November 3, 2008

Natural or Anthropogenic Effects on Atlantic Hurricanes, Past, Present, Future?

Filed under: Climate Extremes, Hurricanes

We have often discussed the observed patterns of Atlantic tropical cyclone activity and what may lie behind them, and we generally have concluded, based upon both our analysis of the data, along with a thorough review of the scientific literature, that identifying a statistically significant and robust human signal in the observed history of Atlantic basin tropical cyclones, whether over the past 100+ years, or in recent decades, is untenable.

We have largely come to this conclusion as the observed increases in hurricane activity in recent decades far exceeds that generally projected by climate models run with observed changes in anthropogenic emissions, and there is ample (and growing) evidence that the Atlantic hurricane record is characterized by multi-decadal oscillations that are tied to multi-decadal oscillations in ocean circulation, atmospheric circulations, and patterns of sea surface temperature variability. That these multi-decadal oscillations can be traced backward in time for at least several centuries, is strong indication that they are a natural part of the earth’s climate system, rather than being primarily driven by human alterations of the earth’s atmosphere.

This conclusion has important implications for the future, as it suggests that as the sign and strength of the natural cycles controlling hurricane behavior wax and wane, so to will the future activity of Atlantic tropical cyclones, both in frequency and intensity. The contrary conclusion—that anthropogenic “global warming” is largely controlling the activity of Atlantic tropical cyclone activity—portends, conversely, an ever stormier future.

While we have tried to present clear evidence that the scientific tide seems to be turning in the direction of a predominately “natural” origin of past, present, and future, Atlantic tropical cyclone variability, there are still many prominent groups, including the U.S. Climate Change Science Program (CCSP) and the U.S. Environmental Protection Agency (EPA) that choose to rely on out-dated findings to support their claims of a significant anthropogenic impact on current and future Atlantic hurricane activity in their current draft versions of climate change summary documents. As public reviewers of these documents, we have continually stressed that their conclusions are ill-founded and out-of-date and must be amended and modified to reflect the current state of scientific knowledge on this topic. We hope that they will choose to do so in when the final versions of these documents are released.

As further support to our contentions concerning the underlying influences on Atlantic tropical cyclone behavior, hurricane researchers Gabriel Vecchi, Kyle Swanson, and Brian Soden published a ‘Perspectives’ piece in this week’s Science magazine which summarizes their view of the subject.

They lay out the arguments for each case:

Anthropogenic case:

There is a strong correlation between sea surface temperatures (SSTs) in the tropical Atlantic Ocean and Atlantic tropical cyclone activity. And, in recent decades, as the global temperatures have risen (presumably from human activities) so too have the SSTs in the tropical Atlantic which has promoted an increase in the frequency and intensity of Atlantic hurricanes. As climate models run with increasing levels of atmospheric greenhouse gases indicate Atlantic SSTs will increase in the future, so too will Atlantic tropical cyclone activity.

Natural case:

There is a strong correlation between the SST changes in the tropical Atlantic Ocean relative to tropical SSTs in other ocean basins and Atlantic tropical cyclone activity. In recent decades, the tropical Atlantic Ocean has warmed faster than other tropical oceans and thus, Atlantic tropical hurricane activity has picked up, both in frequency and intensity. As climate models run with increasing atmospheric concentrations of greenhouse gases do not project that the tropical Atlantic will warm faster than other tropical oceans, future tropical cyclone in the Atlantic will be driven by natural fluctuations in the patterns of tropical SST increases rather than simply an overall SST increase.

Vecchi et al. (2008) suggest that empirical evidence is insufficient at the current time to draw a distinction between the two scenario.

However, if you were to turn to purely physical arguments or to the latest state-of-the-science dynamical calculations from high temporal and spatial resolution modeling efforts, you would begin to gather enough weight to start to tip the scale in the direction of natural cycles. Vecchi et al. (2008) lay out these lines of evidence and summarize their conclusions in Figure 1.

Figure1. Observed tropical cyclone activity in the Atlantic basin, 1946-2007 (black lines) and the fit to absolute tropical Atlantic SST (thick brown line, top) and relative tropical Atlantic SST (thick light blue line, bottom). Climate model projections to the year 2100 based upon the observed tropical cyclone/absolute SST relationship (orange lines, top) and observed tropical cyclone/relative SST relationship (blue lines, bottom). The projections made by high resolution dynamic hurricane models are indicated by the green symbols on the right of each chart (see Vecchi et al., 2008 for additional details).

The top chart in Figure 1 shows a cumulative measure of annual Atlantic tropical cyclone activity (thick black line), a statistical fit to the observed activity using absolute tropical Atlantic SSTs (thick brown line) and the climate model projections of the future Atlantic tropical cyclone activity based upon that statistical fit (thin orange line are individual model projections, the thick orange line is the model average). Cleary, under this scenario, Atlantic hurricane activity is projected to increase dramatically in the future driven by anthropogenic global warming.

The bottom chart of Figure 1 shows the results of the scenario in which Atlantic tropical cyclone activity (thick black line) is driven by relative changes in the tropical Atlantic SSTs (thick light blue line). Climate model projections of this relationship are indicated by the thin dark blue lines and the thick blue line model average. In this scenario, global warming has little impact on Atlantic tropical cyclone activity.

The current “best thinking” as to the impact of global warming on Atlantic tropical cyclone activity from high resolution dynamical hurricane models is indicated by the elements in green (stars, squares, triangles, bars) at the far right-hand side of each chart. In each case, the high-resolution model results fall within the spaghetti of the model projections depicted in the bottom chart and not within the spaghetti of the top chart. This implies that our best hurricane models are lending their support to side maintaining that there is little impact from global warming, and instead, tropical cyclones are largely modulated by natural variability.

Granted, there is still a lot of work that needs to be done in the arena of hurricane modeling before this issue can be cleared up, which is the primary message that Vecchi et al (2008) want you to take home with you, but, along the way, Vecchi et al. (2008) strongly demonstrate that based upon what we now know, it seems that natural multi-decadal oscillations in the climate of the Atlantic Ocean trump anthropogenic global warming, when it comes to being the dominant driver of 20th and 21st century Atlantic hurricane activity.


Vecchi, G. A., et al., 2008. Whither Hurricane Activity? Science, 322, 687-689.

No Comments

No comments yet.

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Powered by WordPress