In a recent breakthrough for environmental sensing, researchers have developed a technique that uses structured light beams and machine learning to extract precise data on temperature and wind speed from atmospheric turbulence.
Study: Weather sensing with structured light. Image Credit: Alex Photo Stock/Shutterstock.com
In a recent paper published in Communications Physics, researchers introduced a clever new method that uses structured light beams and machine learning to measure environmental conditions like temperature and wind speed. The trick? They’re using the distortions caused by atmospheric turbulence—the very thing that usually messes up optical systems—as a source of valuable information.
At the core of their approach are Orbital Angular Momentum (OAM) light beams. These aren’t your typical laser beams—they have a corkscrew shape that gives them unique properties. When these beams pass through turbulent air, they get distorted in specific ways. The team figured out how to read those distortions to figure out what’s going on in the atmosphere.
Why This Matters
Turbulence is a constant headache in optics. When light travels through air, even small fluctuations in temperature or wind speed can bend it, scatter it, or blur it beyond recognition. That’s bad news for communication systems, telescopes, and imaging tools.
But here's the twist: those distortions actually carry a fingerprint of the conditions that caused them. Instead of trying to filter them out, the researchers in this study opted to lean into the mess—treating it as data. With the right type of light and a smart enough algorithm, they showed it’s possible to turn noisy optical signals into detailed environmental measurements.
Inside the Experiment
To test this idea, the team built an experimental system with three core components: preparing the beam, simulating turbulence, and analyzing the received light. The setup was designed to mimic the kind of distortions that occur during long-range optical propagation, but in a controlled lab setting.
They started by using a spatial light modulator (SLM) to encode a basic Gaussian beam with OAM, creating vortex beams with specific spatial structures. These beams were then passed through a folded mirror system that created an effective propagation path of 36 meters. Along the way, the team introduced artificial turbulence by simulating random phase aberrations across multiple reflections—emulating real-world atmospheric effects.
The next step was analysis. The distorted beams were captured and run through a machine learning algorithm—in this case, a Support Vector Machine (SVM) regression model. This model was trained to detect subtle changes in the beam’s shape and structure and correlate them with specific environmental conditions like temperature and wind speed.
To improve the model’s performance, they used a polynomial kernel to map the features into a higher-dimensional space, helping the system identify patterns that wouldn’t be obvious in the raw data. They also fine-tuned the wavelength and focus of the optical system to ensure stable, noise-resistant readings across different test scenarios.
Results and Discussion
The results were impressive. The system was able to detect temperature variations as small as 0.49 °C and wind speed changes as low as 0.029 m/second—a strong indication that structured light, paired with machine learning, can offer high-sensitivity environmental sensing.
Beyond just raw accuracy, the team also explored how different OAM modes performed under turbulent conditions. They observed that the temporal intensity changes in the light—caused by eddies and turbulence—offered insight into the dynamics of the surrounding air. Higher-order OAM modes sometimes yielded better sensitivity, but they also ran into challenges like clipping, where the beam is partially lost or distorted due to the physical constraints of the system.
This pointed to a trade-off: while more complex beam modes may offer richer data, they also demand more careful tuning of the system to avoid degradation. The study suggests that selecting the right mix of beam modes and system settings will be key to maximizing performance in real-world applications.
The researchers also evaluated scintillation effects—rapid fluctuations in light intensity caused by turbulence—as another way to understand how optical distortions relate to environmental changes. By comparing the signal-to-interference ratios across different OAM modes, they gained further insight into how turbulence affects light in structured ways.
Looking Ahead
While the experiment was conducted over a 36-meter test range, the authors see clear potential for scaling it up. They believe this approach could be adapted for longer-distance sensing or outdoor deployments, especially in areas where conventional tools like lidar and radar face limitations. Unlike those systems, which can be affected by humidity or airborne particulates, light-based sensing using OAM could offer a more flexible alternative.
Applications could range from localized weather stations to mobile sensing platforms on drones or satellites—especially useful in remote or data-scarce regions. And with the addition of machine learning, these systems can continuously adapt to changing conditions, improving accuracy over time.
Journal Reference
Chen Z., Daly U., et al. (2025). Weather sensing with structured light. Communications Physics 8, 105. DOI: 10.1038/s42005-025-02004-5, https://www.nature.com/articles/s42005-025-02004-5