Power kites can map high-altitude turbulence beyond conventional sensors while generating electricity, opening new possibilities for wind control
Tethered kites built for wind power can now map atmospheric turbulence at altitudes that radar masts and lidars have never reached — while generating electricity
Hundreds of meters above the ground, the atmospheric boundary layer churns with turbulence that directly shapes how next-generation wind energy systems perform. For decades, that zone has been effectively invisible to instruments. Measurement masts run out of height. Lidars lose resolution. The data stops well before the altitudes where tethered kites now fly.
What researchers have found is that those kites may already be collecting what the instruments can’t. A tethered kite harvesting wind energy is also quietly sensing the turbulence around it — in ways that conventional fixed-point sensors were never designed to replicate.
The methodology was developed and validated entirely within the UniSimAWE framework, using the Mann turbulence model to generate realistic atmospheric conditions.
A measurement gap hundreds of meters in the making
The atmospheric boundary layer — the lowest few kilometers of the atmosphere — is where wind energy happens. Turbulence within it determines how much load a structure endures, how efficiently a rotor or kite extracts energy, and how accurately weather models predict wind behavior. Understanding that turbulence at altitude has direct consequences for how systems are designed, sited, and controlled.
The tools built to measure it have hard ceilings. Meteorological masts, the gold standard for surface-layer wind data, rarely exceed 200 meters. Fixed in place, they capture only what passes through a single point in space. Lidars reach higher but suffer from volume-averaging effects that blur fine-scale turbulent structures — exactly the features that matter most for load and control calculations.
Above the surface layer, the data thins out fast. For energy developers siting projects or grid operators forecasting output, that gap means working with incomplete information about conditions that will actually drive performance.
What airborne wind energy systems actually do
Airborne wind energy systems — AWES — take a fundamentally different approach to harvesting wind. In ground-generation configurations, a tethered kite flies in crosswind maneuvers, pulling a tether that drives a generator on the ground. The system alternates between a power-producing phase, where the kite sweeps through the sky under high tension, and a return phase where it reels back in using minimal energy before the cycle repeats.
To execute those maneuvers reliably, the kite needs to know exactly where it is and how the air around it is moving. Onboard sensors — inertial measurement units, GPS receivers, airspeed probes — are standard equipment, not optional additions. They’re there for flight control.
These systems operate well into the upper atmospheric boundary layer, where wind is strongest and where the kites are designed to fly. That also happens to be precisely where the measurement gap exists.
Two flight phases, two complementary data streams
The research, developed within Kitemill’s UniSimAWE simulation framework, treats the two operational phases not as an energy engineering problem but as a measurement opportunity.
During the return phase, the kite follows a trajectory running roughly in line with the wind — a near-streamwise path. That geometry allows direct measurement of turbulence intensity and integral time and length scales for all three velocity components. The study confirmed that Taylor’s frozen turbulence hypothesis holds for this sampling strategy, verified through two independent checks: the ratio of integral length to time scales, and a comparison of power spectral densities in both wavenumber and frequency domains. That validation is the theoretical foundation making it legitimate to convert time-series measurements into spatial turbulence statistics.
The production phase tells a different story. As the kite traces its helical crosswind path under high tension, the dominant crosswind motion unlocks integral scales in the crosswind direction — something fixed-point sensors simply can’t provide. The two phases together produce complementary datasets covering what neither could capture alone.
A new metric born from flight path and turbulence interacting
Beyond characterizing existing turbulence quantities, the research introduces something new: an interaction time scale. This metric describes how atmospheric turbulent structures are experienced along the AWES trajectory itself — not just how the turbulence exists in the atmosphere, but how a moving kite actually encounters it.
That distinction matters for control. Turbulence-adaptive control algorithms need to anticipate how gusts and eddies will affect a kite’s flight path and tether tension. A metric capturing the interaction between turbulent structures and the kite’s trajectory gives those algorithms something more actionable than raw atmospheric statistics. The methodology was developed and validated entirely within the UniSimAWE framework, using the Mann turbulence model to generate realistic atmospheric conditions.
What this means for wind energy and atmospheric science
The most immediate implication is practical: AWES fleets could generate electricity and collect high-altitude turbulence data simultaneously, with no additional infrastructure. The sensors are already onboard. The flight paths already cover the altitudes that matter, so data collection becomes a byproduct of normal operation rather than a separate undertaking.
That dual capability has downstream value for wind farm siting, turbine load forecasting, and atmospheric boundary layer research — fields that have been working around the same measurement gap for decades.
Field validation is the next step. The methodology is currently simulation-confirmed, built on the Mann model inside a controlled framework. Real AWES hardware flying in real atmospheric conditions will be needed to confirm the approach holds at scale and across varying wind regimes. That work hasn’t been done yet — but the framework for doing it now exists.
Learn more here: Porta Ko, A., Kelly, M., Nguyen, D. H., and Oland, E.: Using an Airborne Wind Energy System as a turbulence sensor, Wind Energ. Sci. Discuss. [preprint], https://doi.org/10.5194/wes-2026-150, in review, 2026.
Daniel García is an Editor-in-Chief with strong expertise in structural work and engineering principles. He combines this technical foundation with deep knowledge of energy, spatial design, and emerging technologies, bringing a forward-thinking and analytical approach to editorial leadership.
