the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Dec 2024
Relationships Between Surface Fluxes and Boundary Layer Dynamics: Statistics at the Land-Atmosphere Feedback Observatory (LAFO)
Abstract. We used a combination of two Doppler lidars (DLs) and an eddy covariance station at the Land-Atmosphere Feedback Observatory (LAFO), Stuttgart, Germany, to investigate relationships between surface fluxes, convective boundary layer (CBL) height, and profiles of vertical wind variance, horizontal wind variance and turbulent kinetic energy (TKE). One DL was operated in vertical-pointing mode and the other in six-beam scanning mode. Daytime statistics were derived from 20 convective days from May to July 2021. In this data set, the mean CBL height ⟨𝑧𝑖⟩ showed a maximum of (1.53 ±0.07) km between 13:00 and 14:00 UTC, which is about 1.5 to 2.5 hours after local noon. We found counterclockwise hysteresis patterns between the CBL height and the surface fluxes. In the development phase, these relationships were approximately linear. In the early afternoon, the relationships reached a peak phase with both large fluxes and high values of ⟨𝑧𝑖⟩. At 12:00 UTC, just after local noon, the maximum values of vertical, horizontal, and total TKE were 0.55 m2s-2, 1.26 m2s-2 and 1.71 m2s-2 at heights of (0.30 ± 0.06)⟨𝑧𝑖⟩ , (0.56 ± 0.06)⟨𝑧𝑖⟩, and (0.40 ± 0.06)⟨𝑧𝑖⟩, respectively. In the decay phase in the late afternoon, the relationships show non-linear patterns with larger values of ⟨𝑧𝑖⟩ for the same surface fluxes than in the morning. Furthermore, we show relationships between the vertical and horizontal components and total TKE.
- Preprint
(1819 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 30 Jan 2025)
-
RC1: 'Comment on egusphere-2024-3878', D. D. Turner, 26 Dec 2024
reply
This paper analyzes 3 months of turbulence data (TKE, vertical velocity variance, and horizontal velocity variance) using observations from a flux tower at the LAFO site at the University of Hohenheim in southwestern Germany. Generally speaking, this paper is well-written and the results are both interesting and relevant. I am very supportive of the analysis of multi-month datasets to help understand what drives turbulent motions in boundary layer.
I have only a small number of comments, all of which should be addressed before the paper is accepted for publication.
First: while the abstract is clear, I think the title of the paper should be modified to state “Daytime” somewhere – perhaps “…and Daytime Boundary Layer…”
Second: given the physical arrangement of the LAFO site, it would seem that the measurements made by the DLs and flux station are really only connected if the wind direction was either from NNW-to-SSE or vice-versa; i.e., largely along the line between EC1 and the remote sensing site (Fig 1). However, no information was given on the mean wind direction in the CBL for the 20 days used in this analysis. If the analysis was restricted to winds that are largely along this axis, would the results be the same?
Third: on line 197, the authors indicated they removed points that were 4 standard deviations away from the mean. This is a symmetric test; however, it is well-known that there can be significant skewness in vertical velocity profile (e.g., the Berg et al. 2017 paper that was referenced). Have the authors considered using an asymmetric test to identify outliers that preserves the skewness? This might have an impact on the variances they derive, which might be contributing to the occasional times when TKE_v > TKE_tot (line 375).
Fourth: on line 207, the authors indicated that they used 15 lags for all of their autocorrelation analysis. This only works if the integral time scale is larger than 150 s for all of their cases; I would be surprised if this was true. They should only use lags from lag 1 to lag N where all N of the lags have significantly positive autocorrelations (not zero within uncertainty or negative). Would the authors please speak to this?
Fifth: it is not clear what scanning strategy is used for the 6-beam scanning approach in section 3.2. In particular, they indicate on line 226 that they are using the same approach to remove instrument noise from each of the beams (presumably they mean the AC method outlined in section 3.1); however, how was the integral scale (and hence number of lags) used here? I presume that at each beam angle, multiple 1-s samples were collected before the scanner moved to the next beam – however, I also assume that data was collect at each beam angle multiple times in the 30-min time window. More details are needed here to truly understand how they are doing their analysis.
Lastly: in Figure 9, I believe they are showing the mean values throughout the CBL of TKE_v, TKE_tot, and TKE_h. Have the authors investigated the how these relationships change from the middle of the CBL (where all three of these terms are near their maximum values) vs near zi? In other words, would it be possible to break figure 9 into two figures; one to look at these scatter plots (relationships) between 0.3 to 0.6 zi, and other for 0.8 to 1.0 zi?
Citation: https://doi.org/10.5194/egusphere-2024-3878-RC1 -
AC1: 'Reply on RC1', Syed Saqlain Abbas, 08 Jan 2025
reply
Dear Dr. David D. Turner,
Thank you for your valuable comments on our manuscript. We will address yours and other reviewer comments, in the revised version of the manuscript.
Best regards,
Syed Saqlain AbbasCitation: https://doi.org/10.5194/egusphere-2024-3878-AC1
-
AC1: 'Reply on RC1', Syed Saqlain Abbas, 08 Jan 2025
reply
-
RC2: 'Comment on egusphere-2024-3878', Anonymous Referee #2, 07 Jan 2025
reply
Review on Sved Abbas et al.: Relationships Between Surface Fluxes and Boundary Layer Dynamics:
Statistics at the Land-Atmosphere Feedback Observatory (LAFO).
The authors used data from two Doppler lidars and one eddy covariance station from the Land-Atmosphere Feedback Observatory (LAFO) in Germany, to investigate relationships between surface fluxes, CBL height, and profiles of vertical and horizontal wind variance and TKE. In total, 20
convective days from May to July 2021 were used. The days include cloud-free and cloud-topped CBLs.
Overall, the analysis and evaluations lag behind what has been provided so far in previous studies on that topic. For example, a solid comparison with results from previous studies (some possible studies for comparison are listed below) is missing. However, as the turbulence data in this study are not normalised by w*, a comparison is difficult anyway. Even the aim of the study is not clearly specified so that even the last sentence in the summary about future work is nearly consistent with the objective of this study. I suggest major modifications.
Section 1 (Introduction)
General comments: The introduction is not well structured and should be modified.
Instead of citing the original papers about basic findings/knowledge on turbulence and ABL, quite new papers are referred to (so, at least cite the originals, too).
The introduction contains several unnecessary repetitions.
Not all statements are precise or correct.
Just some examples of the above statements:
Line 22-23: “In daytime, the ABL becomes convective and turbulent processes dominate.” By far not all ABLs are or become convective during daytime.
Line 24: …”with a morning transition to a maximum.” To a maximum of what?
Line 25 and 26: “Turbulence in the CBL is driven primarily by surface heating (Manninen et al., 2017), wind shear, and tropospheric entrainment (Wulfmeyer et al., 2016).”
This info is not new since 2016 or 2017 (as you suggest with these citations) but basic knowledge since half of a century (e.g. Kaimal et al 1976 or Caughey and Palmer 1979). So, you should primarily refer to the corresponding textbooks from e.g. Stull (1988: An introduction to the atmospheric boundary layer), Sorbjan (1989: Structure of the atmospheric boundary layer), Garratt (1994: The atmospheric boundary layer, Wyngaard (2010: Turbulence in the atmosphere) …… Lee (2023: Fundamentals of boundary-layer meteorology).
Concerning the conditions at the surface, soil and the energy balance equation of the Earth’s surface (Munn, 1966; Oke, 1987; …… Arya, 2012 and Foken, 2017) you can also easily refer to the basics (i.e., radiation, soil heat flux, sensible and latent heat flux depend all on the conditions at the Earth’s surface) – available in SVAT model descriptions. In the introduction the same information (repetition) is distributed over different lines like in line 28 “vegetation characteristics, and soil moisture”, line 30 “and soil/vegetation fluxes”. Line 33: “transpiration of vegetation”.
Repetitions: information in lines 59-63 (e.g. “The land surface properties influence the partioning of turbulent fluxes (Chen and Lo, 2023”) is already mentioned several times before, e.g. in lines 20-36!
In line 69 you mention that you use data from one EC station only instead of the two stations which are available. I don’t know if you justify this in the following but with respect to surface and soil heterogeneity and corresponding spatial differences in fluxes, as mentioned in the previous part of the introduction, the restriction to one EC station only surprises me.
The aim of the paper (lines 71-72) is not very specific with respect to the problems raised above. E.g. precipitation in line 33. In lines 56-58 the energy balance closure problem is mentioned but is it part of the aim of this paper? Turbulence differences between cloud-free and cloud-topped CBLs. What exactly is intended to attack here? Parameterization? Model evaluation? Could you be more precise!
Section 2.1
Line 82: If elevation is given in m AMSL please add AMSL
Line 84: “As part of this synergy, Doppler lidars (Halo Photonics), a cloud Doppler radar (Metek GmbH), a water vapor differential absorption lidar (DIAL) (Späth et al., 2016), and a water vapor and temperature rotational Raman lidar (RRL) (Lange et al., 2019) are deployed to observe land surface fluxes …”. How can you measure surface fluxes with the instruments listed in the first part of the sentence?
Line 95: Specify “among other instruments” or delete.
Figure 1: could you add a figure which shows the surroundings of LAFO, e.g. where in Germany is LAFO situated.
Line 84: why is the synergy of ….. the LAFO instruments unique? As far as I know there are several other similar observation platforms (Foken, 2021: atmospheric measurements, e.g. chapter 4.7). So, LAFO is one of these platforms and not unique.
Line 106-107: repetition from the introduction. Delete!
Section 2.1.1
General question: What are the measurement heights of the sensible and latent heat flux? Concerning radiation and soil heat flux: Figure 1 suggests that these measurements are made between the two fields. For which of the two vegetations are the radiation and soil heat flux measurements representative? E.g. albedo, surface temperature and soil moisture could be very different!
Line 120: How can the footprint always be within …. fields 4 and 5. For my understanding the footprint could be either within 4 or 5.
Section 2.1.3
What is the vertical resolution of the radar data? Vertical extent, Temporal resolution?
Section 2.1.4
Line 143-144: How did you calculate the cloud cover or better cloud fraction? Is it low, middle or high cloud fraction? Is it from temporal or spatial observations?
“days with cloud cover <40% during daytime, to ensure that the CBL was indeed developed by convection.” Be careful with this statement. The role of ABL clouds on convection is not that clear. This is because CBL clouds might be a significant source convection due to elevated heat release, i.e. influencing the vertical wind variance and TKE in the upper part of the CBL. There are several studies on that topic, e.g. Garratt (1994), Neggers et al. (2003), e.g. Hogan et al. (2009), Chandra et al. (2010) and Lareau et al. (2018).
Lines 144 and 154: Another example of repetiting information: Line 144: “we selected only days with cloud cover <40% during daytime”
and
Line 154: “For our analysis, only those days are selected when cloud cover is <40%, otherwise the days were excluded.” Even in this last sentence the information in the first part and second part of the sentence is redundant!
Line 159ff: it would be good to define the different kinds of averages: bar and bracket. See also Eq. 8.
The description of the energy balance components is too superficial. Example: “In Fig. 2a and Fig. 2b, the mean net radiation ⟨ 𝑄𝑛 ⟩ and the mean incoming solar radiation ⟨ 𝑄𝑠⟩ follow typical diurnal patterns. Before sunrise, ⟨ 𝑄𝑛⟩ is negative while ⟨ 𝑄𝑠⟩ is zero. During daytime, both are positive.”
“In Fig. 2d, the mean of the latent heat flux ⟨𝐿⟩ exhibits a similar pattern.” Similar pattern in comparison to what? Compared to the sensible heat flux? The latent heat flux is not negative in the afternoon.
What I mean with superficial: the information which is of most interest is not given or discussed. This is - concerning the fluxes - what about the energy balance closure problem (mentioned in the introduction)? Does the imbalance depend on the wind direction? Are there differences from day to day? What are reasons/differences for times with sensible heat fluxes of about 20 W/m2 or 250 W/m2 during daytime? How do you explain latent heat flux values of nearly 200 W/m2 between 02 and 03 UTC or 21 to 24 UTC. Any explanation for the ground heat flux of about 130 W/m2 before sunrise?
I would prefer to see and discuss diurnal cycles of wind speed and wind direction as well as cloud fraction of low, middle and high clouds in combination with the energy balance components and energy balance closure. E.g. it is essential for the CBL evolution to know whether the sensible heat is -50 W/m2 or +100 W/m2 at 15 UTC. What happens with the turbulence fluxes when the wind direction changes? Does the wind direction change during daytime at all? This would also have an effect on humidity and temperature. I also wonder about days when the 2 m temperature was 0 °C. Normally, the soil should be frozen on those days.
Section 3.1
Line 209: “The inertial subrange is the time interval in which the turbulence scales are locally homogeneous and isotropic within the CBL (Wulfmeyer et al., 2016).” Ones again: this is not a new finding from Wulfmeyer et (2016) but already stated by Obukhov (1941).
Line 190 and 201: repetition: “known as autocovariance technique, is based on the fact that 𝑤′(𝑥, 𝑡) is correlated in time whereas 𝜀(𝑥, 𝑡) is not.” And “because noise is not correlated on time”.
General question and comment: When you use 30 min averages, you already cut the contribution of convective cells in the turbulence spectrum for periods of about ≥ 15 min. I expect that would be quite high when a larger convective cell dominates your 30 min time interval (see Figure 3, e.g. on 14 and 15 June). Have you look at that? Have you looked at the diurnal cycle of to see how much of the turbulence is cut by you short averaging time? This also holds for the calculation of TKE in section 3.2. If you consider Lenschow’s “how long is long enough”, your time interval could be too short (statistical problem). I don’t know if these problems are discussed in the result sections.
Section 3.2
Eq. 8: Why are variances (bars) and covariances (bars and brackets) treated differently? i.e. different kind of averaging.
Line 226: Section 2.2.1 does not exist.
Line 210 and 234: The TKE-section and boundary layer depth-section are numbered equally: 3.2! The numbering of the sections is very sloppy. See comments before.
Line 245: the criterion “fit best for our dataset.” is a very subjective one. A more objective argumentation is needed here. Have you considered to use the noon-time radiosounding from Stuttgart for comparison.
Line 240: How can you state that your turbulence is surface-based? “The maximum height of surface-based turbulence can be estimated”. I hope you will discuss/explain the strong zi-increase on 17 June (elevated turbulence) and on 19 June in connection with surface-based processes.
To gain some insight into these day-to-day variations of zi, information on the wind speed and wind direction in the CBL could be helpful. Wind direction also could give hints concerning footprint of the convection measured by the wind lidars.
Section 4.1
Line 295: How do you explain a CBL growth when the sensible heat flux is negative?
Line 304: “…. a cooling effect …” Where?
305: “…. due to vigorous convection can inhibit the sensible heat ….”. a) Do you mean “sensible heat flux”? 2b) what generates the vigorous convection when the sensible heat flux is quite low? Representative of the fluxes measured at EC1.
Lines 307-308ff: “Incoming solar radiation heats both ground and subsequently the air above it” It’s the sensible heat flux divergence that heats the air above the ground and not the radiation!
Line 308: Figure 2f does not show a temperature gradient – just the 2 m temperature.
Line 310: It would be helpful to add an equation which includes the physical processes you refer to when you write “consequently TKE leads to increase the CBL during daytime in a linear
manner.”
Line 311: What do you mean with “accumulated surface heating”?
Line 314-315: To which levels do you refer to when you write “leads to lowering the temperature gradient”?
General comment on Section 4.1:
(a) In order to see if your zi evolution is mainly based on the sensible heat flux and entrainment – as assumed for this investigation it would be helpful to check this. What would you expect concerning the zi-growth rate over homogeneous terrain when applying common models/equations describe the correlation between the sensible heat flux (Stull 1988) or Batchvarova and Gryning 1991) which? This would also help to understand some of the days with extreme (high and low) zi.
(b) to investigate the mean behaviours of S and zi will always show a correlation. I guess that using the flux data from EC2 would result in a similar. However, to achieve a better knowledge of the dependence of zi on S, the understanding of the individual days would be helpful. Clustering concerning CBL clouds, CBL wind speed and direction.
(c) What is the footprint for zi during daytime hours? How representative is the EC1? May the CBL be affected by the forest in the west?
Section 4.2
Lines 334-335: “Our findings for ⟨𝑇𝐾𝐸𝑉 ⟩𝑚𝑎𝑥 agree with the results of previous studies (Lenschow et al., 2000; Dewani et al., 2023; Wulfmeyer et al., 2024). What do you mean with TKEmax agrees with previous results? As your values are not normalised by w* (e.g. Lenschow et al. (1980) and Sorbjan (1989), a comparison is not very meaningful.
Lines 350: “The sunlight heating the land surface triggers convection …”. Physically, it’s the sensible heat flux that causes convection! E.g. if solar radiation is absorbed by the surface completely, convection would hardly be triggered at all. What I mean: be more precise with your statements!
General question: You state in line 233 that you calculated w’ based on the 6-beam scanning technique. In parallel, you received w’ based on the vertically pointing lidar (section 3.1). What are the differences between both calculation of the vertical wind variance?
Section 4.3
Line 375-376: Are trivial sentences like “As expected, ⟨𝑇𝐾𝐸𝑡𝑜𝑡 ⟩ is generally larger than ⟨𝑇𝐾𝐸𝑉 ⟩ for the same range and time as expected because the vertical wind variance is one component of TKE.” really necessary?
Lines 380ff: here you start discussing the turbulence for different periods of the day based on Figure 9. However, in Figure 9 the data are not separated by hours but for 9-18 UTC only!
Section 4.2, 4.3
The analysis of the observations in general is quite superficial. Mean values (without clustering) do not really improve our knowledge of the CBL. This is one of the reasons why the authors do not provide more insight into CBL processes but just come up with such sentences in comparison to previous findings like “findings for ⟨𝑇𝐾𝐸𝑉 ⟩𝑚𝑎𝑥 agree with the results of previous studies (Lenschow et al., 2000; Dewani et al., 2023; Wulfmeyer et al., 2024).” Only considering different conditions would have allowed to gain a deeper insight in the CBL processes.
A thorough comparison and discussion with results from e.g. Dewani et al 2013, Maurer et al 2016, Kiseleva et al 2024 could improve the benefit of this paper. For example, if days with cloud-topped CBLs and clear-sky CBLs would have been discussed separately (clustered), the observations could have provided a contribution to open question like the impact of CBL clouds on CBL turbulence (e.g. Hogan et al. (2009), Chandra et al. (2010) and Lareau et al. (2018). Another clustering would be CBL wind speed.
Miscellaneous and Typos
If you use abbreviations for variables you should introduce the abbreviations when you mention them for the first time. When abbreviations are introduced for variables you should only use the abbreviations in the following (e.g. vertical wind variance in line 36 but the abbreviation is only introduced in line 45).
As mentioned, the numbering of the sections as well as referring to sections in the text is very sloppy. A little more care would make the reviewer's job easier! (See details above and below).
Line 50: “solar radiation are result of time” should be “solar radiation as result of time”
Line 61: partitioning in of portioning
Line 127: “2.1.3 Vertical Pointing and Six-Beam Scanning Doppler Lidars” This should be 2.1.2
Line 147: “ms−1.” Without a blank, ms means millisecond! Use m s-1 (occurs several times - check in the whole text!)
Line 150: through instead of though?
Figure 2f: °C instead of C°
Line 196, 199, 204, 206 …… : “.” Instead of “,” because you start with a capital letter in the following line (197, 200, 205, 207).
Figure 4 (figure caption): “…. of the CBL 𝑧̅𝑖. for all selected days. (b) same as (a) but as color plot (c)”: delete full stop after zi and add a full stop before (c). “same” should be “Same”, i.e. use capital letter.
Line 300: 11:30 instead of 11.30 UTC
Line 302: 10:00 should be 10:00 UTC
Eq. 9: normally, v’ is defined as v’ = and not the other way around.
Citation: https://doi.org/10.5194/egusphere-2024-3878-RC2 -
AC2: 'Reply on RC2', Syed Saqlain Abbas, 09 Jan 2025
reply
Dear Reviewer,
Thank you for your valuable comments on our manuscript. We will address yours and other reviewer comments in the revised version of the manuscript.
Best regards,
Syed Saqlain AbbasCitation: https://doi.org/10.5194/egusphere-2024-3878-AC2
-
AC2: 'Reply on RC2', Syed Saqlain Abbas, 09 Jan 2025
reply
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 140 | 42 | 10 | 192 | 2 | 1 |
- HTML: 140
- PDF: 42
- XML: 10
- Total: 192
- BibTeX: 2
- EndNote: 1
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1