the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Aug 2024
Statistical and neural network assessment of climatological features of fog and mist at Pula airport in Croatia: from local to synoptic scale
Abstract. A study was conducted on the climatological characteristics of fog and mist at Pula Airport in the northeastern Adriatic, using statistical and machine learning approaches. The study utilized meteorological data from Pula Airport, along with satellite sea surface temperature (SST) data from two coastal areas west and east of the airport, to gain insights into the influence of sea temperature on fog formation. To identify weather patterns associated with the occurrence of fog and mist, wind and mean sea-level pressure (MSLP) data from the ERA5 reanalysis were analyzed using Growing Neural Gas (GNG), a machine learning method. A notable finding was a declining trend in the frequency of fog and mist at the airport, which can be linked to the results of the GNG analysis of the ERA5 data. This analysis showed a decrease in synoptic patterns favorable for fog and mist. Fog occurs mainly between October and March and is primarily associated with weak easterly and northeasterly winds. Additionally, fog is more likely to occur when the sea surface temperature is higher than the air temperature. Mist has similar characteristics to fog, although it is more likely to occur with easterly winds.
- Preprint
(2728 KB) - Metadata XML
-
Supplement
(24164 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on npg-2024-18', Anonymous Referee #1, 15 Sep 2024
Line 143
Remove the ‘24’
Figure 2b
The legends are overlapped.
Line 212
‘Summarizing the data ‘ should be ‘A summary of the data’
Figure 5
Explain what ‘VRB’ is in the caption.
Figure 9
Use a dot to indicate where the airport is.
The coastal line is not clear. Should use bold line
In the caption, say that you only plot the wind vectors and shadings for the sea
indicate the contour lines are the MSP
The wind vectors are not visible near the airport. It is better to plot additional panels (e, f, g, h) focusing on the region near the airport. This smaller region can be the one in Figure 1b. Same for figure 10-13.
Figure 10-13.
The captions can be simplified to ‘similar to Figure 9 but in March’, and so on.
Line 433-540
The Discussion and summary section is too long. It is difficult to understand the main point.
Many contents should be integrated into section 3.
Citation: https://doi.org/10.5194/npg-2024-18-RC1 -
RC2: 'Comment on npg-2024-18', Anonymous Referee #2, 20 Nov 2024
1. General comments
Using simple statistics and a more sophisticated artificial neural network approach, this study aims at describing synoptic conditions leading to the formation of fog and mist in the vicinity of Pula airport (Croatia). It is an interesting research topic, knowing the negative impact of fog/mist on the airport operations.
However, there are several problems, listed below, that need to be addressed before proceeding further. Also, the text is sometimes very scholar. It would be very beneficial for this work to have a more scientific writing.
I therefore recommend major revisions.
2. Specific comments
Following are some important comments on the work that authors should consider during the revision.
2.1. Content
The “Discussion and conclusion” part is too long and often difficult to follow. Most of the content is just a repetition of what has been presented in the previous section, and should be merged with section 3. According to the journal guidelines (https://www.nonlinear-processes-in-geophysics.net/submission.html#manuscriptcomposition), the last section of the main text should only presents the conclusions.
Several parts need to be rewritten to improve the quality of the manuscript: - L.296-298: What is the meaning of these values? - L.298-304: I could not follow this part. Please clarify. - L.413-428: Please rewrite to include information on linear trends and implication for the fog/mist.
Also, there are unnecessary information throughout the main text that should be remove to make the manuscript clearer. For example: - L.75-80 are not necessary - Same on L.98 (information on population) - L.439-441 (or rephrase and include in the introduction)
2.2. Methodology
The choice of 9 best matching units (BMUs) needs to be discussed more. This choice seems to follow Matić et al. (2022, who did not discuss this choice either). Also, Matić et al. (2022) were working on September only, while the current study applies the Growing Neural Gas (GNG) to six months. So the question of a fixed number of BMUs for different months rises question and should be discussed.
For example, different quality measurements are available to assess the clustering obtained from the self-organizing maps (SOM), such as the quantization error, the topographic error or the percentage of explained variance (Bauer et al., 1999; Kaski and Lagus, 1996; Kohonen, 2001; Kohonen et al., 2009). So it could be possible to use similar measurement with the GNG to select the best number of BMUs for each month. Also, please add a short supplementary text to explain more what “the method outlined in Matić et al. (2022)” (L.182) is and how it has been adapted to the current work.
The authors say that “[...] the GNG algorithm was applied only to 10-m wind data from ERA5, and the derived pressure fields were subsequently extrapolated” (L.316-318). What is a derived pressure field? And why is it extrapolated? Also, removing the MSLP from the analysis is an adaptation of the original work by Matić et al. (2022), which needs to be described either in the main text or as a supplementary text.
The selections described in L.337-341 are not scientifically based and must be discussed. Particularly, the study states that GNG is applied to all the months of the year, to finally focus to only half of them. One solution could be to use Fig.2b (with the changes suggested below) to get rid of the arbitrary choice.
2.3. Figures 9-13
The comment “[...] local variations such as topography or land features can have a large influence on the results.” (L.330) contradicts the statement “[…] considering the specific terrain and coastline features...” (L.93). Displaying the arrows over land is important to capture the regional patterns of atmospheric circulation near the surface. Please include the wind over land.
For some BMUs, the orientation of the arrows looks different from the wind origin reported in Table 1. For example, the wind of BMU-1-8 is WNW/NW, but in the associated map, the main flux over the Istrian Peninsula is more SW. Same for BMU-3-1. Please check.
Also, for some panels, there is a discrepancy between the contour and the orienation of the arrows (e.g., BMU-3-1, BMU-3-3, BMU-12-1). Please check.
3. Technical corrections
3.1. Figures
Figure 1: Please move labels “a)” and “b)” at the top of each panel. What is the meaning of “20” and “24” in the legend? If not necessary, please delete. a) The bathymetry is not necessary, please simplify the information. Also, the inset should not go beyond the edges of the upper panel.
Figure 2b: It is better to center the barplot on boreal winter, so we have a clearer view. And this will help for the choice of the months the GNG is applied to (see comment above). Also, the legends are overlapping, please fix.
Figures 5 and 6: What it the meaning of “VRB”?
Figure 9: Reshape the panels to have January (February) in top (bottom) row. Also, it is better to have the label above each panel, rather than on the right. Finally, please add in the caption the meaning of the arrows, contours and shading.
Figure 10–13: Please move the labels above the panels. Also, please simplify the figure captions by referencing to the Fig. 9.
Figure 14: Add labels (a)-(q) and use in the main text.
3.2. Tables
Table 1: “The slope coefficients describe the linear trends of the most common BMUs, i.e. the yearly change in frequency”. Is it related to Fig.14? If yes, please add a reference to it.
3.3. Text
The title is rather long. Authors could consider to shorten it (Deng, 2015).
L.90: Citation is incorrect/incomplete.
L.146: What is the meaning of “METAR” and “SYNOP”?
L.161: What is the meaning of “level 4”?
L.174: Why this region has been chosen? Is it based on Matić et al. (2022), who used the same region in their study?
L.203-205: Please use the value of the slope to describe the trend.
L.296: “[…] confirmed by calculations”. What does it mean?
L.312: It is not clear what the author mean by “to extract characteristic temporal and spatial patterns”. Please clarify.
L.326-328: Not sure about the meaning of the sentence. Could the authors explain further?
L.380-381: What mechanism(s) are in action?
L.395-396: “cyclonic conditions” rather than “cyclone” maybe?
L.469: “VRB” must be defined earlier.
L.513: There was not previous mention of global warming/climate change and their effect on the frequency of fog/mist days. This should be mention in the introduction.
L.557-674: Please provide the article titles in sentence case.
L.575: Journal abbreviation.
L.586: Please supply the full author list with last name followed by initials. (https://www.nonlinear-processes-in-geophysics.net/submission.html#references).
L.589: Same as L.586 comment.
L.591: Same as L.586 comment.
L.616: Incorrect title of the paper.
L.618: Same as L.586 comment.
L.620: Same as L.586 comment.
L.621: Journal abbreviation.
L.670: Same as L.586 comment.
4. References
Bauer, H.-U., Herrmann, M., and Villmann, T.: Neural maps and topographic vector quantization, Neural Netw., 12, 659–676, https://doi.org/10.1016/S0893-6080(99)00027-1, 1999.
Kaski, S. and Lagus, K.: Comparing self-organizing maps, in: Artificial Neural Networks — ICANN 96, Berlin, Heidelberg, 809–814, https://doi.org/10.1007/3-540-61510-5_136, 1996.
Kohonen, T.: Self-organizing maps, Springer Berlin Heidelberg, Berlin, Heidelberg, https://doi.org/10.1007/978-3-642-56927-2, 2001.
Kohonen, T., Nieminen, I. T., and Honkela, T.: On the quantization error in SOM vs. VQ: a critical and systematic study, in: Advances in Self-Organizing Maps, edited by: Príncipe, J. C. and Miikkulainen, R., Springer Berlin Heidelberg, 133–144, https://doi.org/10.1007/978-3-642-02397-2_16, 2009.
Matić, F., Džoić, T., Kalinić, H., Ćatipović, L., Udovičić, D., Juretić, T., Rakuljić, L., Sršen, D., and Tičina, V.: Observation of abrupt changes in the sea surface layer of the Adriatic sea, J. Mar. Sci. Eng., 10, 848, https://doi.org/10.3390/jmse10070848, 2022.
Citation: https://doi.org/10.5194/npg-2024-18-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 190 | 26 | 14 | 230 | 29 | 12 | 10 |
- HTML: 190
- PDF: 26
- XML: 14
- Total: 230
- Supplement: 29
- BibTeX: 12
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1