Near-surface turbulence and intermittency observations and processed data at the Kelameili Station in the Gurbantunggut Desert during winter 2023–2024

Turbulence bursts were identified when the 5 s moving-average turbulent kinetic energy exceeded the 30 min mean plus two standard deviations. The derived indicators include the burst time fraction, the contribution of bursts to the absolute kinematic heat flux, and the ratio of absolute heat-exchange intensity between burst and non-burst periods. The gradient-tower observations were used to calculate the 2–10 m gradient Richardson number and to classify the records into unstable, weakly stable, moderately stable, and strongly stable conditions.The original high-frequency eddy-covariance and EddyPro records use Urumqi local time (UTC+6), whereas the gradient-tower data retain a Beijing Standard Time field (UTC+8).

The datasets were matched using a consistent relationship between half-hourly start and end times. The accompanying Python scripts support reproduction of the stability statistics, velocity spectra, turbulence intermittency characteristics, heat-exchange effects, and representative burst-event analyses reported in the associated study. The dataset provides observational support for studies of stable boundary layers, intermittent turbulence under weak-wind conditions, and land–atmosphere interactions over snow-covered arid surfaces.