Characterising the properties of the atmospheric emission at Teide Observatory in the 10-20 GHz range with QUIJOTE data

TL;DR

The study characterizes atmospheric emission at the Teide Observatory in the $10$--$20$ GHz range and empirically probes turbulence using QUIJOTE MFI and MFI2 data. By combining radiosonde, GPS PWV, and wind data with cross-correlation, structure-function, and cross-power spectral analyses, the authors demonstrate a water-vapour density profile that decays exponentially with a half-height near $1$ km, and a median PWV of about $3.3$ mm during 2012--2018. For high PWV, the observed structure function aligns with Kolmogorov turbulence with a slope near $5/3$; MFI2 data further support a Kolmogorov-like spectrum in the inertial range, with an outer scale around a few hundred meters. The findings imply that Teide Observatory exhibits atmospheric behavior comparable to ACT sites but with larger PWV, providing actionable insights for atmospheric corrections, reanalysis of QUIJOTE data, and planning for future Tenerife-based instruments such as the Tenerife Microwave Spectrometer.

Abstract

QUIJOTE is a CMB experiment composed of two telescopes, QT1 and QT2, located at the Teide Observatory in Tenerife, Spain. The MFI instrument (2012-2018), installed on QT1, observed the sky at four frequency bands (11, 13, 17, and 19 GHz) with one degree angular resolution. Its successor, MFI2, began operations in 2024 and operates in the same bands. This paper has two main goals: first, to characterise the atmospheric conditions at Teide Observatory to improve existing models at these frequencies, and second, to empirically characterise atmospheric turbulence using QUIJOTE MFI and MFI2 observations. This work has implications for both atmospheric physics and CMB studies and can support future reanalyses of MFI data or the preparation of upcoming instruments such as the Tenerife Microwave Spectrometer. We used data from GPS antennas, the STELLA observatory, and radio soundings to derive median profiles and distributions of key atmospheric parameters for 2012-2018. MFI data were analysed to compute atmospheric structure functions at 17 and 19 GHz and to study the correlation properties of the atmospheric signal through cross-correlation between horns at the same frequency. MFI2 observations were used to estimate the atmospheric power spectrum and compare it with the structure function derived from MFI data. The water vapour density profile follows an exponential decay with a characteristic half-height of about 1000 m. Median PWV in 2012-2018 is 3.3 mm. For high PWV conditions, the structure function agrees with the Kolmogorov turbulence model. The slope of the power spectrum also matches the model prediction, within the frequency range limited by the outer scale and instrument noise. Finally, from the correlation function, we find that atmospheric conditions remain stable for about 1-2 hours.

Figures (17)

• Figure 1: Top: brightness temperature of the atmosphere as a function of frequency between 1 and 150 GHz for different water vapour content computed with the software amPaine, using the atmospheric conditions at the Teide Observatory. The frequency domain explored by the various QUIJOTE instruments is shown in orange. Bottom: same figure zoomed in between 15 and 25 GHz. The bandpasses of the QUIJOTE MFI horn 2 and 4 are shown in arbitrary units in dark blue and purple for horn 2 at 17 GHz and 19 GHz, respectively, and in cyan and pink for horn 4 at 17 GHz and 19 GHz, respectively.
• Figure 2: Median seasonal profiles of atmospheric temperature (left), water vapour density (middle), and atmospheric pressure (right) measured by radio-sounding, spanning altitudes from 105 m to 33 km. In the three panels, the median profiles for autumn (orange), winter (deep blue), spring (red) and summer (green) are shown. The altitude of QUIJOTE is shown as a grey line.
• Figure 3: Binned median water vapour density as a function of the altitude above Güímar for the four seasons during the day (red curves) and night (blue curves) calculated using equation \ref{['eq:rho']}. The QUIJOTE altitude is denoted as a grey line. The full lines represent the binned data, and the dashed lines represent the exponential fits obtained with equation \ref{['eq:exp_decay']}.
• Figure 4: Distribution of PWV measurements recorded during each QUIJOTE-MFI wide survey observation (period 2012-2018). Each PWV value corresponds to the median PWV during a single observation. The overall median PWV for the entire wide survey is approximately $3.3$ mm. The first quartile indicates that 25 % of the observations have PWV values below 2.1 mm. Similarly, the third quartile indicates that 75 % of the observations have PWV values below 5.2 mm during the QUIJOTE MFI survey.
• Figure 5: Distribution of the wind direction (top) and the wind speed (bottom) recorded during each QUIJOTE-MFI wide survey observation (period 2012-2018). The wind velocity is measured in degrees as the direction the wind flows from north to east ($0^\circ$ is North, $90^\circ$ is East). The median wind speed was about 12 km/h.