The manuscript describes the calibration of two of the three model components that constitute the DWM model and the comparison of the full model result to wake flow measurements.

Data base for calibration are measurements from a SpinnerLidar in the wake field of one of the V27 wind turbines at the SWIFT facility.

Three different scanning strategies are applied to collect different kind of flow field representation.

The total amount of collected scans comprises roughly 142 hours that are sorted in three different atmospheric stability classes.

By means of Bayesian inference the model parameters of two of the model components of the DWM model, the deficit in the meandering frame of reference and the wake-added turbulence model are calculated.

The methodology allows for a calculation of an uncertainty of the parameters established from a probability distribution.

The authors show and discuss the results of the parameter fitting also in comparison to other calibration studies of the DWM model.

The final evaluation is conducted by comparing the full DWM implementation with the calibrated submodels to a subset of measurements.

The manuscript is in general well written and the different steps of the analysis are well documented and explained.

The content is relevant to the scientific discussion of using lidar measurements for model validation/calibration, thus I recommend to publish the manuscript after minor revisions.

My main criticism of the manuscript is the lack of discussion of the generalisation of the results.

The authors compare their results of the wake deficit modelling with previous studies and claim e.g. that the model parameters in the IEC standard lead to conservative wake predictions.

This might be the case in this specific campaign, but I am missing the evidence that the results are fully transferable from the 190 kW V27 to multi-MW wind turbines.

This would imply a fully non-dimensionalized problem. It is not the fault of the study that this can not be proved on a single wind turbine, but the generalisation of the results should be handled with more care. I would have expected that discussion at the end of the manuscript.

Furthermore the final validation of the results compares the calibrated model with the measurements that are used for calibration 

(even if these are technically not the same measurements as measured in a different period). Thus, it represents rather a verification of the calibration than a model validation.

Further Comments:

1. Title - As part 2 of this study is only mentioned in the very last sentence of the manuscript and as it includes further measurements, my suggestion is to remove Part I from the title
2. L10 - We demonstrate that  Please edit according to the previous comments on the generalisation of the study
3. L42 - load responses mostly  Please rephrase
4. L46 - still under judgement -> still to be assessed
5. L71 - and at the DTU ...
6. 3.2.1 Atmospheric stability - Can the authors please explain the benefit of sorting the measurements into stability classes for the analysis. as the model has no dependency to atmospheric stability, but just to TI. Why not just sort by TI?
7. L249 - T is not a variable in equation (5), virtual potential temperature needs to be introduced
8. 262 - this is sufficient ... -> Why is this sufficient?
9. 442 - These deviations are mainly due ... Can you please elaborate what you mean here?
10. Figure 8 - As far as I understood all 4 distances are used for the parameter calibration. In the near wake there is not only a quantitative but also a qualitative disagreement between modeled and measured profiles as the double-Gaussian structure is not measured by the lidar. Have the authors considered to discard the near-wake measurements so they do not disturb the parameter fitting? Also, the uncertainty of the model predictions lie in, I would say most of the plots, outside the measurements. How is that possible? Might the uncertainty be underestimated by the uncertainty quantification approach?
11. Figure 9 - This is an interesting figure, but it should be pointed out that the parameters from the other studies were derived from completely different data sets. The different results might also be due to a lack of transferability to larger dimension turbines.
12. Figure 11 - I am confused by the Figure 11 and the corresponding description in the text. From the text I would expect two versions of the wake-added turbulence in the DWM model, one based on the parameter-fitting of eq. 3 and one based on the parameter fitting of eq. 17.
13. 6.1 Correction for rotor induction effect - It's unfortunate that the measurements had to be conducted in the induction zone of the turbine. The uncertainty introduced by the model correction has to be at least mentioned and preferable quantified.
14. Figure 14 - Please also show and discuss the results of DWM**, because the DWM* is as far as I understood not the DWM model as it would be applied based on wind measurements at or upstream of the turbine.
15. L655 -  Our result indicate ... Please also here consider the limits to generalise from these results

Overall Impression:

This is a very interesting manuscript that describes the steps to calibrate the DWM and how it compares to the measurement data. The manuscript is very well written and can be easily followed. The analysis is explained in a structured way and the reader can follow the steps carried out by the author. The content is very useful for the scientific discussion to discuss methods to calibrate the DWM, I thus recommend publishing the manuscript after minor revisions.

My main criticisms are the following:

• The manuscript can be shortened as some information’s are written multiple times.
• Some steps/ assumptions require additional description to clarify the used approach
• The validation can be more comprehensive by using data used in other studies to show that the newly proposed model is applicable for a wide range of conditions (turbine type, site conditions etc.)

Comments:

1. L145: instantaneous wake radius – ‘instantaneous’ would indicate that the wake radius is dependent on time. However, to my understanding the averaged N-S is used here. I would rephrase it.
2. L160: Is it assumed that the parameter k1 and k2 are independent of wind turbine design and ambient conditions? To my understanding, these parameters are dependent on the assumption that the pressure region ends at ~2D. Wouldn’t the pressure region change at different turbine designs and inflow conditions, such as the atmospheric conditions, shear, or veer, and require a new calibration of k1 and k2? Please add a reference where it states you can assume that these parameters can be considered constant
3. L165: In the entire manuscript the focus is on the wake-added turbulence. The first sentence and the equation of the wake turbulence can be removed to shorten the paragraph.
4. L205: Change the reference Herges et al. 2018, as it refers to Berg et al. 2014 regarding the power output
5. L205: Please also add the cut-in and rated wind speed of the turbine
6. L215: I would rephrase the sentence “The SpinnerLidar has been mounted either in the spinner or on top of the nacelle of a wind turbine.” In this study, only results have been used when the SpinnerLidar was installed on the nacelle
7. L235: How was the yaw-offset defined? Is it defined as the difference between the nacelle orientation and the wind direction or is it determined using a wind vane on the nacelle?
8. L250: Please indicate how the friction velocity is determined
9. L250: According to Pena et al. 2019 the distribution of stability changes with height. Why is the 18m – Sonic used in this study, while the 10m – Sonic is available?
10. L260: Please rephrase “sufficient”.
11. L270: Is it valid to assume that w=0 at 1D or 2D? At these distances, you are measuring in the near wake region, which is a very complex flow.
12. L280: How is the longitudinal difference of the points within each scan accounted for, while interpolating onto a 2m-grid, as the SpinnerLidar measures over a sphere?
13. L290: A bivariate Gaussian shape is fitted to obtain the wake center, where the wake center location is within ~10% of the lateral bounds of the scanning area. Would this mean that a part of the wake is outside the scanning area, while the fit is being performed? If this is the case is there a reference that shows that you can still accurately estimate the wake center position?
14. L305: ”... under varying stability regimes during strategy I”
15. L315: Is the assumption of no velocity gradient along the beam valid when the measurement is conducted at the edge of the wake?
16. L330: Please rephrase “low turbulence (indicate a range)
17. L350: Is strategy II used here to analyse the wake turbulence? (according to the description on p.8) It is unclear to me which data points are used here. According to the description on p.8 strategy II is used. However, Table 2 indicates that there are no measurements in neutral conditions at 7m/s or only one case in unstable conditions at 8m/s
18. L365: Was it not possible to determine the thrust coefficient in the free field during the measurement period? If it was possible, why was the thrust coefficient of the aeroelastic model used instead of the free-field data?
19. L440: The assumption of the near-wake region also introduces uncertainty on the deficit predictions at short distances
20. L455: I suggest shifting the comparison of the calibrated DWM and the accompanying figure to section 6 (Validation). Furthermore, I would also show how the calibrated DWM compares to the data set used in the other studies. As the measurement data in this study are used to calibrate the DWM it is no ‘surprise’ that the DWM has a better performance. The data sets of the other studies could be used as validation cases to show that the newly proposed model is also applicable for a wide range of conditions.
21. L550: “The wind speed is mainly lower than 9 m/s, thus WTGa1 operates below rated power.” This sentence can be removed
22. L555: It is not clear to me which sector of the measurements was corrected for the induction as the wake would not be affected by the induction zone with south-westerly winds (according to Figure 1). Were all the measurements obtained while the wind was coming from the south? Furthermore, I suggest showing the uncertainty of applying this correction and how it compares with measurement data not influenced by the induction zone.
23. Either use Figs or Figure to reference a figure.

1. Figure 1: Is the position of WTGa2 correct? In the description, it states that in strategy III a Lidar scan is performed at 5D behind the rotor. This would mean that the Lidar scan is performed at the same position where WTGa2 is located
2. Figure 4: Please indicate the number of lidar scans used to average the velocity deficit
3. Figure 6: Please indicate the number of lidar scans used to average the variance
4. Figure 10: No reference to this figure is found.
5. Figure 11: Please also include the profile of the wake-added turbulence with the ‘original’ analytical formulation.

1. Table 1: Please indicate the wind speed range e.g. [3 - 4] or [2.5 – 3.5].