Quantitative Analysis of Lightning Rod Impacts on the Radiation Pattern and Polarimetric Characteristics of S-Band Weather Radar

Highlights

What are the main findings?

• Lightning rods significantly increase the sidelobe levels of an S-band dual-polarization weather radar by up to 4.55 dB, while leaving the main-beam pointing and beamwidth essentially unchanged.
• Lightning rods introduce localized, azimuth-dependent polarimetric disturbances, producing a roughly 0.24–0.25 dB positive $Z_{\mathit{DR}}$ bias in snowfall and mixed-phase precipitation, whereas freezing-rain cases show no persistent azimuthal anomaly.

What are the implications of the main findings?

• The clear directionality and posture-dependent behavior of the interference indicates that lightning rod geometry and placement should be explicitly considered in radar-antenna structural design, installation, and site planning.
• The quantified $Z_{\mathit{DR}}$ biases provide a practical reference for developing polarimetric quality-control schemes and bias-correction approaches for lightning-rod–induced disturbances in dual-polarization radar products.

Abstract

Lightning rods, while essential for protecting weather radars from direct lightning strikes, act as persistent non-meteorological scatterers that can interfere with signal transmission and reception and thereby degrade detection accuracy and product quality. Existing studies have mainly focused on X-band and C-band systems, and robust, measurement-based quantitative assessments for S-band dual-polarization radars remain scarce. In this study, a controllable tilting lightning rod, a high-precision Far-field Antenna Measurement System (FAMS), and an S-band dual-polarization weather radar (SAD radar) are jointly employed to systematically quantify lightning-rod impacts on antenna electromagnetic parameters under different rod elevation angles and azimuth configurations. Typical precipitation events were analyzed to evaluate the influence of the lightning rods on dual-polarization parameters. The results show that the lightning rod substantially elevates sidelobe levels, with a maximum enhancement of 4.55 dB, while producing only limited changes in the antenna main-beam azimuth and beamwidth. Differential reflectivity ($Z_{DR}$) is the most sensitive polarimetric parameter, exhibiting a persistent positive bias of about 0.24–0.25 dB in snowfall and mixed-phase precipitation, while no persistent azimuthal anomaly is evident during freezing rain; the co-polar correlation coefficient (${\rho }_{hv}$) is only marginally affected. Collectively, these results provide quantitative, far-field evidence of lightning-rod interference in S-band dual-polarization radars and provide practical guidance for more reasonable lightning-rod placement and configuration, as well as useful references for $Z_{DR}$-oriented polarimetric quality-control and correction strategies.

1. Introduction

Weather radars, as core instruments in modern meteorological observation networks, are widely used in quantitative precipitation estimation, hazardous weather monitoring, and early warning applications [1,2,3]. As an all-weather, automated sensing system with high temporal and spatial resolution, the data quality directly affects key operational processes such as quantitative precipitation estimation (QPE), cloud-microphysical analysis, and severe-convective-storm identification. To ensure safe operation during convective storms and intense lightning activity, lightning rods are routinely installed on or near radar antennas to prevent equipment damage and avoid interruptions of critical observations. However, fixed metallic or dielectric structures such as lightning rods and radomes can act as stable, azimuth-locked non-meteorological scatterers when illuminated by the radar beam, generating additional echo signals and compromising both measurement accuracy and polarimetric consistency [4]. This effect is inherently system-dependent, and as dual-polarization radars have become the mainstream instruments in national radar networks worldwide [5,6], lightning-rod-induced biases in differential reflectivity ($Z_{DR}$) and other polarimetric variables have become an increasingly important source of systematic error. From the perspective of numerical weather prediction and data assimilation, the spatial representativeness and error correlation of observations are known to strongly influence variational analyses and precipitation forecasts [7]. These considerations imply that systematic distortions caused by fixed structures such as lightning rods can reduce the reliability of radar data for quantitative applications.

In recent years, numerous studies have examined the impacts of lightning rods and other antenna-mounted structures on weather radar performance. Most results indicate that such fixed structures can form stable scattering centers in the antenna near field or sidelobe regions, leading to deviations in radar echo power distribution and polarimetric responses, thereby affecting the quantitative accuracy of key polarimetric parameters. Cornelius Hald et al. [8] reported that the lightning rods used in the German Meteorological Service’s polarimetric C-band radar network produce local scattering in the antenna near field, which diverts part of the transmitted power from the main beam and redistributes it into the sidelobes, leading to elevated apparent sidelobe levels and a reduction in system gain. Hui Wang et al. [9] configured lightning rods of two different diameters on a Beijing X-band dual-polarization radar and showed that small-diameter rods markedly reduced the $Z_{DR}$ bias and narrowed the azimuthal range over which lightning rods affect $Z_{DR}$, compared with the original large-diameter rods. Using long-term X-band dual-polarization radar observations from Beijing and Foshan, Wang Chao et al. [10] found that $Z_{DR}$ generally exhibited positive biases within ±15° of the lightning-rod azimuth, while reflectivity ($Z$) and the co-polar correlation coefficient (${\rho }_{hv}$) showed slight reductions, and the initial differential propagation phase (${\Phi }_{DP}$) displayed small irregular fluctuations. Beyond lightning rods, large wind farms located within the radar beam have also been shown to generate persistent, non-meteorological wind turbine clutter and to interfere with Doppler weather radar data, highlighting the broader impacts of man-made infrastructures on radar data quality [11]. From a data-quality-control perspective, deep-learning-based inpainting approaches, such as conditional GAN (Generative Adversarial Network) frameworks for correcting partial beam blockage in S-band WSR-88D observations, have recently been proposed to restore missing or low-quality radar echoes behind terrain or building blockages [12]. These studies provide an important foundation for understanding structural interference, yet most rely primarily on empirical statistics or software simulations and still exhibit several limitations: (1) existing work has focused mainly on X-band and C-band radars, which typically employ smaller apertures and hence wider beams, making it difficult to generalize the conclusions to large-aperture S-band systems; (2) many experiments were conducted under non-far-field conditions, complicating the separation of lightning-rod scattering from main-beam energy coupling; and (3) the lack of quantitative, repeatable measurement protocols has limited rigorous assessment of polarimetric parameter errors. Consequently, quantitative analyses based on high-precision far-field measurements are urgently needed to clarify the interference mechanisms of lightning rods on S-band dual-polarization radar performance and to provide transferable guidance for operational radar networks.

To quantitatively evaluate the impacts of lightning rods on S-band dual-polarization weather radar performance, this study employed a Far-field Antenna Measurement System (FAMS), provided by Chengdu Jinjiang Electronic System Engineering Co., Ltd., Chengdu, China, together with a controllable tilting lightning rod whose tilt angle can be manually set to predefined positions, thereby forming an interference analysis platform featuring adjustable orientations, multi-azimuth configurations, and multi-structure combinations. By combining antenna pattern measurements with typical precipitation observations, we analyzed the interference characteristics of lightning rods on S-band radar main-beam radiation properties and polarimetric parameters such as $Z_{DR}$ and ${\rho }_{hv}$. The findings provide theoretical and technical support for radar site optimization, improvement of long-term observation accuracy, and the design of polarimetric parameter quality-control and correction procedures under complex environmental and structural configurations.

2. Materials and Methods

2.1. Instruments and Experimental Setup

This study was conducted using the facilities of the Changsha Meteorological Radar Calibration Center (CRCC), the only dedicated weather radar calibration center in China. CRCC is equipped with high-precision hardware and software platforms, including an S-band dual-polarization weather radar (SAD radar) manufactured by Huayun Metstar Radar (Beijing) Co., Ltd., Beijing, China, and the FAMS. These facilities provide comprehensive and advanced capabilities for weather radar testing, calibration, and performance evaluation. An overview of the experimental site and lightning-rod configuration is shown in Figure 1.

FAMS is located northeast of the SAD radar at a straight-line distance of approximately 18.1 km, with a feed-height difference of about 25 m, ensuring an unobstructed line of sight that comfortably satisfies conventional S-band far-field criteria for an 8.5 m antenna aperture. Controllable lightning rods are installed around the radar radome, and the directional radiation from FAMS enables a quantitative assessment of lightning rod interference on the SAD radar’s performance parameters. To support subsequent analysis, the major instruments involved in the experiment and their key technical specifications are described below.

2.1.1. SAD Radar

The SAD radar provides high temporal and spatial resolution and excellent polarimetric measurement capabilities. Its radome adopts a polyhedral structural design, effectively suppressing scattering loss and improving the consistency of the antenna radiation pattern. As the reference radar at CRCC, it is used to evaluate the consistency of other radars. In this study, however, it mainly serves as a high-precision receiving terminal to acquire the echo signals used for antenna-pattern extraction. Technicians routinely conduct strict high-precision calibration, including UAV-suspended metal sphere calibration [13], active radar calibrator–based calibration [14], satellite–ground cross calibration [15,16], and far-field antenna measurements, ensuring the accuracy and consistency of observational data. The main technical specifications are listed in

Table 1. Technical specifications of the SAD radar system.

Parameter	Value
Antenna diameter (m)	8.5
Operating frequency (GHz)	2.845
Detection range (km)	460
Resolution (m)	62.5/150/250
Peak power (kW)	≥650
Pulse width (µs)	0.42/0.83/1.57/4.70
PRF (Hz)	300–1300
Beamwidth (°)	0.91° (H)/0.86° (V)
Antenna gain (dB)	45.23 dB (H)/45.29 dB (V)

. Serving as the receiving terminal in this study, the radar’s high stability provides reliable support for the quantitative assessment of lightning rod interference effects.

2.1.2. Lightning Rod

To investigate the influence of lightning rods on radar performance, four controllable tilting lightning rods were installed around the SAD radar antenna platform at azimuths of 37°, 127°, 217°, and 307°, corresponding to the north, east, south, and west directions. One of the lightning rods is located near the alignment between FAMS and the SAD radar. Their distribution is shown in Figure 2. Each lightning rod supports continuous elevation adjustment, with the 0° position defined as the folded state lying along the ground.

The lightning rods adopt a segmented and detachable structure, with representative structural parameters listed in

Table 2. Structural parameters of lightning rods.

Segment Length (m)	Segment Diameter (mm)	Deployment
3 + 3 + 3 + 3.3 + 3.2	120 + 100 + 80 + 62 + 44	North, South
2.6 + 2.6 + 2.6 + 2.6 + 2.5 + 2.4	127 + 100 + 85 + 69 + 54 + 41	East, West

. Because the rods were supplied by different manufacturers, those installed at different azimuths exhibit minor structural differences. This study, however, focuses on the effects of rod orientation and geometric alignment rather than structural differences, and no type-specific analysis is conducted. The rod bodies are made of fiberglass, and each rod is fitted with a stainless-steel air terminal at the top, which has typical dimensions of about 1 m in length and about 25 mm in diameter, although minor variations may exist among suppliers. The down conductor is embedded within the fiberglass body. These features provide high mechanical strength and effective lightning-current conduction while maintaining corrosion resistance, enabling controlled interference experiments under various mechanical and electromagnetic configurations.

2.1.3. FAMS

FAMS is mounted on a 35 m steel truss tower, located at an azimuth of 35.787° northeast of the SAD radar. Using a high-performance signal generator to transmit standard signals to the radar, the system enables precise measurement of key antenna parameters such as main-beam azimuth, beamwidth, and sidelobe levels [17,18]. The geometric relationship between the SAD radar and FAMS is illustrated in Figure 3.

Technical specifications of the far-field antenna measurement system are summarized in

Table 3. Specifications of FAMS.

Parameter	Accuracy
Antenna gain	±0.5 dB (wind speed < 3.3 m/s)
Beamwidth	±0.05°
Beam pointing	0.05° RMS
Sidelobe level	±2.0 dB (within ±2°)
Cross-polar isolation	±3 dB (for isolation ≥ 35 dB)
Dual-polarization orthogonality	±0.1° (tower-top swing < 0.02°)

. The system uses a Keysight E8267D signal generator (Keysight Technologies, Inc., Santa Rosa, CA, USA) as the standard source to provide stable, controllable test signals for far-field measurements. Its excellent frequency stability and amplitude linearity ensure a reliable radiation source for far-field measurements. These capabilities allow the system to support the systematic evaluation of lightning-rod interference on the radiation characteristics of the SAD radar.

2.2. Data Processing and Analysis

To investigate the impact of different lightning rod postures on the radiation characteristics of the SAD radar, multiple antenna pattern measurement experiments were conducted, focusing on the variations in the main-beam azimuth, half-power beamwidth (HPBW), and sidelobe level. All antenna pattern data were derived from the radar’s measured IQ signals [19], and a consistent data processing and parameter extraction procedure was used to ensure comparability among different test conditions. First, the raw IQ data were converted into normalized power distribution curves. Then, the main-beam azimuth and HPBW were obtained by processing the antenna pattern. Finally, sidelobe characteristics were extracted from the normalized power distribution for statistical analysis. The entire processing workflow remained identical for all experimental configurations, guaranteeing the reproducibility and comparability of the results.

2.2.1. Power Normalization

During the lightning rod experiments, the SAD radar received the standard signal radiated by FAMS and recorded it in complex form as $IQ=I+jQ$. For each azimuth angle, the corresponding complex samples were extracted, and the average power was computed. Let ${IQ}_x(m,n)$ denote the complex samples at azimuth $x$, where $m$ is the sample index and $n$ is the pulse index. The mean power is given by:

$$P\left(x\right)={\displaystyle \frac{1}{MN}}\sum _{m=1}^M\sum _{n=1}^N{\left|{IQ}_x\left(m,n\right)\right|}^2$$

(1)

The power is then converted to decibels:

$$P_{dB}\left(x\right)=10{log}_{10}\left(P\left(x\right)\right)$$

(2)

To facilitate comparison under different experimental conditions, all power values were normalized relative to the maximum:

$$P_{norm}\left(x\right)=P_{dB}\left(x\right)-\underset{x}{\max }\left(P_{dB}\left(x\right)\right)$$

(3)

where x denotes the azimuth angle. This step removes the effect of absolute gain variations and preserves only the shape of the antenna pattern.

2.2.2. Main-Beam Interpolation and HPBW Calculation

To reduce errors caused by sampling noise and the finite azimuth sampling interval, linear interpolation was applied to the antenna-pattern samples around the main beam to obtain a quasi-continuous representation that enables more robust estimation of main-beam parameters. Let the original data points be $(x_i,y_i)$, where $x_i$ is the azimuth angle and $y_i$ is the normalized power. For each interval $[x_i,x_{i+1}]$, the interpolation function is:

$$S_i\left(x\right)=y_i+{\displaystyle \frac{y_{i+1}-y_i}{x_{i+1}-x_i}}(x-x_i)$$

(4)

The main-beam azimuth ${\theta }_c$ corresponds to the maximum of the continuous curve:

$${\theta }_c=\arg \underset{x}{\max }S\left(x\right)$$

(5)

The HPBW is defined as the azimuth span where the power drops by −3 dB from the peak. Let $S_{\mathrm{m}\mathrm{a}\mathrm{x}}$ be the peak value and $S_{\mathrm{h}\mathrm{a}\mathrm{l}\mathrm{f}}=S_{\mathrm{m}\mathrm{a}\mathrm{x}}-3$. Using linear interpolation, the left and right half-power points ${\theta }_L$ and ${\theta }_R$ are obtained, and the beamwidth is:

$$HPBW={\theta }_R-{\theta }_L$$

(6)

This method effectively improves the robustness and accuracy of beam-parameter extraction.

2.2.3. First Sidelobe Detection and Statistical Analysis

After determining the main-beam region, the first sidelobe characteristics were quantified using a peak search method implemented with SciPy (v1.13.1) [20]. Let the half main-beam width be $W=\mathrm{H}\mathrm{P}\mathrm{B}\mathrm{W}/2$. The sidelobe search region $\Omega$ is defined as:

$$x<{\theta }_c-Worx>{\theta }_c+W$$

(7)

Within this region, the first sidelobe is identified as:

$${\theta }_{s1}=\arg \underset{x\in {\Omega }_{\mathit{peak}}}{\max }\left(P_{norm}\left(x\right)\right)$$

(8)

$$P_{s1}=P_{norm}\left({\theta }_{s1}\right)$$

(9)

where $\Omega$ is the set of discrete azimuth samples that satisfy Equation (7), and ${\Omega }_{peak}$ denotes the set of local maxima within $\Omega$. The first sidelobe azimuth is denoted as ${\theta }_{s1}$. For each scan, the following parameters were extracted: main-beam azimuth ${\theta }_c$, HPBW, and first sidelobe level $P_{s1}$. Their mean and standard deviation were computed to quantify pattern stability. To provide a clear and reproducible workflow, Algorithm 1 summarizes the procedure for extracting the main-beam and sidelobe parameters from the IQ data.

Algorithm 1. Extraction procedure for antenna pattern parameters

Input: IQ data ${\mathit{I}\mathit{Q}}_{\mathit{x}}\left(\mathit{m},\mathit{n}\right)$, azimuth sequence $\mathit{x}$ Output: ${\theta }_c$, HPBW, first sidelobe level $P_{s1}$ 1. Power Calculation and Normalization  For each azimuth $x$:    Compute mean power $P\left(x\right)={\displaystyle \frac{1}{MN}}\sum {\mid {IQ}_x\mid }^2$  Convert to dB: $P_{\mathrm{d}\mathrm{B}}\left(x\right)=10{\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\left(P\right(x\left)\right)$  Normalize: $P_{\mathrm{n}\mathrm{o}\mathrm{r}\mathrm{m}}\left(x\right)=P_{\mathrm{d}\mathrm{B}}\left(x\right)-\underset{x}{\max }\left(P_{dB}\left(x\right)\right)$ 2. Main-Beam Interpolation  Apply linear interpolation to obtain $S\left(x\right)$  Determine ${\theta }_c=\mathrm{a}\mathrm{r}\mathrm{g}\underset{x}{\mathrm{m}\mathrm{a}\mathrm{x}}S\left(x\right)$ 3. HPBW Calculation  Compute $S_{\mathrm{h}\mathrm{a}\mathrm{l}\mathrm{f}}=S_{\mathrm{m}\mathrm{a}\mathrm{x}}-3$ dB  Find ${\theta }_L$ and ${\theta }_R$  HPBW = ${\theta }_R-{\theta }_L$ 4. First Sidelobe Detection  Define $\Omega =\{x\mid x<{\theta }_c-\mathrm{W}orx>{\theta }_c+\mathrm{W}\}$  Identify local maxima within $\Omega$ to form ${\Omega }_{peak}$ $${\theta }_{s1}=\mathrm{a}\mathrm{r}\mathrm{g}\underset{x\in {\Omega }_{peak}}{\mathrm{m}\mathrm{a}\mathrm{x}}(P_{\mathrm{n}\mathrm{o}\mathrm{r}\mathrm{m}}\left(x\right))$$ $$P_{s1}=P_{\mathrm{n}\mathrm{o}\mathrm{r}\mathrm{m}}\left({\theta }_{s1}\right)$$

Figure 4 shows an example of the antenna pattern and key parameter extraction. Red dots indicate raw normalized power samples, the blue curve represents the data curve, the green line denotes linear interpolation around the main-beam region, and the blue star marks the first sidelobe peak. All antenna pattern results reported in the following analysis are averaged over at least three repetitions and presented as mean ± standard deviation.

3. Results

This section aims to systematically evaluate the influence of lightning rods under different postures, rod elevation angles, and azimuthal arrangements on the radiation characteristics of the SAD radar antenna. Using the measured antenna patterns from FAMS far-field directional transmissions, the analysis focuses on key indicators including the main-beam azimuth, HPBW, and sidelobe level. Comparative assessments are conducted for cases with and without lightning rods, under different tilt angles, and at various azimuthal positions to reveal the directional behavior of scattering and blockage effects. In addition, radar observations during representative precipitation events are analyzed to illustrate azimuth-locked signatures associated with lightning-rod directions. These findings provide the basis for subsequent discussions on long-term observational impacts and engineering recommendations.

3.1. Antenna Radiation Pattern Differences Between Upright and Folded Lightning-Rod States

This subsection compares antenna patterns under two lightning-rod postures: upright and folded (near-horizontal). The rod is mounted on the north side of the radar. When folded, it lies close to the ground and falls outside the radar’s illuminated region, causing virtually no interaction with the received signal. When upright, it extends into the antenna’s illuminated region and produces measurable scattering. The two postures are illustrated in Figure 5.

Table 4. Antenna pattern parameters for the lightning rod on the north side of the radar in upright and folded states.

States	Main-Beam Azimuth (°)	HPBW (°)	First Sidelobe Level (dB)
Upright	35.763 ± 0.006	0.932 ± 0.005	−26.139 ± 0.031
Folded	35.760 ± 0.010	0.932 ± 0.006	−30.202 ± 0.009

summarizes the main-beam azimuth, HPBW, and first sidelobe level for the two postures. The main-beam azimuth and HPBW show almost no change between upright and folded cases, indicating that the rod posture has minimal influence on the main-beam direction or beam shape. In contrast, the first sidelobe level increases by 4.06 dB in the upright posture compared with the folded posture, demonstrating a clear sidelobe enhancement effect and indicating that even a geometrically slender lightning rod can redistribute a non-negligible fraction of the radiated energy into the near-sidelobe region.

Figure 6 presents the normalized power patterns for the two conditions. As shown in Figure 6a, the main-beam shape is nearly identical, with negligible differences in beam azimuth and HPBW. However, significant differences appear near the main beam, especially in the sidelobe region. Figure 6b,c show power difference curves within ±10° and ±15°, which are used to compare pattern changes from the near-sidelobe region to farther-off sidelobes. Within ±10°, the mean uplift is 4.85 dB; within ±15°, the mean uplift reaches 4.39 dB. All antenna patterns were peak-normalized to highlight relative energy-distribution changes induced by the lightning rod. Overall, the present S-band experiments indicate a limited impact on main-beam shape but more pronounced perturbations in sidelobe energy.

The quantitative analysis in this study is primarily based on experiments at a single radar scan elevation angle. To illustrate the spatial variations in echo structures, a fixed-range slice at approximately 10 km from the radar was extracted from multi-elevation radar scans for visualization. Figure 7 illustrates the two-dimensional distributions of reflectivity $Z$ and horizontal channel signal-to-noise ratio (SNRh), providing a qualitative visualization of the spatial structures under the two postures. In the folded posture, the lightning rod lies close to the ground and remains outside the radar’s illuminated region, resulting in smooth fields that are consistent with system noise. In contrast, in the upright posture, a narrow reflectivity-enhanced band appears near the main-beam direction, accompanied by increased SNRh.

The combined evidence from Figure 6 and Figure 7 indicates that the observed features are consistent with the geometric configuration of the lightning rod. The results show that the rod introduces direction-dependent scattering within specific azimuth sectors near the main-beam direction. These effects may generate persistent directional artifacts in the far-field signals, potentially reducing the accuracy and stability of radar measurements. Supplementary simulations further show that the presence of lightning rods causes limited changes in main-beam gain, indicating that lightning-rod impacts are dominated by sidelobe distortions rather than by substantial main-beam effects.

3.2. Impact of Lightning Rod Elevation Angle on the Antenna Radiation Pattern

To assess how rod elevation angle affects antenna performance, the north side lightning rod was tested at seven rod elevation angles: 90°, 75°, 60°, 45°, 30°, 15°, and 0°. The corresponding configurations are illustrated in Figure 8, and the measurement results for main-beam azimuth, HPBW, and first sidelobe level are listed in

Table 5. Comparison of antenna pattern parameters with different lightning rod elevation angles.

Rod Elevation Angle (°)	Main-Beam Azimuth (°)	HPBW (°)	First Sidelobe Level (dB)
90°	35.763 ± 0.006	0.932 ± 0.005	−26.139 ± 0.031
75°	35.753 ± 0.021	0.929 ± 0.004	−27.932 ± 0.031
60°	35.763 ± 0.015	0.930 ± 0.002	−30.194 ± 0.040
45°	35.757 ± 0.015	0.938 ± 0.004	−29.959 ± 0.046
30°	35.753 ± 0.015	0.932 ± 0.006	−29.921 ± 0.040
15°	35.760 ± 0.000	0.934 ± 0.005	−29.919 ± 0.025
0°	35.760 ± 0.010	0.932 ± 0.006	−30.202 ± 0.009

. The 90° and 0° configurations correspond to those presented in Section 3.1, and the same measurement results are used for consistency.

As shown in Table 5, the main-beam azimuth remains stable near 35.76°, and the maximum HPBW variation is within 0.01°, indicating that changes in lightning-rod elevation angle have a negligible influence on main-beam pointing and shape at S-band. In contrast, first sidelobe levels rise significantly at high rod elevation angles. Relative to the 0° case, first sidelobe levels increase by +2.27 dB at 75° and +4.06 dB at 90°, indicating markedly stronger scattering coupling as the rod approaches the vertical state and intersects a larger fraction of the illuminated aperture.

Figure 9 shows that first sidelobe levels remain nearly constant from 0° to 60°, but exhibit a sharp increase at 75° and 90°, reaching the highest value at 90°. This demonstrates that scattering is strongest when the lightning rod is nearly vertical, indicating a pronounced directional interference effect.

3.3. Impact of Lightning Rods at Different Azimuth Configurations on the Antenna Radiation Pattern

Lightning rods are located at four fixed azimuths around the radar. In this experiment, the south rod remained folded in all configurations, while the postures of the north, east, and west rods were varied to produce four test cases: (a) all rods folded, (b) north rod upright, (c) west rod upright, and (d) east rod upright. These posture configurations are illustrated in Figure 10. Throughout all measurements, the far-field illumination from FAMS was kept fixed, and only the rod postures, not their azimuthal positions, were altered. The complete antenna-pattern parameters are summarized in

Table 6. Antenna pattern parameters under different lightning rod azimuth configurations.

Configuration	Main-Beam Azimuth (°)	HPBW (°)	First Sidelobe Level (dB)
All folded	35.760 ± 0.017	0.931 ± 0.001	−30.275 ± 0.039
North upright	35.763 ± 0.006	0.926 ± 0.006	−25.727 ± 0.014
West upright	35.750 ± 0.010	0.935 ± 0.006	−30.258 ± 0.032
East upright	35.750 ± 0.010	0.932 ± 0.007	−30.149 ± 0.046

.

The results show that the main-beam azimuth remains stable between 35.75° and 35.76°, and the HPBW varies only from 0.926° to 0.935°, indicating a negligible impact of azimuth placement on main-beam direction or beamwidth. In contrast, the first sidelobe level changes significantly: the north-upright rod increases the sidelobe level by 4.55 dB, whereas the west and east upright rods remain close to the all-folded condition, with differences below 0.15 dB.

Figure 11 further shows that when the rod is on the north side, several sidelobe peaks around the main-beam direction become noticeably higher. By contrast, the east and west placements produce sidelobe curves nearly identical to the folded condition. These results indicate that significant scattering occurs only when the lightning rod lies close to the main-beam direction. This indicates that the azimuth placement of the lightning rod primarily modulates sidelobe characteristics, while leaving the main-beam azimuth and HPBW effectively unchanged.

3.4. Analysis of Lightning Rod Effects on Dual-Polarization Parameters

To examine how lightning rods affect the polarimetric observations of the SAD radar under different weather conditions, three representative precipitation events were selected: a snowfall event beginning at 21:00 UTC on 21 January 2024, a freezing-rain event starting at 23:42 UTC on 3 February 2024, and a convective–stratiform mixed precipitation event starting at 06:32 UTC on 24 June 2024. Each event consists of three consecutive volume scans covering the full 360° azimuth, ensuring representativeness and comparability for statistical analysis.

In the preprocessing stage, azimuthal alignment was applied by selecting the 360 samples closest to integer-degree azimuths so that all polarimetric variables correspond at identical azimuth angles. $Z_{DR}$, $Z$, SNRh, and ${\rho }_{hv}$ were then reordered and processed using an identical workflow. To isolate lightning-rod-induced effects from background variability, non-precipitation and low-quality echoes were removed using the following thresholds: reflectivity $Z$ between 18 and 25 dBZ, SNRh > 20 dB, ${\rho }_{hv}$ > 0.95, and height ≤ 3 km. Finally, azimuthal mean curves of $Z_{DR}$ and ${\rho }_{hv}$ were computed by averaging all valid samples at each azimuth angle, and the global mean $Z_{DR}$ was superimposed as a reference to assist in identifying lightning rod-induced anomalies.

Figure 12 marks the azimuths of the four lightning rods at approximately 37°, 127°, 217°, and 307°, using black vertical dashed lines, and

Table 7. Comparative analysis of lightning rod interference under different precipitation types.

Precipitation Type	${\mathit{Z}}_{\mathit{D}\mathit{R}}$ Anomaly Amplitude (dB)	Affected Azimuth Width (°)	Severity
Snowfall	±0.24	40–55	Moderate
Freezing rain	±0.25 *	N/A	Not significant
Mixed precipitation	±0.25	20–40	Moderate

* The $Z_{DR}$ anomaly in freezing-rain cases was not azimuthally locked to the lightning rod positions.

summarizes the corresponding interference characteristics. Each of the three precipitation types was analyzed independently to compare differences in interference characteristics. Since previous studies have shown only weak lightning-rod effects on differential phase ${\Phi }_{DP}$ [9], the analysis here focuses on $Z_{DR}$ and ${\rho }_{hv}$.

During the snowfall process, hydrometeors were primarily solid snow crystals [21]. Because snowflakes have complex shapes and low effective refractive indices, $Z_{DR}$ exhibits stronger intrinsic fluctuations. As shown in Figure 12a, reflectivity is spatially uniform but relatively weak, consistent with stratiform snowfall. ${\rho }_{hv}$ remains near 0.98 with only small azimuthal variations. In contrast, pronounced localized disturbances in $Z_{DR}$ appear near the lightning-rod azimuths, with anomaly amplitudes of approximately 0.24 dB over widths of 40–55°, especially near 37° and 307°, and to a lesser extent near 217°. This demonstrates that even in predominantly solid-phase precipitation, lightning rods can introduce systematic, azimuth-locked polarimetric interference.

Freezing rain is a typical mixed-phase precipitation type involving the transition between liquid and ice phases [22], whose dielectric contrasts make polarimetric variables highly sensitive to microphysical changes. As shown in Figure 12b, the event is dominated by layered stratiform echoes with localized enhancement. $Z_{DR}$ fluctuates by approximately ±0.25 dB, but no clear persistent azimuthal anomaly is evident at the lightning rod azimuths. ${\rho }_{hv}$ remains near 0.98 with only slight reductions, indicating stable signal quality and weak lightning rod interference. A plausible explanation is that icing on the lightning rod during freezing-rain conditions modified its effective dielectric properties and local geometry, thereby reducing the strength and coherence of the scattering coupling with the radar beam.

In the mixed precipitation event, echo structures are more complex and vary substantially in intensity [23]. As shown in the reflectivity field in Figure 12c, strong convective cells appear near the radar, with pronounced spatial variability in reflectivity. Localized $Z_{DR}$ anomalies of approximately 0.25 dB occur at the lightning rod azimuths, particularly at 37°, 217° and 307°, with typical widths of 35–40°, and a narrower sector of about 20° near 127°. ${\rho }_{hv}$ also exhibits small fluctuations but overall remains near 0.98. These results indicate that scattering from the lightning rods can still affect polarimetric observations in complex precipitation environments.

4. Discussion

Overall, the experiments indicate that lightning rods have a limited impact on the main-beam radiation characteristics of the S-band dual-polarization radar but significantly modify the sidelobe energy distribution. The main-beam pointing and beamwidth remain essentially stable, whereas the sidelobe level increases notably with different lightning rod configurations, including azimuth and rod elevation angle. The maximum enhancement reaches 4.55 dB, indicating that scattering coupling from the lightning rod redistributes radiated energy and produces a substantial effect on the overall antenna pattern.

Lightning rod interference exhibits clear directional and posture-dependent behavior. The interference strength is primarily determined by the geometric relationship between the radar beam and the lightning rod. When the lightning rod aligns with the main beam or is configured at a high rod elevation angle, sidelobe enhancement becomes most pronounced. Conversely, when the rod is located away from the main-beam direction or at lower elevations, the interference is greatly reduced. This demonstrates that the rod’s azimuth and configuration are key factors governing interference magnitude.

The primary polarimetric impact of the lightning rod manifests as a bias in $Z_{DR}$, with precipitation-type-dependent characteristics. During snowfall and mixed-phase precipitation, $Z_{DR}$ increases by approximately 0.24–0.25 dB in azimuth sectors aligned with the lightning rods, whereas no persistent azimuthal anomaly is evident at the lightning-rod azimuths under freezing-rain conditions. Changes in ${\rho }_{hv}$ remain small across all precipitation types, with no clear azimuthal dependence on the lightning-rod directions. These findings suggest that lightning rod scattering alters the energy balance between the two polarization channels, leading to $Z_{DR}$ biases whose magnitude and stability appear to be influenced by both the hydrometeor phase and the background signal-to-noise ratio.

5. Conclusions

This study, based on the high-precision FAMS and SAD radar platforms at the CRCC, established an adjustable experimental system using a tilting lightning rod with different azimuth configurations and rod elevation angles. The system enabled quantitative evaluation of lightning rods’ induced changes in S-band radar radiation characteristics and polarimetric interference, complemented by analyses of representative precipitation observations. This study provides observational evidence and quantitative insights into lightning rod interference in S-band dual-polarization radars, filling a gap in previous measurement-based research. The results offer a scientific basis for optimizing radar-antenna structural design, site planning, and polarimetric data quality control.

In practical terms, these findings suggest that the positioning and structural configuration of lightning rods should be explicitly considered in radar site planning and routine calibration, particularly for dual-polarization performance. Future work may incorporate full-wave numerical electromagnetic simulations and scattering-matrix modeling to further investigate the underlying scattering mechanisms under various structural and material configurations, and to develop quantitative, operationally implementable correction models for lightning-rod interference in dual-polarization radar products.