General comments:

This paper proposes a novel second-order structure function model for lidar measurements. The model can be used to correct turbulence filtering caused by probe-volume averaging and to estimate the turbulence energy dissipation rate. The proposed method is shown to provide energy dissipation rate estimates from a six-beam lidar that are in good agreement with traditional spectra-based methods using a sonic anemometer. Overall, the paper is well written, and the proposed method makes a significant scientific contribution across multiple fields. I have several comments, mostly minor, that the authors may consider to further improve the manuscript.

Major comments:

1. L20 in introduction. The authors claim that hot-wire sensors are commonly used for atmospheric turbulence measurements. Based on my experience, hot-wire anemometers are used far more often in laboratory experiments. Hot-wire instruments are sensitive and fragile; the thin wire is extremely delicate and breaks easily under real atmospheric conditions. Rain, dust, insects, humidity, and temperature swings can affect their accuracy or damage the sensor. In addition, hot-wire anemometers exhibit calibration drift and require frequent recalibration, making them impractical for remote or long-term field deployments.
2. Some explanations or references on why a pulsed lidar does not measure point but rather a probe volume is helpful. Also, a CW lidar is possible to measure a point-like turbulence as a hot-wire or sonic anemometer if the focused distance is very short (smaller than several meters) to the probe lens.
3. Line 140, Some additional details on the six-beam lidar are recommended. Commercial pulsed lidars typically have a radial (range-gate) resolution of 15–60 m, which is determined by the temporal window used in the Fourier transform applied to the heterodyne signal. Can the reported 3 m resolution be achieved simply by selecting a shorter window length, or is a different signal-processing strategy or hardware configuration being used?
4. Line 182. Is the mean obtained by averaging over time—for example, every 6 seconds for each beam? In Step 2, the six beams will produce different LOS wind-speed spectra because each beam is contaminated by different velocity components according to its projection vector, and because atmospheric turbulence is anisotropic. Please explain why the average is taken over the six beams. In particular, the vertical beam measures only the vertical velocity component, yet it is still included in the averaging.
5. The impact of the mean vertical wind profile (related to atmosphere stability) on the LOS velocity distribution within the probe volume also requires further discussion.
6. It is recommended that the authors discuss how the proposed method for detecting energy dissipation rates could be utilized to improve turbulence intensity estimation in wind-energy applications.

Minor comments:

1. Abstracts: in areas ->in areas or at heights
2. Line-of-sight some times written as line of sight
3. Line 172, please explain why these wavenumber range is chosen
4. Line 177, Is the Taylor’s theory is also considered in the accumulation time (is), please clarify
5. Please explain N in eq (11)
6. Figure 5, should be k1 not k in the left figure
7. Line 215, Can the author further explain why the errors increase with heights?
8. Figure 7. The point density needs to be explained.
9. Line 249, DSW already defined