October 07, 2011

Is West Coast Weather Getting More Extreme?

A question that is often asked is whether Northwest or West Coast weather is getting more extreme.  More heavy rain and flooding events?  More windstorms?  Extreme losses of snowpack in the mountains?   Some groups have answered these questions in the affirmative and have stated, suggested or hinted that such extremes are connected with anthropogenic global warming.  They often provide hand-waving explanations of such extreme behavior--for example, it is accepted that the atmosphere will gain water vapor as the planet warms (by roughly 7% per degree C).  Won't that cause extreme precipitation to increase?


My students and I have looked into this issue, as have other weather/climate scientists.  As I shall discuss, this is not a simple issue and uncertainties abound, but the bottom line is that there is little evidence for area-wide increases in extreme weather due to human-induced global warming during the past several decades over our region.  That is not to say there won't be any in the future, but the impacts are anything but clear.

Lets take extreme precipitation. With UW graduate students Mike Warner and Adam Skalenakis, we looked at the trends of heavy precipitation over the past 60 years at coastal stations from southern CA to British Columbia. Specifically we looked at the trends for the top 20, 40, and 60 events of two-day precipitation.  Some of the results are shown in the figure below. Southern/central CA is a mixed bag, but northern CA and southern Oregon show a DECLINE in heavy events. On the other hand, northern Oregon and WA have seen an increase in the big rainstorms.  BC show weak upward trend.


Now heavy precipitation, of course, has a big influence on rivers...what do big river discharge events look like?  Well, we looked at the trends over the same period of maximum annual discharge of unregulated rivers (see figure below).  Importantly, we got the same pattern! Decline over southern Oregon and northern CA, increase to the north.
Up arrows indicate increasing trend, down arrows, declining trend for max annual river discharge.
So there is no uniform change in extreme precipitation along the West Coast:  some places are up and some are down.  Some nice work by members of the Climate Impact Group at the UW, looking at the output of a variety of global climate models, suggests that this pattern is the result of natural variability and not the result of any kind of global warming signal.
    What about big windstorms?  Storms such as the Inauguration Day Storm and Chanukah Eve storm?   Fascinatingly, the pattern is the same over the past half-century--- upward trend over Washington and northern Oregon and decreasing frequency of such storms to the south.  We are trying to understand exactly what is causing this configuration.
The Chanukah Eve Storm of 2006
   What about the next fifty years or so?  Will global warming change extreme weather events over our region?   The honest answer---we don't know.   The computer models do not agree.   And one can easily think of scenarios where there will be a decline of extreme precipitation and storms.   Extreme weather events are closely associated with the jet stream.  Most climate models indicate the jet stream will move northward under global warming.   Will the most extreme events gets displaced northward?

   In short, the changes in extreme weather along the West Coast has been non-uniform and don't suggest a human-induced global warming signal.  Global warming due to our greenhouse gas emissions will be very significant, but the local implications regarding extreme weather are quite uncertain.

22 comments:

  1. How are top 20/40/60 events defined and what is being measured in the comparison between these top events?

    ReplyDelete
  2. I think this post shows some important differences between GLOBAL warming vs. local, and climate vs. weather. One of the biggest hang-ups I get when talking to "non-believers" about global warming is that they see all cold weather here in the Pac NW and draw the conclusion that it must not really be getting warmer.

    Its hard to get it through to them that no matter how we see things on a regional level it doesn't change how things are changing on a global level.

    ReplyDelete
  3. ecuadork...two-day precipitation. So these are the top 20, 40, and 60 two-day events over a 60-year period

    ReplyDelete
  4. As someone who grew up in the plains, and dealt with extreme weather fairly regularly, I would welcome anything "extreme" here in the PNW. Often what we consider extreme here (with the exception of one or two windstorms every 365 days)is hardly a variation on the otherwise mundane 'rails' in which we normally glide. bring on some snow this year!

    ReplyDelete
  5. Cliff - Thanks for being a voice of reason in the global warming debate and using actual data and rational analysis.

    Are you saying that Al Gore is reaching too far with his latest statements?: http://www.guardian.co.uk/environment/2011/sep/28/al-gore-proof-climate-change

    ReplyDelete
  6. A thought good sir. What if the same weather scaring that scours Africa's midsection is happening here in the US?

    ReplyDelete
  7. I guess the real challenge for our species..we build and build use climate for a large basis underlying regulations..we can only walk the plank so long before we have moved into an ongoing consequence analysis challenge.

    ReplyDelete
  8. So, you looked at the data and determined that big storms have increased in the north, but storms became less frequent in the south.

    This is what has been predicted by those "arm-waving" warmists: the south dessicates while the north gets soaked during its rainy season.

    You say it is "natural variability" based on what, precisely? Is that the conclusion of actual climate scientists, or is it more arm-waving by someone looking for excuses to deny the problem?

    ReplyDelete
  9. Ruthless -

    The study Prof Mass decsribes concerns extreme precipitation events. Overall, there has been no significant change in southern CA precipitation in the last century.

    But thanks for the arm waving!

    ReplyDelete
  10. Thank you for the interesting post, Professor Mass.

    I am going to wave my arms a bit, though, and make a bold prediction. I'm going to predict the Pacific Northwest is going to see a lot of rain over the next 6-8 months.

    ReplyDelete
  11. Bob:

    How does that establish that the CA result (and more generally the western coast result is from natural variation?

    Climate scientists have predicted that the northern US wets coast should get wetter while the southern part gets drier.

    So, combining the two, of course the north will tend to balance out the south (or in this case OR and N. CA).

    One of the reasons is the expansion of the subtropics and the northward shift of storm tracks, together with storms in the storm track becoming more extreme.

    What I am getting at is, did the authors of the study conclude it was natural variability, or is someone interpreting the study as having said that.

    Too often, people who intensely do not want to believe in global warming read studies and convince themselves they say things they do not.

    Secondly, even if one study concludes something, that does not make it true. You have to see how it withstands scrutiny over time.

    ReplyDelete
  12. It would be instructive to review some work that has already been done on extreme precipitation and then decide how likely it is that we will face increases in extreme precipitation, or whether we already have seen such increases.

    Some rivers in western WA have had *four* 100+ year floods in the past two decades, for example. When you actually plot the floods, it is pretty obvious they are getting worse and more frequent.

    This is precisely what has been predicted by climate scientists as a consequence of global warming.

    So, for someone to come along and say, "oh, that's just a coincidence; its 'natural variation'" I think is pretty weak. They'd better have a really good reason why a 40% increase in CO2 is not doing what it was predicted to do.

    In the mean time, Google shcolar is a good resource. A search for "extreme precipitation": http://scholar.google.com/scholar?hl=en&q=extreme+precipitation&btnG=Search&as_sdt=0%2C48&as_ylo=&as_vis=0

    And the work of Kevin Trenberth at NCAR is useful too. For example:
    http://portal.iri.columbia.edu/~alesall/ouagaCILSS/articles/trenberth_bams2003.pdf

    ReplyDelete
  13. Ruthless,
    Modeling studies by scientists at the UW Climate Impact Group, among others, have found that there is no reason to think that the pattern we have seen...heavier precip over WA, less extreme precip from central Oregon south, has anything to do with global warming, human caused or otherwise. The variability is not consistent with a global warming signal. If you read the global warming literature you will find a number suggesting heavier precipitation over CA with increased moisture content in atmospheric rivers.

    ReplyDelete
  14. Cliff:

    What was the study where they determined that our increase in extreme weather was only due to natural variation?

    Is it a model study?

    And beside all of that, with all due respect, the UW team is maybe not who one should look to for climate advice.

    What I am getting at is that models chronically underestimate effects which are observed to be worse.

    If models predict something, and then you observe it to be worse, that does not mean the observations were caused by something else.

    It just as plausibly could mean that models are missing something.

    The whole idea of attribution is exceedingly hard, because the models are not that good. Read the work of Hansen's team as one example.

    You end up with an "absence of evidence is not evidence of absence" problem.

    So, if you are using model results, which are known to underestimate warming effects, as a justification that we are not yet seeing the effects of global warming, your claim would be very poorly supported.

    Given the seriousness of the problem, I would even say it is an irresponsible conclusion to make.

    ReplyDelete
  15. In terms of mountain rains washing out roads, that trend certainly seems to be accelerating. I suppose part of it could be just the decay of cheaply built logging roads, but it sure seems like the term "100 year flood" has lost whatever meaning it might once have had.

    ReplyDelete
  16. Snoqualman,
    Heavier events ARE increasing around her lately...but not south of us. That is the key point. And keep in mind that you can get two 100-year floods back to back without any long-term trend.

    ReplyDelete
  17. The Chehalis River in Lewis County has had four floods of at least 100-year magnitude in the last two decades. One of them was a 500-year flood. You can download historic flood crests from USGS. The trend is pretty obvious as soon as you plot the data, same with one-day rainfall.

    I don't know what UW study all of this comes from, but their web talks about one from 2009 in which they ran a single climate model exactly TWO TIMES (!) to arrive at the conclusion that the range of model outputs was too wide to be useful for engineering design criteria.

    Not exactly a heroic effort.

    The thing about rains easing up hundreds of miles to the south proves what? That drying associated with the subtropics is expanding northward? We knew that:
    http://agwobserver.wordpress.com/2010/07/27/papers-on-hadley-cell-expansion/

    ReplyDelete
  18. The world is warming via the same mechanisms it warmed after the last ice age, namely increases in greenhouse gases. That produced the deserts in So Cal and the rain in the Northwest. Why wouldn’t you expect expansion of those effects as the atmosphere increases still further in greenhouse gases?
    Tom

    ReplyDelete
  19. The hype about 4 "100 year" events misses the point about statistics. It's like say that if you roll 2 12s on a pair die in a row that the probability must of changed and it's not a 1/36 event. People forget that 1 in 100 year event doesn't mean it will litterally happen once ever 100 years and never more commonly.

    Natural variability means just that. Use statistical analysis of past events to see if the current trend is some sort of anomoly or just one of those unusual quirks of probability that can naturally occur.
    People forget that there's many overlapping cycles that seem to take place across weather, some spanning decades. I mean the growth of Mt. Rainier ice could be misinterpretted if not compared to known patterns. This is where the rigorous analysis of science helps us discern a real change in trend from just statistical "noise".

    To go with this thread, Cliff, Why does weather data seem to rely on averages so much. Knowing 2 regions have similar average temperatures tells me little about the actual likelihood I'll be sweating or shivering. Similarly, the average monthly rain charrts shown on this blog always have a smooth line as we'd expect when averaging this sort of data. It's hard to see if it's really that unique to have 10 dry days at the start without some sort of variability measure.

    ReplyDelete
  20. Th UW group determined it was natural variability how? Was it a formal attribution study? Using models that are known to underestimate increases in rainfall (1-3%/K modeled vs 7%/K observed)?

    I'd like to see the paper where they concluded this.

    Has anyone actually calculated the probability of having three 100-year events and one 500-year event in 20 years? It's vanishingly small.

    To say the sothwest US coast's climate cancels out the northwest's is like looking at length-of-day measurements in the northern hemisphere and in the southern, and then concluding that Earth has no global seasons. Mathematically true but also virtually meaningless.

    ReplyDelete
  21. Cliff,

    To change the topic slightly, is it true that we have fewer "neutral years" and more EL Nino and La Nina years than we did thirty years ago? Witness this third La Nina since the summer of 2009! I remember hearing way back in the 70's that El Nino averaged once in about seven years. Perhaps La Nina was less understood? Or are there more of them?

    ReplyDelete
  22. Given the seriousness of the problem, I would even say it is an irresponsible conclusion to make.

    What problem? This is like saying "it's not about the truth, but about the seriousness of the charge" stuff with Rathergate.

    ReplyDelete

Please make sure your comments are civil. Name calling and personal attacks are not appropriate.