the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Jun 2024
Tropospheric Links to Uncertainty in Stratospheric Subseasonal Predictions
Abstract. Variability in the stratosphere, especially extreme events such as Sudden Stratospheric Warmings (SSWs), can impact surface weather. Understanding stratospheric prediction uncertainty is therefore crucial for skillful surface weather forecasts on weekly to monthly timescales. Using ECMWF subseasonal hindcasts, this study finds that stratospheric uncertainty is most strongly linked to tropospheric uncertainty over the North Pacific and Northern Europe, regions that can modulate but also respond to stratospheric variability, suggesting a two-way propagation of uncertainty. A case study of the 2018 SSW event shows an initial poleward and upward propagation of uncertainty from tropical convection, followed by a downward propagation where ensemble members that accurately predict the SSW also better at predicting its downward impacts. These findings highlight the locations in the troposphere that are linked to stratospheric uncertainty and suggest that improved model representation of tropospheric mechanisms linked to polar vortex variability could enhance both stratospheric and extratropical surface prediction.
- Preprint
(4736 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 17 Jul 2024)
-
RC1: 'Comment on egusphere-2024-1652', Anonymous Referee #1, 20 Jun 2024
reply
General comments
The paper is well written and relevant for the community. Its aim is to evidence the link between tropospheric and stratospheric uncertainty in a climatological context, using subseasonal ensemble hindcasts issued by ECMWF. The focus on the example of the 2018 SSW prediction is useful to make a link with individual subseasonal forecast cases. After the comments below have been addressed, I will be happy to advise pubblication of the manuscript in Weather and Climate Dynamics.
Specific comments
Lines 23-24 : From this sentence it seems as though the only sources of uncertainty for the stratospheric variability are the mean state of the stratosphere and the vertical wave propagation from the troposphere. Could you make clear that vortex preconditioning, also regulated by internal stratospheric oscillations, is an important factor modulating vertical wave propagation from the troposphere (hence stratospheric variability) ? For discussing this, I would advise to check the references in the second and third paragraph of the introduction of De la Cámara et al. 2019.
Line 78 : Your measure of (average) spread, used to separate hindcasts in large and small uncertainty groups, mixes short-term and subseasonal time ranges. Would the hindcast selection change if the spread was averaged starting from time +1 week (or +10 days) until the end of the hindcast period? Can you re-compute Figure 1 with this new selection as a sensitivity test ?
Figure 2 : It would be useful to show significance from the hindcast climatology (in a,b,c,d) and between the two hindcasts groups (in c,f).
Lines 218-232 : While the effect of mid-latitude precursors on the wave-2 propagation and the SSW development is quite clear and interpretable, the impact of the tropics is more subtle, and is suggested with a certain caution also in Statnaia et al. 2020. In Figure 5 a,b you show the tropical OLR before the event. How do you connect this directly to the anomalous propagation of wave-2 ? Does the mean error of the two ensemble means from the reanalysis in terms of OLR provide evidence that the tropics are better captured by the SSW ensemble ? In my opinion, in the Conclusions you stress a lot the impacts of the Tropics although the evidence supporting this is rather weak. The same also applies to a sentence in the Abstract.
Line 207 : It would be interesting to mention the reason for the large-spread ensamble having a stronger vortex at initial times.
Technical corrections
Line 7 : missing verb.
Lines 18-22 : Break up in two sentences.
Lines 30-31 : Check incorrect wording.
Lines 37-41 : Break up in two sentences.
Line 48 : « To » or « with » instead of « into ».
Lines 69-70 : Why do you mention the computation date ? I find this quite confusing when talking about hindcasts for a different date in 2018.
Line 72 : I would suggest to change u1060 to something more clear like U10hPa60N
Line 124 : Uncertainty is large also where the ensemble mean flux is anomalously negative.
Line 240 : Repetition of « in this region ».
Cámara, A. d. l., T. Birner, and J. R. Albers, 2019: Are Sudden Stratospheric Warmings Preceded by Anomalous Tropospheric Wave Activity?. J. Climate, 32, 7173–7189, https://doi.org/10.1175/JCLI-D-19-0269.1.
Citation: https://doi.org/10.5194/egusphere-2024-1652-RC1 -
RC2: 'Comment on egusphere-2024-1652', Anonymous Referee #2, 30 Jun 2024
reply
review of "Tropospheric Links to Uncertainty in Stratospheric Subseasonal Predictions" by Wu et al
This paper demonstrates that uncertainty in stratospheric S2S prediction originates in uncertainty in the troposphere. Using a large set of hindcasts from the IFS, they find that weak vortex cases tend to have more certainty as compared to strong vortex cases in the first few weeks. Also, large uncertainty cases tend to have larger mean values of 100hPa heat flux. They then perform a detailed investigation of the 2018 SSW, and show that intermodel spread in its prediction is specifically associated with intermodel spread in wave-2 in the troposphere.
The paper is already in good shape, however I think the arguments could be made stronger. Please see my suggestions below
general comments:
1. Is it possible to produce Figure 2 for a tropospheric level? The authors are connecting v'T' at 100hPa with the underlying tropospheric stationary waves. While this is, to first order, an ok assumption to make, the connection would be more immediate if the level shown was lower down. This is because the tropospheric waves can sometimes have zonal propagation and strong phase tilt near the tropopause, and hence the connection with the longitude of the underlying wave source can sometimes be lost.
2. The authors define the high uncertainty composite and low uncertainty composite by averaging over the entire 46 days of the hindcasts. I wonder if the message regarding the direction of propagation of uncertainty would be clearer if the averaging was performed only over certain days. Specifically, if the averaging was over the second half only, the upward propagation of uncertainty could be better constrained by focusing on tropospheric and lower stratospheric conditions before. Similarly, if the averaging was over the first half only, the downward propagation of uncertainty could be better constrained by examining tropospheric conditions afterwards. The reason I suggest this is that the current Figure 3 and its accompanying text can be hard to interpret, as it isn't clear the direction in which uncertainty is propagating. (It could be that there is substantial overlap as to which hindcasts have uncertainty in the first vs. second half, in which case this suggestion won't be very helpful)minor comments:
section 5: Cho et al 2023 perform a very similar exercise but with a focus on the 2021 SSW. They find that for this event, transmission in the lowermost stratosphere is crucial. It would be helpful if the authors could try to diagnose the effect Cho et al found, so as to explore its relevance for the 2018 SSW. Even if this is not practical, this paper should be cited near e.g., line 37 and 41 among other places, and also included in the discussion. Furthermore, there are two recent papers that should also be discussed in the discussion: Spaeth et al 2024 and Rupp et al 2023. These papers are somewhat less directly related to this paper than Cho et al., but certainly the discussion section should place the present results in the context of this work
line 48: include Garfinkel and Schwartz 2017 here too.figure 1 caption: I don't see any indication of the ensemble spread for the 2018 case in the panels. Please remove this from the caption, or otherwise add it to the figures.
line 185, 226/227: are the convection anomalies this similar to the MJO event that was happening at the time? That is, are the events which capture the SSW also those which better simulate the MJO? See Garfinkel and Schwartz 2017 who argue that models which better capture the MJO related convection also better capture the SSW, but of course couldn't consider this particular event.
line 187, 223/224: I don't see much evidence for a PNA pattern. Rather, there is a low in the far western Pacific and a ridge over Western North America. This is in-phase with wave-2, and hence constructively interferes with it. See e.g., Garfinkel et al 2010 (already cited) and Cohen and Jones 2011. Specifically for wave-2 the more relevant pattern is a low in Eastern Siberia and ridge over Alaska, and not a PNA per se.
line 192: "the North Pacific" -> "Alaska, but reduced SLP over Eastern Siberia,"
line 203-204 I found this sentence confusing. Please add some commas or rewriteRupp, P., Spaeth, J., Garny, H., and Birner, T.: Enhanced polar vortex predictability following sudden stratospheric warming events, Geophys. Res. Lett., 50, e2023GL104057, https://doi.org/10.1029/2023GL104057, 2023.
Spaeth, J., P. Rupp, H. Garny, T. Birner: Stratospheric impact on subseasonal forecast uncertainty in the Northern extratropics, Commun. Earth Environ., 5, 126, https://doi.org/10.1038/s43247-024-01292-z, 2024
Cho, H.-O., Kang, M.-J., & Son, S.-W. (2023). The predictability of the 2021 SSW event controlled by the zonal-mean state in the upper troposphere and lower stratosphere. Journal of Geophysical Research: Atmospheres, 128, e2023JD039559. https://doi.org/10.1029/2023JD039559Garfinkel, C. I., and C. Schwartz. "MJO‐related tropical convection anomalies lead to more accurate stratospheric vortex variability in subseasonal forecast models." Geophysical research letters 44, no. 19 (2017): 10-054.
Citation: https://doi.org/10.5194/egusphere-2024-1652-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 111 | 41 | 13 | 165 | 8 | 8 |
- HTML: 111
- PDF: 41
- XML: 13
- Total: 165
- BibTeX: 8
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1