 V required by the equipment housed within the chassis. A ripple suppressor is applied to further refine the output current to pure DC. For active components within the analog front end, such as the LNA and noise calibration sources, additional linear voltage regulators are employed for voltage conversion at each individual component.

Design and Implementation for Extreme Environ

The environmental temperature of the Antarctic inland is approximately –40 to –30°C during polar day, and the lowest temperature can reach –70°C during polar night. Providing energy during the polar night is difficult, because no solar power is available during the long polar winter when the Sun is below the horizon for several months. We therefore design our instrument to be able to survive unpowered during polar night, but still operate when power is restored. To ensure our instrument func- tions normally in the extreme Antarctic environment, con- siderable design, improvement, testing, and experimental work have been done.

(1) Material selection. Many of the modules in the instrument are custom-designed, including the LNA, fre- quency source, and RF filters. These are constructed using military-grade chips, with operating temperature ranges of –55°C to 150°C, suitable for the extreme Antarc- tic environment. For the cables, we used ultra-flexible sili- cone rubber cables, with a low-temperature tolerance of –60°C. For the batteries, we selected a variety with the highest available charge and discharge efficiency at low temperatures (see Section 3.4). (2) Reinforcement. Various commercial products used in our experiment, such as the microcomputer, solar con- troller, and serial port controller, were reinforced by replac- ing some of their components. For example, we replaced electrolytic capacitors, which are prone to damage at low temperatures, with solid state capacitors. (3) Thermal insulation. Some modules only have good performance at standard temperatures, such as microwave switches, mechanical calibrators, hard disks, and batteries. We have taken thermal insulation measures for these modules, wrapping the cavity where these mod- ules are placed with special thermal-insulating cotton. (4) Laboratory test. We conducted low-temperature tests on most of the modules in a high- and low-tempera- ture chamber. These modules were tested under operating conditions, with the chamber set at –50°C. To verify that these modules will not be damaged during polar night, specifically at a temperature of –70°C, we powered off the modules, then set the chamber to –70°C. When the tem- perature rose to –50°C, we powered them on again to check whether they could be successfully rebooted. (5) Field experiment. We have also conducted sys- tem-level tests by deploying an identical set of equip- ment in the electromagnetic quiet zone at Hongliuxia Sta- tion in Xinjiang. The temperature there during winter can drop below –20°C. This is still much higher than Antarc- tic temperatures, but is suitable for some preliminary testing.

RESULTS

RFI Measurement We measured the RFI along the route from Zhong- shan Station (69°22'24.76"S, 76°22'14.28"E) to Kunlun Station (80°25'01"S, 77°06'58"E). At each stop of the day, RFI measurements were taken at a distance of approxi- mately 2 km from the camp of the convoy.

The RFI is measured with a spectrum analyzer, using a rod antenna for lower frequencies (10 kHz–30 MHz), and a biconical antenna for higher frequencies (30–300 MHz), as shown in . The RFI is measured to a spectral res- olution of 1 kHz, in three orthogonal directions by each antenna: vertical, north-south horizontal, and east-west hori- zontal.

Unfortunately, measurements can only be taken near the convoy camp, and for safety reasons, convoy activi- ties inevitably generate significant RFI, from sources such as the tractor engine, computer and communication equip- ment, and other instruments. Consequently, measured results may overestimate the true RFI values. Major interfer- ence, measured in four typical locations, is distributed at frequencies below 30 MHz, while there is little RFI in the 30–400 MHz region (see ). Even with the convoy present, the RF environment is excellent in the 30–400 MHz band.

Ground Effect As noted in Section 1, ground effects have signifi-

Amplitude/dBm

V H 0° H 90°

Amplitude/dBm

V H 0° H 90°

Amplitude/dBm

V H 0° H 90°

Amplitude/dBm

V H 0° H 90°

cant impacts on high-precision global spectrum measure- ments. In most locations, soil is a poor conductor, and its conductivity and dielectric constant change with tempera- ture and humidity. The conductivity of soil is also highly dependent on its water content, which increases substan- tially below the water table line and in areas with conduc- tive minerals, so that strong reflected waves can be gener- ated in underground layers, forming standing waves with the antennas located at the surface. These difficult-to-mea- sure changes ultimately cause systematic errors, produc- ing complex and difficult-to-identify effects. To address this challenge, most experimental setups employ a metal ground plane—typically a high-conductivity metallic mesh installed around the detection antenna—to mimic an ideal conductive plane for simplified analysis. However, the phys- ical size of the metallic ground plane and unavoidable envi- ronmental couplings cause constraints, so complete elimina- tion of such interference remains unachievable.

Antarctica is covered with an ice cap with an aver- age thickness of 2 km, reaching 3 km in some places.

The surface of the ice cap is flat, so that no strong reflec- tion or scattering occurs, except for the interface between the ice surface and the air. In our frequency band of inter- est, the attenuation of electromagnetic waves within the

V H 0° H 90°

V H 0° H 90°

V H 0° H 90°

V H 0° H 90°

Antarctic ice sheet is rapid. Simulations show that transmit- ted waves in the ice sheet are almost completely absorbed by a 100-m-thick ice layer, so reflections from lower lay- ers do not have a significant impact. Reflection occurring at the surface can be precisely simulated using electromag- netic simulation software, and it serves as part of the antenna response. In addition, Antarctica temperatures are extremely low, and the ice and snow surface are stable, which can alleviate or even avoid problems caused by changes in dielectric constant and conductivity on the ground.

To further investigate the effect of the ice sheet, we used GPR to probe the ice layers during the expedition.

The GPR unit was moved on the ice surface along a straight line while beaming radar waves downward. shows the ice sheet reflections.

Most reflections occur near the ground surface. In the inland ice cap, we do not observe significant structure.

The reflected waves are stronger near the surface, but that is mostly due to the fact that the incident wave is stronger there. These randomly reflected waves will not generate spectral features that could be confused with the 21 cm signal. By contrast, the seashore figure shows clear layers, which could generate confusing spectral features.

Frequency/MHz Frequency/MHz Frequency/MHz Frequency/MHz Frequency/MHz Frequency/MHz Frequency/MHz Frequency/MHz

Depth/m Distance/m In Situ Observation After installation, we powered up the system. The coefficient of the antenna and receiver measured by its internal VNA are shown in . For the antenna, the red line represents the measured values, while the black line denotes the simulated results for the antenna posi- tioned above an 80 m ice sheet. The simulation and mea- surement results are generally consistent, affirming the validity of the design. The results of the receiver measure- ment are also consistent with our expectation.

Simulation Measurement Frequency/MHz Frequency/MHz Antenna. (B) Receiver. −10 000 −20 000 −30 000 The calibrator source measurement results are shown , where the colored lines correspond to differ- ent following conditions. “Antenna” is the raw sky signal as observed. “Noise on” and “Noise off” are noise source activated and deactivated, used for relative calibration of the receiver. “High load” and “Ambient load” are spectra from high-temperature load and ambient-temperature load, used for absolute calibration of the receiver. “Open calibra- tor”, “Short calibrator”, “Long open cable”, “Long shorted cable”, “Short shorted cable”, and “Short open cable” are calibrator spectrums for calculation of noise wave parameters . “68 load” and “50 load”, used as antenna simulators to test instrument response.

The strong interference captured by the antenna at 137 MHz and 145 MHz come from satellite radio transmissions.

The shape and amplitude of these spectrums are similar to what we observe in the global spectrum experiments at the Xinjiang site , showing that the system is working normally.

CONCLUSIONS

The low level of RFI, relatively stable sky signal, and simple ground effects make the continental interior of Antarctica a promising site for precise low-frequency radio astronomy. We have carried out RFI and GPR sur- veys along the route from Zhongshan Station to Kunlun Station, and the results confirm our general expectations, showing that Antarctic sites do indeed have great poten- tial for low-frequency radio astronomy in general, and for the 21 cm global spectrum measurement in particular.

We have designed a set of instruments for the 21 cm global spectrum experiment (which comprises an ellipti- cal dipole antenna with a receiver system). These are intended to operate in the environmental conditions of Antarctica while ensuring high sensitivity and stability in Distance/m (A) The Antarctic inland region, where the thickness of the ice cap is over 2 km. (B) A site near Zhongshan Station, where the thickness of the ice cap is approximately 40 m. The color bar shows the strength of the reflected wave (ADC units).

Amplitude/dB Antenna Noise on Noise off High load Long open cable Long shorted cable Ambient load detecting the faint 21 cm signals from the early universe.

The self-calibration mechanisms indicate that the instru- ment has a sensitivity of 0.04 K in the expected fre- quency band. The system has been successfully deployed in Antarctica. Preliminary tests show resilience to the harsh environment, with a good alignment with simu- lated performance, confirming its capability to mitigate a number of frequently encountered challenges in such experi- ments, such as the ground effects. When the data from the 21 cm global spectrum experiment is returned (follow- ing the next annual expedition) and analyzed, it will fur- ther deepen our understanding of the characteristics of the Antarctic environment, and hopefully confirm that the Antarctica can provide an ideal site for low-frequency radio astronomy. Further data collection, longer observa- tions, and refinements in instrumentation are essential to achieve a higher level of detection accuracy for faint cosmo- logical signals. The results can also be compared with other sites, such as measurements taken in Xinjiang. This multi-site approach may help to mitigate site-specific sys- tematic errors and offer a more comprehensive understand- ing of the experiments.

ACKNOWLEDGEMENTS

We are grateful to the 40 th and 41 st Chinese National Antarctic Research Expedition team, supported by the Polar Research Institute of China and the Chinese Arctic and Antarctic Administration. This work is supported by the Chinese Academy of Science Key Instrument (ZDKYYQ20200008) and the National Natural Science Foundation of China (12473094 and 12273070).

AI DISCLOSURE STATEMENT AI-assisted technology is not used in the preparation of this work.

Open calibrator Short calibrator Short shorted cable Short open cable AUTHOR CONTRIBUTIONS Xuelei Chen conceived the idea and supervised the project. Fengquan Wu and Shijie Sun designed the instru- ment. Jiaqin Xu, Minquan Zhou, Shenzhe Xu and Hao- ran Zhang did the simulation work, installed the instru- ment, and carried out the laboratory calibration. Shijie Sun and Bin Ma implemented the instrument in the inland of Antarctica and conducted in situ observation. Juyong Zhang and Zhaohui Shang managed and coordinated the research activities. Shijie Sun wrote the original draft.

Xuelei Chen and Fengquan Wu reviewed and edited the manuscript. All authors read and approved the final manuscript.

DECLARATION OF INTERESTS

Xuelei Chen is the editorial board member for Astro- nomical Techniques and Instruments. He was not involved in the editorial review or the decision to pub- lish this article. The authors declare no competing interests.

REFERENCES

Madau, P., Meiksin, A., Rees, M. J. 1997. 21 centimeter tomography of the intergalactic medium at high redshift.

The Astrophysical Journal (2): 429.

Chen, X. L., Miralda-Escudé, J. 2004. The Spin-Kinetic temperature coupling and the heating rate due to Ly α scattering before reionization: Predictions for 21 centimeter emission and absorption.

The Astrophysical Journal , 602 (1): 1.

Chen, X. L., Miralda-Escudé, J. 2008. The 21 cm signature of the first stars.

The Astrophysical Journal (1): 18.

Furlanetto, S. R., Oh, S. P., Briggs, F. H. 2006. Cosmology at low frequencies: The 21 cm transition and the high- redshift Universe.

Physics Reports , 433 (4−6): 181−301.

Barkana, R. 2016. The rise of the first stars: Supersonic streaming radiative feedback, and 21-cm cosmology. [ 5 ]

Frequency/MHz

Physics Reports : 1−59.

Bowman, J. D., Rogers, A. E., Hewitt, J. N. 2008. Toward empirical constraints on the global redshifted 21 cm brightness temperature during the epoch of reionization.

The Astrophysical Journal , 676 : 1.

Bowman, J. D., Rogers, A. E. E. 2010. A lower limit of ∆ z > 0.06 for the duration of the reionization epoch.

Nature , 468 (7325): 796−798.

Sokolowski, M., Tremblay, S. E., Wayth, R. B., et al. 2015.

BIGHORNS-broadband instrument for global HydrOgen ReioNisation signal.

Publications of the Astronomical Society of Australia : e004.

Patra, N., Subrahmanyan, R., Raghunathan, A., et al. 2013.

SARAS: a precision system for measurement of the cosmic radio background and signatures from the epoch of reionization.

Experimental Astronomy , 36 : 319−370.

Singh, S., Subrahmanyan, R., Shankar, N. U., et al. 2018.

SARAS 2: a spectral radiometer for probing cosmic dawn and the epoch of reionization through detection of the global 21-cm signal.

Experimental Astronomy , 45 (2): 269−314.

T., J. N., Subrahmanyan, R., Somashekar, R., et al. 2021.

SARAS 3 CD/EoR radiometer: design and performance of the receiver.

Experimental Astronomy : 193−234. Philip, L., Abdurashidova, Z., Chiang, H., et al. 2019.

Probing radio intensity at high-Z from marion: 2017 instrument.

Journal of Astronomical Instrumentation

de Lera Acedo, E., de Villiers, D. I. L., Razavi-Ghods, N., et al. 2022. The REACH radiometer for detecting the 21-cm hydrogen signal from redshift z ≈ 7.5–28.

Nature Astronomy , 6 : 984–998.

Price, D., Greenhill, L., Fialkov, A., et al. 2018. Design and characterization of the Large-aperture Experiment to Detect the Dark Age (LEDA) radiometer systems.

Monthly Notices of the Royal Astronomical Society (3): 4193−4213.

Voytek, T. C., Natarajan, A., Jáuregui García, J. M., et al.

2014. Probing the dark ages at z~20: The SCI-HI 21 cm all-

sky spectrum experiment. The Astrophysical Journal (1): L9.

Nhan, B. D., Bordenave, D. D., Bradley, R. F., et al. 2019.

Assessment of the Projection-induced Polarimetry Technique for Constraining the Foreground Spectrum in Global 21 cm Cosmology.

The Astrophysical Journal , 883 (2): 126.

Bustos, R., Monsalve, R. 2021. Mapper of the IGM Spin Temperature. Available from mist/ Thekkeppattu, J. N., McKinley, B., Trott, C. M., et al. 2022.

System design and calibration of SITARA—a global 21 cm short spacing interferometer prototype.

Publications of the Astronomical Society of Australia : e018.

Burns, J. O., Bradley, R., Tauscher, K., et al. 2017. A space- based observational strategy for characterizing the first stars and galaxies using the redshifted 21 cm global spectrum.

The Astrophysical Journal (1): 3. Burns, J., Bale, S., Bradley, R., et al. 2021. Global 21-cm cosmology from the farside of the Moon. arXiv: 2103.

Chen, X. L., Yan, J. Y., Deng, L., et al. 2020. Discovering the Sky at the Longest wavelengths with a lunar orbit array.

Philosophical Transactions of the Royal Society A:

Mathematical, Physical and Engineering Sciences Chen, X. L., Yan, J. Y., Xu, Y. D., et al. 2023. Discovering the sky at the longest wavelength Mission-A pathfinder for exploring the Cosmic Dark Ages.

Chinese Journal of Space Science (1): 43−59. (in Chinese) Shi, Y., Deng, F. R., Xu, Y. D., et al. 2022. Lunar orbit measurement of the cosmic dawn’s 21 cm global spectrum.

The Astrophysical Journal (1): 32. Bowman, J. D., Rogers, A. E., Monsalve, R. A., et al. 2018.

An absorption profile centred at 78 megahertz in the sky- averaged spectrum.

Nature (7694): 67−70. Barkana, R. 2018. Possible interaction between baryons and dark-matter particles revealed by the first stars.

Nature (7694): 71−74. Yang, Y. P. 2020. The abundance of primordial black holes from the global 21cm signal and extragalactic gamma-ray background.

The European Physical Journal Plus : 690.

Berlin, A., Hooper, D., Krnjaic, G., et al. 2018. Severely constraining dark-matter interpretations of the 21-cm anomaly.

Physical Review Letters , 121 (1): 011102.

Hills, R., Kulkarni, G., Meerburg, P. D., et al. 2018.

Concerns about modelling of the EDGES data. Nature , 564 (7736): E32−E34.

Bradley, R. F., Tauscher, K., Rapetti, D., et al. 2019. A ground plane artifact that induces an absorption profile in averaged spectra from global 21 cm measurements with possible application to EDGES . The Astrophysical Journal (2): 153.

Sun, S. J, de Lera Acedo, E., Wu, F. Q., et al. 2024.

Calibration error in 21-centimeter global spectrum experiment.

Universe , 10 (6): 236.

Singh, S., T., J. N., Subrahmanyan, R., et al. 2021. On the detection of a cosmic dawn signal in the radio background. arXiv: 2112.06778 Shang, Z. H. 2020. Astronomy from Dome A in Antarctica.

Research in Astronomy and Astrophysics (10): 168.

George, E. M., Reichardt, C. L., Aird, K. A., et al. 2015. A measurement of secondary cosmic microwave background anisotropies from the 2500 square-degree SPT-SZ survey.

The Astrophysical Journal , 799 (2): 177.

Carlstrom, J., Ade, P. A., Aird, K., et al. 2011. The 10 meter south pole telescope.

Publications of the Astronomical Society of the Pacific (903): 568.

Chiang, H. C., Ade, P. A., Barkats, D., et al. 2010.

Measurement of cosmic microwave background polarization power spectra from two years of BICEP data.

Astrophysical Journal (2): 1123. Crouzet, N., Guillot, T., Agabi, A., et al. 2010. ASTEP South: an Antarctic Search for Transiting ExoPlanets around the celestial south pole.

Astronomy & Astrophysics Yuan, X., Su, D. Q. 2012. Optical system of the three antarctic survey telescopes.

Monthly Notices of the Royal Astronomical Society (1): 23−30.

Sun, T. R., Li, X. Y., Hu, L., et al. 2022. Antarctic Survey Telescope 3-3: overview, system performance and prelimi- nary observations at Yaoan, Yunnan.

Universe , 8 (6): 303.

Ma, B., Shang, Z. H., Hu, Y., et al. 2020. Night-time measurements of astronomical seeing at Dome A in Antarctica.

Nature : 771−774. Yang, X., Shang, Z. H., Hu, K. L., et al. 2021. Cloud cover and aurora contamination at dome A in 2017 from KLCAM.

Monthly Notices of the Royal Astronomical Society 3614−3620.

Shi, S. C., Paine, S., Yao, Q. J., et al. 2016. Terahertz and far-infrared windows opened at Dome A in Antarctica.

Nature Astronomy : 0001. Wu, F. Q., Xu, J. X., Zhu, J. C., et al. 2024. Progress of cosmic dawn global spectrum experiments.

Chinese Journal of Radio Science (4): 589−594. (in Chinese)

Price, D. C., Tong, C. Y. E., Sutinjo, A. T., et al. 2023. A new technique to measure noise parameters for global 21-cm experiments. arXiv: 2305.11479 .