the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Intensified Aleutian Low induces weak Pacific Decadal Variability
William James Dow
Christine M. McKenna
Manoj M. Joshi
Adam T. Blaker
Richard Rigby
Amanda C. Maycock
Abstract. The Aleutian Low drives decadal variability in North Pacific sea surface temperatures (SST), but its role in basin-wide Pacific SST variability is less clear owing to the difficulty of disentangling coupled atmosphere-ocean processes. We apply local atmospheric nudging to isolate the effects of an intense winter Aleutian Low using an intermediate complexity climate model. An intensified Aleutian Low produces a basin-wide SST response with a similar pattern to internally-generated Pacific Decadal Oscillation (PDO). The amplitude of the SST response in the North Pacific is comparable to PDO, but in the tropics and southern subtropics the anomalies induced by the intense Aleutian Low are a factor of 3 weaker. The tropical Pacific warming peaks in boreal spring, though anomalies persist year-round. A heat budget analysis shows the northern subtropical Pacific SST response is predominantly driven by anomalous surface heat fluxes in boreal winter, while in the equatorial Pacific the response is mainly due to meridional heat advection in boreal spring. The propagation of anomalies from the extratropics to the tropics can be explained by the seasonal footprinting mechanism, involving the wind-evaporation-SST feedback. The results show that low frequency variability and trends in the Aleutian Low could contribute to basin-wide anomalous Pacific SST, but the magnitude of the effect cannot explain the full amplitude of the PDO. This finding suggests that external forcing of the Aleutian Low is unlikely to explain observed shifts in the phase of PDO in the late 20th and early-21st centuries.
- Preprint
(1716 KB) - Metadata XML
-
Supplement
(319 KB) - BibTeX
- EndNote
William James Dow et al.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-1595', Anonymous Referee #1, 22 Aug 2023
This paper examines the response of the Pacific climate system to forcing associated with Aleutian low variability using an intermediate complexity coupled atmosphere-ocean model (e.g., T42 atmospheric resolution; 2° ocean). The experiment design involves first performing a long control simulation. Nudging was used to insert atmosphere forcing associated with the Aleutian low, which is: i) applied between 30°-65°N 160°E-140°W; ii) begins in November, ramps up to a maximum between December-February and then ramps down in March; iii) is obtained from the average SLP anomaly over the N. Pacific (North Pacific Index) for when the NPI < -3σ in winter months (a very strong low pressure anomaly); iv) is full strength at the surface and decreases exponentially to zero at the tropopause; and v) added to temperature, winds and surface pressure on a daily basis. A 50-member nudged ensemble was performed, where each member is 30 years long and the nudging is applied each winter. The response to the nudged forcing is explored by the ensemble mean difference between the nudged ensemble and the control run.
The results suggest that the influence of the forcing can extend beyond the North Pacific to the subtropics and to a lesser extent to the equator. A regional mixed layer (upper 30 m) heat budget indicates that the temperature changes are mainly though the surface heat flux in the subtropical region, but meridional transport in the equatorial region. The Aleutian low was also identified as the primary driver of the Pacific decadal oscillation (PDO) SST pattern.
Comments
I have a few major concerns about the paper and some additional minor comments
Major comments:
1) A major issue of the results is the low amplitude of SST anomalies in general. The authors note that the Pacific low frequency variability in the control run is a factor of four-five times weaker than observations. In addition, the diffusivity between 5°N-5°S is increased by a factor of 20 to balance upwelling. While the model probably runs rapidly and so a large number of fairly-long simulations can be performed, is it really a good model for addressing the topic they wish to explore in this study? Perhaps statical analysis could be applied to higher resolution GCMS, such as those in the CMIP archive, to investigate the relationships examined here. There are statical methods, e.g., partial correlations, to examine correlation between variables while removing the influence of others.
2) Related to 1) and perhaps of even greater concern, is the very weak response to the forcing even though it’s exceptionally strong (3σ NPI index) and applied over multiple winters. Many of the results are shown as a regression per unit change in the PDO σ (so it’s a little difficult to gauge their actual values) but the values given are generally less than 0.1 K/σ south of 30°N and even smaller on the equator. Indeed lines 187-188 and Fig. 6 indicate that the maximum response in the equatorial band is less than 0.05 K/σ for all months. This is very small even for decadal variations along the equator. While the use of a large number of ensembles may indicate statistical significance in some subtropical locations, many fewer grid points have significant changes south of ~10°N. Even if the changes are significant due to the large sample size, they may not have much practical importance given their small magnitude. If anything, these results suggest that fluctuations in the Aleutian low have little impact on the equatorial Pacific (in contrast to forcing by the North Pacific Oscillation).
3) Given that the forcing is drawn from a coupled control experiment, the Aleutian low variability and the PDO obtained from the control run will include the influence of the tropics (via the atmospheric bridge).
4) While the budget analysis is helpful to understand the processes involved in the anomalies reaching the subtropics and equator, further analysis could help elucidate how the anomalies reach the subtropics and equatorial regions. For example is there propagation of the signal south (and usually westward) associated with WES? Is the wind anomalies consistent with the “trade wind charging hypothesis”?
Chakravorty, S., Perez, R. C., Anderson, B.T., Giese, B., Larson, S., & Pivotti, V., 2020. Testing the trade wind charging mechanism and its influence on ENSO variability. Journal of Climate. doi:10.1175/JCLI-D-19-0727.1
Anderson, B.T., R. Perez, and A. Karspeck, 2013: Triggering of El Niño onset through the trade- wind induced charging of the equatorial Pacific, Geophys. Res. Lett.. DOI:10.1002/grl.50200.
Minor comments:
1) Line 65 Could reference the following paper as well:
Gan, B. L. Wu, F Jia, S. Li, W. Cai, H. Nakamura, M. A. Alexander, and A. J. Miller, 2017: On the response of the Aleutian Low to greenhouse warming. J. Climate, 30, 3907-3925, doi: 10.1175/JCLI-D-15-0789.1
2) Fig. 1:
a) The caption and legend above the figures don’t seem to match.b) The forcing is very different than the control at high latitudes – perhaps not unexpected as the surface temperatures may not be that strongly related to the PDO. It is also somewhat surprising that the negative air temperature anomalies near ~35°N are not of larger amplitude and extend further from Asia into the central/eastern Pacific in the nudging experiment.
3) Lines: 151-152 variability is calculated by multiplying the standard deviation of overlapping 15-year means by √2.
Explain why 15 years (some estimate of decadal variability)? Why multiply by sqrt of 2 (for statistical significance)?
4) Line 156 The mixed layer varies with season and is generally deeper than 30 m especially in winter at higher latitudes, so perhaps it should just be stated as the budget of the upper 30 m rather than over the mixed layer. While it is likely a secondary factor (especially for a fixed depth budget), is penetrating solar radiation out the base of the mixed layer (30 m) considered in equation 5?
5) Line 178, It might be helpful to show the calendar sigma values for the PDO in the Control and compare them to observations. It could be shown in the supplemental.
6) Lines 188-191 Since the response to tropical SST anomalies strongly influences the Aleutian low and the PDO, the basin-wide SST in the Control and included in the nudging would not be solely due to “internally generated coupled variability”.
7) Lines 240-242 and Fig. 4. It looks like fluxes damp the SST anomalies during MAM and JJA over most of the Pacific, as expected from a stochastic model perspective (linear damping of the SST anomalies) and may indicate little dynamic feedback on the atmosphere.
8) The regression of the latent heat flux on the SLP over the North Pacific the (NPI index, Fig. S3) looks very different than the nudged experiment, even north of 30°N in the winter, when it would be expected that they would be fairly similar.
9) Line 250; Fig. 5. Do winds (shading) include the nudged forcing? If so, should the wind forcing be removed to see the response? Otherwise it is primarily showing the forcing over the North Pacific.
It’s also difficult to see a Rossby wave response to the forcing based on the surface winds. Perhaps showing upper level geopotential or stream function would be helpful in this regard.
Note, no significant change occurs in the tropical box except for two points in MAM (and the values are very small).
10) Line 253-255 The zonal bands of wind anomalies that reach the equator in the central Pacific extend from southwestward from Central America not from California.
11) Line 289-293. While the authors note the difference in the timing of the forcing and the type of forcing (fluxes vs. air temperature and winds) between their study and the one by Sun and Okumura (2019), a key difference is that the latter derived the forcing from the NPO as opposed to the Aleutian low. The former has a dipole pattern with the southern lobe being closer to the equator and thus may be more effective at influencing the tropics.
Citation: https://doi.org/10.5194/egusphere-2023-1595-RC1 -
RC2: 'Comment on egusphere-2023-1595', Anonymous Referee #2, 04 Sep 2023
This study investigates the influence of the Aleutian Low on tropical Pacific sea-surface temperatures (SSTs), using wind-nudging experiments in an intermediate complexity climate model. The pathway the authors investigate has been proposed as a key part of the Pacific Decadal Oscillation (PDO) and is critical piece in understanding how tropical Pacific SSTs are influenced by internally generated or externally forced changes in the Aleutian Low. The results of this study are therefore potentially relevant for interpreting historical variability in Pacific SSTs and will be of broad interest within the climate variability community.
While the overall setup of this study is mostly sound, the interpretation of the results is misleading in several substantial ways, which should be addressed before publication:
- One important caveat that needs to be mentioned is that this model uses very low resolution that does not resolve mesoscale processes in the ocean and atmosphere. While this is true of many studies on the PDO, this is important here considering the framing in relation to Klavans et al. (submitted), who propose higher resolution as potentially helping to address the low signal-to-noise ratio of the PDO.
- Another important caveat that needs to be mentioned is that the nudging used is taken from a single winter and therefore may not be representative of Aleutian Low variability in general. It would be possible do more analysis to show that the anomaly used is (or is not) representative of other 3-sigma Aleutian Low anomalies, which could get around needing to include this caveat.
- Some of the framing of the results is about their relevance for decadal variability, e.g., the sentence “Here we find a similar effect on multi-year timescales in response to an anomalous Aleutian Low.” However, the same nudging is imposed every year, so the results could entirely be explained by processes on shorter timescales. Please revisit each instance of ‘decadal’ and ‘PDO’ in the manuscript and reconsider whether these statements are supported considering the annually repeating forcing used. It is of course fine to motivate the study by questions about decadal variability, but considering the model set up, it’s not possible to make strong conclusions about decadal variability without further work. The title and abstract are important to revisit in this regard. In my opinion, the abstract is well written and supported by the study up until “anomalous Pacific SST” on line 29, but after that all the statements about the PDO are only marginally related to the study and can’t really be supported by the results. I also think that the inclusion of “Pacific decadal variability” in the title does not accurately represent what the study is about.
- On the influence of the Aleutian Low not being able to explain the full PDO nor it’s phase shifts in the late 20th and 21st centuries, please consider recent literature showing that the decadal variability of the PDO, including its characteristic phase shifts, is coming almost entirely from the North Pacific (Wills et al. 2018, https://doi.org/10.1002/2017GL076327; Wills et al. 2019, https://doi.org/10.1029/2018GL080716), with the tropical Pacific primarily adding noise on interannual timescales. At the very least, please change “PDO” to “PDO in the tropics” on line 30. However, this also means that the statements on lines 30-31 and 329-332 are not supported by the results without additional analyses.
- More interpretation is needed in Section 3.3 (Figures 5 and 6). What are the take away’s from this analysis?
Minor comments:
- For consistency between Eq. 1 and 2, I would recommend defining functions of z and t in Eq. 2 (as described near the bottom of page 5 / top of page 6).
- Line 152: “multiplying the standard deviation of overlapping 15-year means by √2” requires more explanation of why you are doing this / why this is the relevant measure for statistical significance.
- I think (a) and (b) are switched in the caption of Figure 1
- Line 183: What does it mean to express the anomalies between NUDGED and CONTROL “per standard deviation of the PDO index”? And how does one compare the amplitudes between the two (as discussed on lines 195-196).
- Line 188-189: Considering that Fig. 1a and Fig. 1b have DO NOT look very similar, please also discuss the differences, rather than just saying these patterns closely resemble one another. You could also make this more quantitative with a pattern correlation.
- Figure 2: Why would you expect this to be different depending on the year of the simulation? It doesn’t seem essential to the overall paper to show this for three different averaging periods.
- Figure 4: There is much more that could be said about this figure. The DJF panel shows that heat fluxes from the atmosphere into the ocean are imprinting onto the SST pattern during the nudging period, while the positive heat fluxes in the KOE region in all other seasons show that the cold SST anomalies in this region are reducing the heat loss to the atmosphere.
Citation: https://doi.org/10.5194/egusphere-2023-1595-RC2 -
EC1: 'Comment on egusphere-2023-1595', David Battisti, 11 Sep 2023
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-1595/egusphere-2023-1595-EC1-supplement.pdf
William James Dow et al.
William James Dow et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 172 | 37 | 12 | 221 | 21 | 4 | 3 |
- HTML: 172
- PDF: 37
- XML: 12
- Total: 221
- Supplement: 21
- BibTeX: 4
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1