the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Dust-producing weather patterns of the North American Great Plains
Abstract. The North American Great Plains are a semi-arid and windy environment prone to dust events that produce a variety of hazards to public health, transportation, and land degradation. Dust has substantial spatial variability across the plains, and the weather responsible for that dust is understudied in most of the plains, especially the North and East. Here we identify specific weather patterns associated with dust occurrence across the plains. We make use of an atmospheric classification that defines 21 weather patterns for the Great Plains that includes various stages of warm and cold frontal passages, northerlies, anticyclones, and summertime patterns not associated with mid-latitude cyclones. We use the time series of weather pattern to composite satellite daily dust observations from 2012–2021. We calculate average dust occurrence for each weather pattern, the contribution of each pattern to local dust loads, and identify the specific weather patterns most important to each location and subregion. We find no single weather pattern is responsible for dust occurrence in the plains, but that different patterns are responsible for dust in different subregions of the Great Plains. Passing cold fronts are most responsible for dust events in western Texas and New Mexico, southerlies are responsible in the northeastern plains of from Iowa to the Dakotas, and summer weather patterns produce the majority of dust in the High Plains from Colorado to Canada. Identifying the dust-producing weather patterns of particular subregions is a valuable step toward understanding dust variability and improving dust predictions, both present and future.
- Preprint
(2864 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 19 Nov 2024)
-
RC1: 'Comment on egusphere-2024-2820', Anonymous Referee #1, 28 Oct 2024
reply
This manuscript uses a previously designed classification of 21 different weather patterns conditions over the US Great Plains based on reanalysis data. These weather patterns are used to understand the different large-scale conditions which contribute to dust events in specific regions of the Great Plains. This is novel to previous research which tended to look at either climatology, or specific weather events in specific locations (e.g. ‘Albuquerque Lows’ determining dustiness in El Paso). Satellite dust observations are used in the study, leading to limitations in terms of time of overpass and cloudiness, though this is referred to and sufficiently caveated in Section 4.
Overall, the manuscript is coherently written and laid out with clear figures and appropriate methods for the study. This research provides important, new understanding on the conditions causing dust events in the US, which as the author has explained, have various impacts on populations across the length of the US. I only propose a few technical corrections for the author.
Line 11: should be of or from, not both.
Line 36-37: Something odd happening with the references repeating.
Line 93: Spelling - ‘wins’ should be ‘winds’
Line 110-111: LT or UTC?
Figure 4: It may be useful to include in the caption the patterns associated with each of the categories.
Line 173: ‘southerlies and southerlies’
Figure 5: There appears to be a rogue line on the map extending from the bottom of Texas.
Line 214: ‘Only Pattern 5, increasing at 1.1 days/year, has a statistically significant trend (95% confidence) over the period 2012-2021’. Was this discussed in the paper? This feels like it has come out of nowhere in the conclusions with no reference to looking at trends previously.
Citation: https://doi.org/10.5194/egusphere-2024-2820-RC1 -
RC2: 'Comment on egusphere-2024-2820', Precious Ebiendele & Adeyemi Adebiyi (co-review team), 10 Nov 2024
reply
The paper identifies the weather patterns across the North American Great Plains into 21 categories and connects them to VIIRS-retrieved dust observations over the region.
While the classification method was originally defined in a different paper (Evans et al. 2017), the author provided a brief summary that was largely limited in delivering the needed overview necessary for a reader who may not want to go through the other paper. For example, it is not clear in this paper why 21 patterns were identified or the reason for limiting the Great Plains to the area defined. These issues must be addressed here to have a complete paper and not leave readers in a state of confusion.
In addition, the authors identified several uncertainties associated with the observational data used for dust identification. Despite these uncertainties, it is unclear why the author decided to use this dataset over other equally available and potentially better datasets. For example, the author cited Ginoux’s paper that used MODIS. With a better dust-retrieving algorithm than VIIRS, the author did not provide a valid reason for their decision or how potentially the results would be different if this other dataset was used. We suggest that additional analysis should be done with at least one other similar dust observation that will mitigate some of the uncertainties identified with dust retrieval in VIIRS.
Also – there is a fair amount of confusion around how the composite analysis for dust was performed. The author mentioned that four times daily ERA-interim datasets were used for the weather pattern (Line 75), but daily observation was used for the VIIRS dust (Line 117). How does the author reconcile that difference? Similar to the comment above, this suggests that a better temporal dust retrieval, such as dust RGB from GEOS, maybe a better alternative than the VIIRS used in this analysis.
Other Comments:
Line 11: remove “of”
Line 83: Do you mean A1 instead of S1? Also, why is Fig. A1 in the main manuscript?
Are Fig. 1 and A1 showing for 2012-2021 that are used for dust classification, or are they showing for 1996-2010 used in the other paper? This should be identified in the caption and stated clearly in the text. Given that 1996-2010 is entirely irrelevant to the results of the dust pattern shown in Fig. 3, 4, and 5, I hope Fig. 1 and A1 are showing for 2012-2021. Otherwise, the figure should be changed and results re-interpreted if the pattern is different.
Line 76-80: The authors used a "k-medians classification algorithm to identify and define commonly occurring weather patterns for the region." There are many other clustering methods (e.g., hierarchical, K-means); why do the authors consider the K-medians algorithm? I recommend that the authors provide a more detailed justification for their decision.
Lines 111-115: The authors stated, "We do not capture dust that initiates after the overpass or occurs beneath clouds." Possible bias could result from undercounting, especially during seasons with higher cloud cover. Could the authors provide details on the potential significance of undercounting in regions or seasons with high cloud cover?
Lines 145-146: The authors' findings indicate that "Pattern 4" is the primary source of dust in the region, as illustrated in Figure 3. Could the authors explain how this aligns or differs from previous studies over the northeastern plains?
Line 220-222: "The summer weather patterns (Patterns 17-21) likely undercount dust more than others. Since convective summer dust events are likely frequent and short-lived, this could be a significant constraint. I suggest the authors discuss how to address this bias. The authors could discuss possible alternative satellite products or station-based observations to address this.
Citation: https://doi.org/10.5194/egusphere-2024-2820-RC2
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
159 | 49 | 8 | 216 | 0 | 6 |
- HTML: 159
- PDF: 49
- XML: 8
- Total: 216
- BibTeX: 0
- EndNote: 6
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1