Variations in the Surface Atmospheric Electric Field on the Qinghai–Tibet Plateau: Observations at China’s Gar Station

Abstract

The Qinghai-Tibet Plateau, known as the “third pole” of the Earth with an average elevation of approximately 4500 m, offers a unique natural laboratory for probing the dynamic behavior of the global electric circuit. In this study, we conduct a comprehensive analysis of near-surface vertical atmospheric electric field (AEF) measurements collected at the Gar Station (80.1° E, 32.5° N; 4259 m a.s.l.) on the western Tibetan Plateau, spanning the period from November 2021 to December 2024. Fair-weather conditions are imposed. The annual mean AEF at Gar is ∼0.331 kV/m, significantly higher than values observed at lowland and plain sites, indicating a pronounced enhancement in atmospheric electricity associated with high-altitude conditions. Moreover, the AEF exhibits marked seasonal variability, peaking in December (∼0.411–0.559 kV/m) and valleying around July–August (∼0.150–0.242 kV/m), yielding an overall amplitude of approximately 0.3 kV/m. We speculate that this seasonal pattern is primarily driven by variations in aerosol concentration. During winter, increased aerosol loading from residential heating and vehicle emissions due to incomplete combustion reduces atmospheric conductivity by depleting free ions and decreasing ion mobility, thereby enhancing the near-surface AEF. In contrast, lower aerosol concentrations in summer lead to weaker AEF. This seasonal decline in aerosol levels is likely facilitated by stronger winds and more frequent rainfall in summer, which enhance aerosol dispersion and wet scavenging, whereas weaker winds and limited precipitation in winter favor near-surface aerosol accumulation. On diurnal timescales, the Gar AEF curve deviates significantly from the classical Carnegie curve, showing a distinct double-peak and double-trough structure, with maxima at ∼03:00 and 14:00 UT and minima near 00:00 and 10:00 UT. This deviation may partly reflect local influences related to sunrise and sunset. This study presents the longest ground-based AEF observations over the Qinghai–Tibet Plateau, providing a unique reference for future studies on altitude-dependent AEF variations and their coupling with space weather and climate processes.

1. Introduction

The global electric circuit (GEC) is a planetary-scale direct current system maintained by worldwide thunderstorm activities and electrified precipitation clouds. These global generators inject approximately 1 kA of current into the ionosphere, maintaining an ionospheric potential of 250 kV relative to Earth’s surface. This potential generates a fair-weather atmospheric electric field (AEF) near the ground, typically ranging from 100 to 300 V/m, with an average of around 130 V/m [1,2]. Among various influencing factors, meteorological conditions and anthropogenic factors exert the most direct and significant modulation on the AEF. For example, variations in cloud cover and increases in aerosols or pollutants can substantially modify the vertical profile of atmospheric conductivity, which, in turn, perturbs the fair-weather (FW) electric field structure [3,4,5]. Besides meteorological influences, other factors such as solar activity (e.g., flares and coronal mass ejections) [6], interplanetary conditions (e.g., solar wind–magnetosphere interactions and interplanetary magnetic field sector boundary crossings) [1], and geophysical phenomena (e.g., earthquakes and volcanic activity) [7] also affect the AEF. A comprehensive review of the GEC is provided in Williams and Mareev [8], encompassing a wide array of topics such as atmospheric conductivity, lightning activity, climate-scale variability, and external forcings (e.g., GCRs and gamma-ray flares).

Essentially, the GEC is closely related to the degree of ionization in the Earth’s atmosphere, with galactic cosmic rays (GCRs) being one of its dominant ionization sources. Under strictly fair weather conditions, where meteorological and anthropogenic influences are excluded, any transient disturbances in the GEC induced by solar activity can be attributed to global-scale changes in atmospheric electrical conductivity associated with solar energetic particle events or Forbush decreases in GCR flux. The influence of space weather on AEF has received considerable scientific attention over the past few decades (e.g., Refs. [9,10,11,12,13]). For instance, Li et al. [11] investigated the response of the AEF at eight mid- and low-latitude stations in China during the large geomagnetic storm and associated Forbush decrease event around 24 April 2023, revealing that AEF increased at five stations and decreased at the other three. In a study of 15 geomagnetic storms at Mohe, China, Qiu et al. [14] reported that the AEF consistently exhibited sharp decreases concurrent with changes in the interplanetary magnetic field (IMF) and solar wind proton number density. Recently, Gurmani et al. [15] analyzed fair-weather AEF data collected at Islamabad, Pakistan, and found that solar activity and solar eclipses exerted a pronounced local impact on the AEF during solar cycle 24. However, it is important to emphasize that the prerequisite for such studies is the precise estimation of the background AEF, the removal of which is essential to isolate small perturbations attributable to space weather effects. These highlight the utility of AEF as a sensitive diagnostic tool for probing near-surface atmospheric processes and their coupling with both terrestrial and extraterrestrial forcing (e.g., Refs. [4,6,16,17,18]).

Given its sensitivity to global, regional, and local influences [19], characterizing the diurnal variation of the AEF is crucial to advance our understanding of the GEC. The classical Carnegie curve, first identified from extensive fair-weather marine measurements aboard the Carnegie vessel in the early 20th century, remains the canonical reference for the global diurnal pattern of AEF [20]. This globally recognized curve exhibits a highly consistent unimodal structure, peaking near 19:00 UT and reaching a minimum around 03:00–04:00 UT [2]. In regional AEF studies, observational results are frequently compared to the classical Carnegie curve to evaluate the coupling between global GEC signals and local atmospheric dynamics [21,22,23,24]. However, in regions with complex terrain, such as plateaus, mountainous areas, and continental interiors, AEF measurements often reveal bimodal or even multimodal diurnal structures that substantially deviate from the classical Carnegie curve. These deviations are typically attributed to the combined effects of different latitudes, elevations, meteorological variability, and anthropogenic factors [19,25,26,27,28,29].

To reliably identify fair-weather AEF measurements, it is essential to exclude data influenced by local meteorological and space weather disturbances. Recent studies have increasingly focused on defining strict fair-weather selection criteria to ensure that the observed AEF is representative of undisturbed GEC conditions [30,31]. The absence of precipitation is a fundamental prerequisite, as rainfall and surface wetness can significantly affect electric field measurements [24,32]. High horizontal visibility (typically ≥ 2 km) is required to minimize interference from aerosols and fog, which can alter atmospheric conductivity [4,22,33,34]. Wind speeds are typically constrained to 0–8 m/s to prevent charge accumulation during calm conditions and suppress turbulence-induced noise under strong winds [30,33]. In addition to meteorological influences, space weather events, particularly geomagnetic storms, can significantly disturb near-surface AEF measurements. Consequently, data collected during periods with geomagnetic activity index (Kp) exceeding 3 are typically excluded to ensure that the observations primarily reflect global atmospheric electric circuit behavior rather than magnetospheric perturbations [35,36,37,38]. Such rigorous fair-weather screening is particularly important in complex environments, such as high-altitude plateaus and mountainous regions, where local atmospheric disturbances can significantly affect AEF measurements. This standardization enhances data comparability across sites and enables more robust investigations into the physical mechanisms underlying deviations from the classical Carnegie curve [30].

Despite decades of scientific interest, a comprehensive understanding of AEF variability across global, regional, and local scales remains limited, primarily due to the scarcity of systematic and long-term observational datasets. Historically, AEF monitoring has been confined to isolated stations, resulting in fragmented data coverage and a lack of global coherence. To address this gap, the international Global Atmospheric Electricity Monitoring (GloCAEM) network, led by Dr. Karen A. Nicoll and colleagues at the University of Reading, was established to coordinate global AEF observations, characterize natural variability, and enhance understanding of the GEC [23,39]. In China, the government has supported a large-scale scientific initiative to monitor the near-Earth space environment through a meridional chain of ground-based observatories, named the Chinese Meridian Project (CMP) [40,41]. The CMP was initiated in two phases, beginning in 2008, and was officially approved by national authorities in March 2025. One of its key objectives is to monitor near-surface atmospheric electricity. As part of this effort, seven ground-based AEF observatories have been constructed, including Zhongshan Station (76.4° E, 69.4° S), Zhaoqing Station (112.4° E, 23.0° N), Pixian Station (103.7° E, 30.9° N), Jiufeng Station (114.5° E, 30.5° N), Gar Station (80.1° E, 32.5° N), Nong’an Station (124.9° E, 44.1° N), and Mohe Station (122.4° E, 53.5° N) [42].

At high-altitude regions, the compression of equipotential surfaces can lead to an enhancement of fair-weather AEF values [24,27]. Large-scale signals in these regions are less likely to be obscured by noise from local sources [19,43]. Therefore, observations from high-altitude stations are essential for advancing our understanding of atmospheric electricity and its response to both global and local influences. In this study, we analyze near-surface AEF measurements obtained at Gar Station (4259 m a.s.l.) from 2021 to 2024 to investigate the characteristics of atmospheric electricity over the Qinghai–Tibet Plateau. The structure of this paper is as follows: Section 2 introduces the study area and the monitoring instrumentation at Gar Station; Section 3 describes the methodology; Section 4 presents the analysis of AEF variability across multiple timescales, including interannual, seasonal, and diurnal variations; and Section 5 summarizes the main findings of the study.

2. Observation Site and Instrumentation

2.1. Study Area

This study analyzes near-surface AEF data collected from November 2021 to December 2024 at Gar Station. This station is situated in Gar County, in the westernmost part of Ngari Prefecture within the Tibet Autonomous Region, one of the highest inhabited areas in the world. Geographically, it lies in the Sengge Zangbo and Gar Zangbo river basins. As one of Tibet’s 18 border counties, Gar features an extremely low population density (0.8 persons per square kilometer) and is devoid of industrial infrastructure and high-voltage transmission lines. Moreover, the absence of significant continuous electromagnetic radiation sources in the vicinity ensures an exceptionally clean electromagnetic environment, making it well suited for high-precision AEF measurements. Figure 1 shows a map of the Qinghai–Tibet Plateau, indicating the location of the Gar AEF station. For comparison, another plateau AEF station, Yangbajing (YBJ), is also marked [25].

Gar Station is situated at an altitude of 4259 m above sea level in a cold and arid region, characterized by minimal annual precipitation, large diurnal temperature variations, and classification as a secondary wind zone [44]. These conditions reflect the typical features of a high-altitude plateau climate, with pronounced seasonal variations in temperature, humidity, wind speed, and precipitation (Figure 2). The monthly mean wind speed (Figure 2a) reaches up to 7.22 m/s in summer and drops to about 4.49 m/s in winter. Relative humidity (Figure 2b) increases to approximately 36% during summer and decreases to around 25% in winter. Visibility (Figure 2c) remains consistently high year-round, generally exceeding 29 km, with minor reductions in spring and autumn. The monthly mean temperature (Figure 2d) increases steadily from January to a maximum of approximately 17.5 °C in July, then declines to below −10 °C in January. This unimodal pattern reflects the region’s strong solar forcing and cold, arid conditions. Precipitation (Figure 2e) peaks in July (62.2 mm) and August (57.5 mm), while remaining low in other months. The monthly mean concentrations of PM₁₀ and PM_2.5 (Figure 2f), obtained from the Ngari National Ambient Air Quality Monitoring Station during 2017–2023 and adapted from Wang et al. [44], follow a consistent annual pattern: elevated levels from January to May, a minimum in August, and gradual increases from September to December. Seasonally, concentrations follow the order: winter > spring > autumn > summer. The elevated PM concentrations during the cold half of the year (winter and spring) are likely attributable to a combination of unfavorable meteorological conditions and enhanced anthropogenic emissions, including a shallow and stable boundary layer, frequent temperature inversions, weakened surface winds, and increased residential heating emissions [44].

Figure 3 illustrates the seasonal mean diurnal variations in the parameters shown in Figure 2, plotted in local time (LT). A consistent diurnal pattern is observed across seasons for all parameters, primarily attributable to radiative and thermodynamic forcing over the high-altitude plateau [45]. Wind speed (Figure 3a) reaches a daily minimum in the early morning (approximately 07:00–09:00 LT) and peaks in the late afternoon (around 17:00–19:00 LT). Air temperature (Figure 3d) also exhibits a pronounced diurnal cycle, reaching a minimum in the early morning and a maximum in the afternoon, highlighting the dominant role of radiative forcing in controlling surface thermal variations. Relative humidity (Figure 3b) reaches a maximum in the early morning and decreases markedly in the afternoon. Visibility (Figure 3c) remains relatively stable throughout the day in all seasons except winter, during which a pronounced reduction is observed in the morning hours (approximately 08:00–10:00 LT), likely due to temperature inversions, aerosol accumulation, or near-surface moisture [46,47,48]. The annual mean diurnal variation in pollutant concentrations (Figure 3e) shows a daily maximum between 09:00 and 12:00 LT, likely due to a combination of boundary layer development, the release of surface-accumulated pollutants triggered by early morning solar heating, and the onset of local anthropogenic activities. Following the morning peak, concentrations decline, reaching a minimum in the early afternoon (around 14:00–16:00 LT), possibly due to enhanced vertical mixing and convective turbulence that promote efficient dispersion. They then gradually rise again, peaking near midnight, before decreasing steadily through the early morning hours to another minimum around 08:00 LT. The nighttime accumulation is possibly attributed to a shallow boundary layer and suppressed dispersion under stable stratification [49,50]. Such a diurnal cycle is actually a reflection of the interplay between boundary layer dynamics and emission processes.

2.2. AEF Instrument

AEF measurements at Gar are conducted using a mill-type electric field meter (EFM-100), a standard instrument commonly used in the CMP for atmospheric electricity monitoring. The instrument is installed on the rooftop of the main building at the Ngari Seismic Station, approximately 3 $\mathrm{m}$ above ground level, at a carefully selected site that is open, level, and free from nearby vertical obstructions, such as chimneys, trees, or antennae. The EFM-100 probe, which constitutes the external sensing component of the instrument, is mounted atop a $1.5$ $\mathrm{m}$ high non-conductive mast fixed directly to the rooftop, resulting in a total installation height of approximately $4.5$ $\mathrm{m}$. The sensor operates by detecting induced surface charges on a conductor that is periodically exposed to the ambient electric field through rotating shielding electrodes.

Prior to deployment, the EFM-100 was calibrated under laboratory conditions following the procedure described by Li et al. [51]. First of all, the external sensor head of the EFM-100 was placed between two parallel metal plates with a radius of 54 cm and a fixed vertical separation of 24.6 cm. A series of adjustable direct current voltages (approximately 0, 40, 80, and 111 V) was then applied across the plates to generate known electric fields, and each voltage level was maintained for 5 min to examine the stability and accuracy of the probe. The results showed that the measured electric field remained stable within 8 V/m during each 5-min interval, indicating good temporal stability and linearity. It is worth noting that the obtained electric field values were consistently 5–20 V/m lower than the theoretical values, representing a deviation of less than 5%, which may arise from multiple sources, including environmental disturbances in the laboratory setting, intrinsic errors in the measurement system, and limitations in probe performance, such as the actual voltage applied to the probe terminals being lower than the theoretical input. On the other hand, to address potential distortions introduced by the rooftop installation and elevated measurement height, an empirical correction method based on long-term FW AEF observations was applied. Specifically, under conditions with minimal meteorological and environmental interference, the diurnal mean variations in the electric field were analyzed and subsequently fitted to data from ground-based reference stations with standardized installations through regression methods. The resulting correction factor was then used to adjust the measurements at Gar, ensuring consistency with surface-level AEF observations. To maintain the long-term stability and accuracy of the system, such instrument calibration is performed approximately once every two years.

Overall, the instrument continuously records the vertical component of the AEF, with a measurement range of ±50 kV/m, an accuracy better than 5%, linearity within 1%, a resolution of 10 V/m, and a temporal sampling interval of 1 s.

3. Method

To enhance the quality of near-surface AEF observations and minimize the influence of transient noise and local disturbances, rigorous quality control and preprocessing procedures are applied to the raw AEF data recorded at a 1-s resolution. As an initial step, AEF values exceeding the threshold of ±1 kV/m are discarded. These extreme values are typically due to transient phenomena such as lightning discharges or nearby electrical interference, and are unsuitable for climatological analysis.

Subsequently, the filtered 1-s resolution AEF data are first resampled to a minutely resolution. Then a Gaussian-weighted moving average is applied with a time window of 15 min to smooth the data [52]. The smoothed signal $h_g\left(t\right)$ at time t is given by:

$$h_g\left(t\right)={\displaystyle \frac{{\int }_{-\infty }^{+\infty }h\left(\tau \right)g(t-\tau )d\tau }{{\int }_{-\infty }^{+\infty }g(t-\tau )d\tau }},$$

(1)

where $\tau$ is the central time point of the Gaussian kernel, $h_g\left(t\right)$ represents the time-domain AEF signal after Gaussian smoothing, and $g(t-\tau )$ is the Gaussian kernel function defined as:

$$g\left(t\right)={\displaystyle \frac{1}{\sqrt{2\pi }\sigma }}exp\left(-{\displaystyle \frac{t^2}{2{\sigma }^2}}\right),$$

(2)

in which $\sigma$ controls the effective width of the filter. Here, $\sigma$ is selected such that the full width of the Gaussian kernel spans 15 min, effectively covering 95% of the total weight. This Gaussian kernel effectively acts as a low-pass filter, attenuating high-frequency random noise [53]. The resulting smoothed data can reduce short-term fluctuations while preserving the underlying trends of the electric field signal. Finally, the obtained time series are downsampled to an hourly resolution for subsequent analysis. It should be noted that, to ensure data quality, any hour containing fewer than 48 valid minute-level data points (i.e., less than 80% of the theoretical maximum of 60) is excluded from the hourly averaging.

Following the aforementioned quality control step, the hourly averaged AEF values are further screened under FW conditions, based on criteria used in previous studies (e.g., [30,42,54]), as detailed below:

(1)

wind speed < 8 m/s;

(2)

relative humidity ≤ 80%;

(3)

visibility ≥ 15 km;

(4)

no precipitation;

(5)

AEF value between 0 and 1 kV/m;

(6)

geomagnetic activity index Kp < 3.

The meteorological data are obtained from the Shiquanhe meteorological station (Station ID: 55,228; https://q-weather.info, accessed on 1 February 2025), which is located about 2.6 km from the Gar AEF station. The Kp index data are retrieved from the OMNI database (https://omniweb.gsfc.nasa.gov/form/dx1.html, accessed on 1 February 2025). A threshold on the Kp index is imposed to exclude the potential influences from solar and geomagnetic disturbances.

After quality control and FW screening, monthly statistics of valid hourly AEF data are compiled from November 2021 to December 2024. The number of valid data points for each month is listed in

Table 1. Monthly AEF data points under all and FW conditions.

Date	All (Hours)	FW (Hours)	Date	All (Hours)	FW (Hours)
Nov 2021	671	449	Jun 2023	665	286
Dec 2021	744	586	Jul 2023	697	281
Jan 2022	696	448	Aug 2023	684	409
Feb 2022	600	288	Sep 2023	659	212
Mar 2022	719	358	Oct 2023	736	469
Apr 2022	720	267	Nov 2023	716	447
May 2022	736	334	Dec 2023	665	460
Jun 2022	648	306	Jan 2024	575	422
Jul 2022	686	258	Feb 2024	574	266
Aug 2022	715	329	Mar 2024	717	328
Sep 2022	694	309	Apr 2024	632	247
Oct 2022	743	401	May 2024	537	175
Nov 2022	717	379	Jun 2024	617	245
Dec 2022	742	391	Jul 2024	675	304
Jan 2023	744	364	Aug 2024	582	247
Feb 2023	555	144	Sep 2024	502	305
Mar 2023	743	338	Oct 2024	282	160
Apr 2023	678	240	Nov 2024	312	281
May 2023	558	139	Dec 2024	639	327

and plotted in Figure 4b.

4. Results and Discussion

In this section, we investigate the temporal variability in the AEF across multiple timescales, including interannual, seasonal, semiannual, and diurnal variations, using ground-based measurements at the Gar Station.

4.1. Interannual Variation of AEF

A total of 12,199 h of FW AEF data are extracted from 24,575 h of preprocessed valid observations, following the selection criteria outlined in Section 3. Figure 4a displays the monthly mean AEF values under FW conditions, and Figure 4b shows the corresponding number of valid data points used in the calculation. We can see that the AEF profile exhibits a stable annual cycle characterized by a distinct single-peak structure, with maximum values occurring in December (approximately 0.411–0.559 kV/m) and minimum values during July–August (approximately 0.150–0.242 kV/m). This annual variation pattern is consistent with the seasonal evolution of PM₁₀ and PM_2.5 concentrations, as shown in Figure 2f. It is therefore speculated that the observed annual variability in AEF is possibly driven by changes in aerosol loading and meteorological conditions. Specifically, during winter, elevated aerosol concentrations resulting from anthropogenic activities reduce atmospheric conductivity by depleting free ions and decreasing ion mobility, thereby amplifying the near-surface electric field in accordance with Ohm’s law [31,44]. In contrast, cleaner atmospheric conditions in summer enhance conductivity, leading to a weakened AEF.

Furthermore, the annual mean AEF at Gar under fair-weather conditions is approximately 0.331 kV/m, significantly higher than values reported at lower-altitude stations [55,56,57,58]. For instance, at Islamabad, Pakistan, FW AEF values measured from 2015 to 2017 range between 0.100 and 0.275 kV/m [55]; in Beijing, China, an average of about 0.220 kV/m was recorded between August 2004 and November 2005 [56]. In Kolkata, India, values between 0.120 and 0.193 kV/m were observed from January 2006 to February 2009 [57], while the Ramon station in Israel reported an annual mean of approximately 0.186 kV/m during June 2013 to June 2015 [58]. The markedly elevated AEF at Gar is primarily attributed to its high-altitude environment, which leads to undulations and compression of equipotential surfaces, thereby enhancing the near-surface electric field relative to lower-altitude regions [24,27].

4.2. Diurnal Variation in AEF

Besides interannual variability, the AEF also exhibits pronounced seasonal and semiannual variations. Figure 5 illustrates the annual evolution of AEF diurnal variations, with the x-axis indicating both UT and LT, and the y-axis denoting calendar months. This figure provides a comprehensive view of how diurnal patterns evolve throughout the year, enabling the determination of both UT/LT dependencies and seasonal signatures. We can see that from October to the following May, the AEF exhibits an obvious double-peak and double-trough structure. In contrast, during the summer months (June–August), this pattern becomes markedly suppressed, and the AEF amplitude drops to its annual minimum. These seasonal contrasts likely reflect strong modulation by both global-scale processes and local influences, including meteorological conditions (e.g., boundary layer dynamics, wind speed, and precipitation) and environmental factors, such as surface aerosol loading [21,22,33,55].

To further quantitatively characterize the seasonal features of diurnal AEF variations, seasonal mean curves are computed for spring, summer, autumn, and winter, defined according to the conventional Northern Hemisphere classification: spring (March–May), summer (June–August), autumn (September–November), and winter (December–February). In addition, semiannual mean curves are computed for the winter and summer half-years, defined as November through April and May through October, respectively. The resulting seasonal and semiannual mean AEF profiles are presented in Figure 6, where local sunrise and sunset times are indicated by upward and downward arrows, respectively.

From Figure 6a, except for summer, the diurnal AEF curves in the other three seasons consistently exhibit a distinct “double-peak and double-valley” structure, likely associated with the effects of sunrise and sunset. In winter, two pronounced AEF peaks occur at approximately 11:00 LT (03:00 UT) and 23:00 LT (15:00 UT), with troughs near 17:00 LT (09:00 UT) and 07:00 LT (23:00 UT). In autumn, the peaks shift to around 11:00 LT and 21:00 LT, while troughs remain near 17:00 LT and 07:00 LT. In spring, peaks appear around 11:00 LT and 23:00 LT, with troughs at 18:00 LT and 07:00 LT. These seasonal shifts in peak and trough timings reflect variations in sunrise and sunset times, solar elevation angles, and boundary-layer development associated with seasonal changes in solar insolation and atmospheric stability. This characteristic “double-peak and double-valley” structure also aligns temporally with the diurnal variations in PM₁₀ and PM_2.5 concentrations shown in Figure 3e, suggesting that aerosol variability likely influences the observed AEF patterns through its modulation of atmospheric conductivity [31]. Between 00:00 and 08:00 LT, the near-surface AEF (and aerosol concentrations) generally remains low, reaching its first daily minimum shortly before sunrise. This can be attributed to the unique high-altitude conditions of the region: the absence of anthropogenic emissions and strong radiative cooling at night establish a stable boundary layer and suppress vertical mixing. In the absence of new aerosol sources, residual particles gradually settle or disperse, resulting in a transient decrease in near-surface aerosol concentrations. The associated increase in atmospheric conductivity weakens the near-surface electric field [52]. Around 08:00 LT, solar heating at sunrise initiates atmospheric thermal convection and turbulence, altering near-surface water vapor, wind speed, temperature, and aerosol concentrations. As solar radiation accumulates, surface temperature rises further, enhancing convective activity. Concurrently, anthropogenic emissions lead to a gradual increase in aerosol loading, while elevated temperatures enhance ionization, resulting in increased charged ion concentrations. Between 08:00 and 11:00 LT, intensified solar heating strengthens boundary-layer convection, lifting charged aerosols upward. Facilitated by the development of the exchange layer, this vertical transport of space charge leads to a notable increase in the near-surface AEF, producing the first daily peak [22,59,60,61,62]. Notably, the earlier the sunrise, the earlier the accumulation of radiative ionization begins, leading to an earlier occurrence of the morning AEF peak. Following the morning peak, the AEF begins to decline. The afternoon minimum, typically between 16:00 and 18:00 LT, is attributed to enhanced atmospheric conductivity driven by decreased near-surface aerosol concentrations and persistent solar effects. During this period, sustained convective mixing efficiently transports aerosols to higher altitudes, leading to a temporary depletion near the ground. This increased conductivity reduces the near-surface electric field. After sunset, the decay of solar radiation suppresses convection. A stable nighttime boundary layer forms, and aerosols begin to settle back toward the surface. Around midnight (24:00 LT), this reaccumulation of aerosols, combined with reduced ion production and diminished vertical current from the global electric circuit, decreases conductivity, resulting in a secondary AEF peak [39].

In summer, the diurnal AEF curve displays a relatively flat profile, reflecting the suppression of the characteristic double-peak structure evident in the other three seasons. This behavior is primarily attributed to enhanced solar heating, intensified boundary layer convection, elevated humidity, and reduced aerosol loading (as shown in Figure 2). These factors collectively increase atmospheric conductivity, thereby diminishing the near-surface vertical potential gradient and weakening the AEF. Consequently, the typical seasonal diurnal variability is markedly attenuated during the summer months.

Figure 6 further presents the semiannual mean AEF profiles for the winter (September to the following February) and summer (March to August) half-years. A clear semiannual contrast is evident, characterized by higher AEF magnitudes and more pronounced diurnal variation during the winter half-year, and lower magnitudes with a flattened profile during the summer half-year. This semiannual asymmetry is primarily driven by the same physical mechanisms responsible for the seasonal variations discussed above, including differences in solar heating, boundary layer dynamics, humidity, and aerosol loading.

In Figure 7, we present the mean diurnal AEF profile under FW conditions, obtained by averaging data over the entire study period. For comparative analysis, we also include diurnal AEF profiles from other stations: YBJ in China [25], Islamabad in Pakistan, Kolkata in India, and Roman in Israel [55,57,58], as well as the classical Carnegie curve [2]. Figure 7a shows that, both the Gar and YBJ curves deviate significantly from the classical Carnegie curve. The Gar curve exhibits a clear double-peak structure, with maxima occurring at approximately 03:00 UT (11:00 LT) and 14:00 UT (22:00 LT). This stands in contrast to the Carnegie curve, which was derived from ocean-based observations. The deviation observed at Gar may be attributed to regional factors such as local aerosol variability, complex boundary-layer processes, and electrode effects associated with the land surface [22,26,63]. A similar double-peak structure is also observed at the YBJ station, another AEF station located on the Tibetan Plateau [25]. However, the timing of the peaks at YBJ is slightly offset, and their amplitudes are more suppressed, particularly during the afternoon and early evening hours. We hypothesize that this attenuation may be related to more frequent convective activity and elevated near-surface humidity at YBJ. The presence of extensive geothermal infrastructure in the area could contribute to higher local moisture levels, which, together with stronger boundary-layer convection, may enhance atmospheric conductivity and reduce the near-surface electric field [64].

Figure 7b extends the comparison to include FW AEF observations from other Asian stations, including Islamabad (Pakistan), Kolkata (India), and the Ramon station in Israel [55,57,58]. These profiles reveal clear regional differences in the fair-weather atmospheric electric field. At the Islamabad station, Gurmani et al. [55] reported a distinct morning peak in the FW potential gradient at approximately 09:00 LT, with no pronounced secondary maxima throughout the day. They attributed this single-peak structure to a combination of continental thunderstorm activity, local sunrise effects, and pollution-related factors. In Kolkata, De et al. [57] identified a primary maximum around 09:30 LT, followed by secondary peaks at approximately 13:30–14:00 LT and 21:00 LT. This multi-peak structure was interpreted as a result of strong modulation of the GEC by continental thunderstorm activity, which is particularly pronounced in tropical regions. At the Ramon station in Israel, although the diurnal electric field pattern broadly follows the Carnegie curve, a secondary morning peak was also observed. Yaniv et al. [58] attributed this feature to increased aerosol concentrations near the surface during the early morning hours, linked to boundary layer processes. In contrast, the Gar station exhibits a more distinct double-peak structure, with two maxima occurring at approximately 11:00 LT and 22:00 LT, corresponding to periods shortly after sunrise and sunset, respectively [22,61].

5. Conclusions

This study systematically analyzes three years of AEF observations (November 2021 to December 2024) from Gar Station (80.13° E, 32.51° N; 4259 m a.s.l.), situated in Ngari Prefecture on the western Qinghai–Tibet Plateau. This region provides an exceptionally favorable setting for atmospheric electricity research due to its high altitude, extremely low humidity, and minimal anthropogenic aerosol loading. These conditions help suppress local disturbances and allow clearer detection of the global fair-weather electric field. Moreover, the elevated altitude enhances the intensity of galactic cosmic ray ionization, increasing the production of free ions and thereby strengthening the atmospheric conductivity response. In addition, the region’s relatively stable meteorological conditions reduce background variability in electric field measurements, making it well suited for long-term trend analyses. The main findings are as follows:

(1)

The annual mean FW AEF at Gar is approximately 0.331 kV/m, substantially higher than values reported at lowland or plain sites, reflecting the pronounced enhancement associated with high-altitude environments. A clear seasonal cycle is evident, with monthly mean AEF values peaking in December (0.411–0.559 kV/m) and reaching a minimum during July–August (0.150–0.242 kV/m). This seasonal variability is likely governed by a combination of meteorological conditions and aerosol concentrations, both of which modulate atmospheric conductivity and, consequently, influence the strength of the near-surface electric field.

(2)

The diurnal variation in AEF is also examined. A distinctive feature of the Gar curve is the presence of a double-peak diurnal pattern (approximately around 11:00 LT and 22:00 LT), primarily influenced by sunrise and sunset effects, as well as by the diurnal evolution of aerosol concentrations. The phase of the diurnal curve exhibits slight seasonal shifts: the earlier the sunrise, the earlier the onset of radiative ionization, which, in turn, leads to an earlier occurrence of the morning AEF peak;

(3)

Comparisons between land-based stations (Gar, YBJ, Islamabad, Kolkata, and Ramon) and the classical Carnegie curve reveal substantial and systematic deviations. All land-based sites exhibit distinct diurnal patterns and amplitudes that diverge markedly from the ocean-based Carnegie reference, which characterizes global fair-weather electric field behavior under minimal local influence. Notably, the AEF curve at Gar shows a significantly larger diurnal amplitude and a well-defined double-peak structure compared to YBJ, while other stations present unique features, such as pronounced morning peaks or irregular profiles. These disparities underscore the critical role of regional and local terrestrial factors, including meteorological conditions, aerosol loading, and geographical setting, in modulating the surface electric field and shaping the regional expressions of the GEC.

This work provides direct observational evidence for the enhancement of AEF at high altitudes, based on measurements from a plateau region. By integrating meteorological and aerosol observations, it offers a comprehensive analysis of the multi-scale variability of the AEF and its possibly underlying physical drivers. These results not only deepen our understanding of the regional background characteristics of the AEF, but also facilitate the extraction of weak signals induced by other disturbances from the background. Future research, leveraging longer time series and broader observational networks, will further advance our understanding of how space weather, climatic variability, and geophysical processes modulate atmospheric electricity in high-altitude environments.