1. You state in your abstract "Whether the characteristics of tropical cyclones have changed or will change in a warming climate — and if so, how — has been the subject of considerable investigation, often with conflicting results."
Can you describe some of the conflicting results? What is the range of projections?
John McBride: All described succinctly in our Nature Geoscience paper
Tom Knutson: An example of conflicting results is in Fig. 1, where depending on which statistical model one accepts, one can either conclude that there may already be a detectable anthropogenic signal of increasing Atlantic PDI, and that the projected change for the 21st century is of order +300% (top), or that there is not a detectable long-term increase signal in PDI as yet, and even the sign of the future change is quite uncertain (bottom). Dynamical models to date strongly support the latter interpretation.
Another example is in Figure 2, where depending on what adjustments are done to the data for 'missing storms', one can either conclude that there is a significant long-term rising trend in Atlantic TC counts or not. The range of future projections is well described in Fig 1 for the Atlantic, in our supplemental material for various basins, and summarized in Box 1.
2. You indicate “For future projections of tropical cyclone activity, the challenge is to develop both a reliable projection of changes in the various factors influencing tropical cyclones, both local and remote, and a means of simulating the effect of these climate changes on tropical cyclone metrics, such as storm frequency, intensity and track distribution. “ How reliable are current models with respect to tropical cyclones?
John McBride: Once again described succinctly in the paper..... The purpose of most of the research reported on was basically addressing this question.
Tom Knutson: The latest downscaling and time slice techniques with statistical/ dynanamic or higher resolution dynamical models are looking quite promising for the Atlantic in particular. See Fig. 3. Also see some examples in the recent Bender et al 2010 Science paper for some Atlantic TC intensity metrics, and other papers referenced in Fig. 3. This is an area where the field has been making notable progress since the IPCC AR4.
3. You state in your paper: “For detection and attribution, the emphasis here is on the Atlantic Ocean basin because the data records for this region are longer and relatively more reliable, though our assessment state ments (summarized in Box1) include consideration of all basins as appropriate. “ Given that the emphasis in your paper is on the Atlantic, is it appropriate to extrapolate these findings to other basins, such as the basin affecting Northern Australian tropical systems?
John McBride: No extrapolations were made from Atlantic findings to other basins. All reported finding on other basins are from work where scientists analysed data or carried out model simulations in those basins.
Tom Knutson: For detection and attribution, we conclude that it remains uncertain whether past changes in tropical cyclone activity have exceeded the variability from natural causes. This statement applies to all basins. There is no study we are aware of that provides a convincing case, at this time, of a long-term TC change in any basin that exceeds the variability from natural causes.
4. You state that “a substantial part of the increase in Atlantic power dissipation since 1950 is likely due to factors other than greenhouse-gas-induced warming. “ Can you name those factors?
John McBride: All described in the paper.
Tom Knutson: The partial quote used here is inappropriate. The full sentence is: "If the relationship between Atlantic power dissipation and this differential warming in Fig. 1b is causal, then a substantial part of the
increase in Atlantic power dissipation since 1950 is likely due to factors other that greenhouse-gas-induced warming." So if the first part of the sentence applies, then the Atlantic PDI
increases (or decreases) are causally linked to warming (or cooling) of the tropical Atlantic relative to the rest of the tropics. Since current climate models forced with substantial increases in greenhouse gases (IPCC A1B scenario) show relatively small increases or decreases in this relative SST metric (Fig. 1b), there is no indication at this time of a strong increase in relative SST warming of the Atlantic due to increasing greenhouse gases. Hence we arrive at the full statement. Proposed factors include: data problems, internal multi-decadal climate variability such as the "Atlantic Multidecadal Oscillation", and changes over time in aerosol forcing of climate.
5. You indicate: “Substantial multidecadal SST variability is evident in the North Atlantic basin (Fig. 2, second green series). The cause of this variability remains uncertain, with possible contributions from both internal climate variability and radiative-forcing changes.” If the cause of variability remains uncertain how confident are you that models are able to project tropical cyclone frequency into the future with any confidence?.
John McBride: The answer to this question is discussed and stated very clearly in the concluding sections of the paper, and is much of the point of the paper.
Tom Knutson: Researchers have been using combinations of observations (with attempted corrections for homogeneity issues), models and theory to shed light on this problem. Uncertainties remain, which is why our projections statements at the regional scale have such limited confidence. For example in Box 1 for TC frequency projection we state: "We have very low confidence in projected changes in individual basins." Nonetheless, there are indications in for example the Bender et al 2010 Science paper that despite strong multidecadal variability in past Atlantic cat 4-5 frequency, an anthropogenic signal may emerge from this noise late in the 21st century according to those model projections. This is still a very active area that researchers are continuing to pursue, and we hope that future work will help to reduce the various uncertainties.
6. Changes in observing capacity brought about by satellite monitoring capability suggest many apparent changes in observational frequency are an artefact of observation methodology. Can you expand on comments around this issue you raise in your paper?
John McBride: Once again discussed at length in our paper and in the references therein.
Tom Knutson: See for example, the references: Vecchi and Knutson (2008)
Journal of Climate and the Landsea et al 2010 Journal of Climate paper (latter paper in press and available on line soon).
7. You cite Mann et al., 2009 study of a 1500 year record of sediment outwash from a number of sites along the US east coast that finds several periods of strong Atlantic hurricane landfalls during several periods over the last 1000 years. How do you reconcile potential natural variation with variation potentially induced by anthropogenic forcing? How would we know if future increases are due to anthropogenic factors? Is there a signature that would allow a confident assignation of anthropogenic factors as the cause of any change?
John McBride: Once again, this is the whole point of the paper, and is one of the many reasons for the “remains uncertain” finding concerning whether past changes can be attributed to anthropogenic warming
Tom Knutson: The recent Bender et al. Science paper is a step toward using a model to estimate what the anticipated signal of anthropogenic warming on Atlantic hurricane activity would be. Also discussed in that paper is an example of estimating the "emergence time scale" for such a signal in cat 4-5 hurricane counts from our current estimate of internal climate variability in the basin. This is a tough problem though, and still be actively worked on.
8. You state that in regard to future projections of tropical cyclone frequency that “Many of these models reproduce key aspects of observed part cyclone activity…” What proportion of models used in your review do not?
John McBride:Not a precise question.
Tom Knutson: Recall that this paper is an assessment of previously published work, and not in itself a research project. Not all of the previous cited works address how well their models reproduce various aspects of observed TC variability. The methods and degree of comparison are uneven across different studies. While that is unfortunate, this is the information that the team had to work with at this stage.
This is where some of the expert judgment of the team must come into play, namely how to weigh various findings in various studies when the methods and degree of comparison between models and observations vary widely across different available studies.
9. You state: “However, confidence in these projections remains very low for individual basins (Supplementary Table S1), owing to uncer tainties in the large-scale patterns of future tropical climate change, as evident in the lack of agreement between the model projections of patterns of tropical SST changes as well as remaining limitations in the downscaling strategies.” How do you reconcile this “very low” level of confidence with your assessment of likelihood of future tropical cyclone frequency which you state is “likely that global mean tropical-cyclone-frequency will either decrease or remain essentially unchanged owing to greenhouse warming.” How do you reconcile this level of confidence with your statement “For individual basins, there is much more uncertainty in projections of tropical cyclone frequency, with changes of up to ±50% or more projected by various models.” What does this say about confidence in systems affecting Northern Australia?
John McBride: Happy to talk about this further. It is one of the major findings of the paper, and so appeared in the summary/abstract.--- That confidence is low for individual basins such as Northern Australia.
Tom Knutson: The remarkably consistent result of a projected reduction in global TC frequency across a range of different modeling studies led the group to its conclusion. (Recall that "likely" here means 67% or more chance that the statement is correct.) But the consistency among different models is much less for individual basins. See table S1 in the supplemental material for more information.
10. You state that “Analyses of globally consistent satellite-based intensity estimates since 1981 indicate that trends in the best track-data are indeed inflated42, but do support an increase globally in the intensities of the strongest tropical cyclones.” Given increase in green house gases since the industrial revolution what does this say about the link between green house gases and tropical cyclone intensity?
John McBride: It supports a link between the highest intensities and global warming; but the length of the data record is too small for any firm conclusions to be made from this analysis.
Tom Knutson: Agreed. There are some concerns about data quality, even of the homogenized satellite intensity record, and there are questions about the amount/nature of internal TC climate variability in the various basins, where our confidence is limited due to relatively short high-quality records we have to work with. This is a key point moving forward. By raising awareness at this point, we hope that our report can help spur better climatological record keeping for TCs worldwide so that some decades from now data bases will be better suited to address these important questions.
11. You state “We judge that a sub stantial increase in the frequency of the most intense storms is more likely than not globally, although this may not occur in all tropical regions. Our confidence in this finding is limited, since the model-projected change results from a competition between the influence of increasing storm intensity and decreasing overall storm frequency. How do you reconcile this lack of confidence with use of the word “likely”?
John McBride: It is what we mean by “likely”
Tom Knutson: Careful here: we use the term "more likely than not" which means essentially that the odds are slighty greater than 50% that the statement is correct. This is not a very confident statement.
12. Given “The uncertainty in climate-model-projected SSTs and related variables can affect even the sign of the projected tropical cyclone activity change in a given region.” Is your likelihood assessment little more than a “gut feeling”?
John McBride: No. It is the result of literally hundreds of scientific papers, enormous numbers of scientific experiments with computer simulations, arguments and discussions among the scientific community, extended analyses of the points of agreement and disagreement between scientists. Also, the scientific journals and scientific process does not include “gut feelings”
Tom Knutson: We provide numerous references and the summarized findings in table S1-S3 to support our assessment. However, rather than merely averaging every result in the literature into some sort of ensemble answer, we have used our judgment to try to assess the relative reliability of different available lines of evidence. I view that as a central charge of our team. Without including this expert judgment element, the whole process would be merely an exercise in data collecting and averaging some outputs from models. It is my hope that our team has added some value beyond that with our report.
13. Finally we note that you acknowledge the Sultan of Oman for sponsoring initial meetings for the report. Given the dire implications of climate change promulgated by the IPCC do you consider this is an extravagance and that less carbon dioxide intensive methods of holding meetings might be more appropriate in the future.
John McBride: Pass
Tom Knutson: Some elaboration of the process might be helpful here. Our team has a budget of zero, and any resources used were drawn from existing resources. Half of our team was able to attend the Expert Team's only face-to-face meeting (in Oman) and half were unable to do so, and therefore attended remotely by telecon (getting up in the middle of the night to do so!) The team members who went to Oman were, on the same trip, able to present a series of invited keynote lectures at the First International Indian Ocean Conference on Tropical Cyclones and Climate Change, which John McBride help enormously in organizing. We felt this was an important step in building the science up on this topic in the Indian Ocean region, and that it would help attract new researchers from that region to this field through the conference and interactions there.
(Recall that one of the limitations of the field are the limited historical TC records in various basins around the world, including the Indian Ocean basin.) Since our report was put together mostly by email, with a few telecons, it is an interesting experiment in whether such less carbon intensive approaches can work. My opinion is that our report was hindered some by not having several face to face meetings of the whole team, but that we did save on science budgets and carbon emissions by not doing so.
