February 2013 Archives

Cod in the Gulf of Maine

| No Comments
This is a very challenging time to be a cod fisherman in the Gulf of Maine, and presumably, it's also a tough time to be a cod.  Fisheries managers have drastically reduced the amount of cod fishermen will be allowed to catch next year, and this follows a large reduction last year. Like many of us working at the intersection of oceanography and fisheries, I'm interested in what's up with cod?  There are a lot of hypotheses being kicked around, but as you might imagine, I'm interested in whether changes in temperature or other environmental factors are playing a role.  Here's a quick and dirty take on whether recent warming could be impacting cod.

First, temperature.  I began with the AVHRR OI data set that I've been using to characterize the 2012 story.  I grabbed the pixel near NERACOOS Buoy E off of the central Maine coast.  Here is the temperature pattern at that location:
The thin blue line is the daily anomalies (smoothed over 15 day window).  The gray circles are the mean for each year.  The black line is the overall trend.  The trend is significant, but there is considerable year-to-year variability.  A trend of 0.026° per year is pretty consistent with the general global warming trend.  This would project to a 1° increase by 2050 and 2.3° by 2100.  The really striking pattern in the figure are the last three, extremely warm years.  If you fit a line to the 2004-2012 anomalies, you get a trend that is 10 times faster than the long term trend. I'm getting on statistically dubious grounds with this calculation, but it does suggest that we may have entered a new temperature regime.

Most of the published relationships between temperature and cod use some form of bottom temperature.  I used the 2002-2012 temperature data from buoy E to develop a simple statistical model for the temperature at 50m as a function of the surface temperature and the day of the year. The long-term mean temperature at this location was 6.89°C, which is a bit lower than the "standard" 8° value that most people report.  So, I added 1.11° to the annual bottom temperatures to get the mean close to right:
Why the big deal about 8°?  Well, Ken Drinkwater's 2005 paper on cod and climate change suggests that for regions above 8°, additional warming tends to reduce the fitness of cod.  For regions below 8°, warming increases fitness.  Using this benchmark, we had a period from 1999-2002 when cod reproduction may have been reduced by temperature.  The 10°C value for 2012 is getting scarily close to the 12°C point where Drinkwater says "cod stocks are not observed much."  

The red line in the figure is the September-October mean (not adjusted).  I then applied a relationship from Fogarty et al. (2008) relating the probability of catching a cod in the NMFS bottom trawl survey with the September-October bottom temperature. 
Not surprising, this has many of the same patterns as the temperature values.  There is a gradual long-term trend (probability declines by 0.004 per year) and a much faster trend over the last 8 years (0.02 per year).  While the trends are interesting, I'm interested in what appears to be a shift in the mean in 1999.  Before 1999, the values are centered around 0.42.  After 1999, the mean probability appears near 0.37.  By this metric, 2012 looks very, very scary.  The probability of finding a cod this year was about half what it was in the early 1980s

More SST videos

| No Comments
The University of Maine moved to Google-based email, and I just realized that it includes a You Tube account.  Now, I can stop writing proposals and fund my lab with kitten videos!  While I'm shopping for a kitten (or perhaps an infant with a penchant for biting), I've been having fun putting together more videos of North Atlantic SST anomalies. Here are my two latest:

This is 2011 and 2012.  Each frame in the video is the anomalies averaged over 15 days.  I also changed the color bar a bit.  I think both changes help emphasize the larger, longer lasting events that I think are really important. You'll notice I also figured out how to use the Ken Burns effect in iMovie. 

 If you want a longer term-view of the North Atlantic , try this:

The anomalies come fast-and furious in this video.  Notice that there is more blue at the beginning than at the end.  This video also helps put the 2012 anomaly that I'm so excited about into perspective.  There are several events earlier in the video that are almost as big, but none of them last as long.

Some notes on how I made these images:
  1. The data come from NOAA's AVHRR OI product.  All anomalies are referenced to the 1982-2011 climatology
  2. I created anomalies for every day in the data set and then averaged 15 together to form one frame.  This is the only smoothing I did.
  3. I saved each image as a PNG file.  
  4. Here are the two Matlab functions I developed (requires the mapping toolbox): generateNAtlAnomalyImages.mmapNAtlSSTanomaly.m.  The interesting trick I use in "generateNAtlAnomalyImages" to replace the "Cdata" field in the surface object created by a call to the other function, rather than calling the plotting function many times.  This skips all of the work that the mapping toolbox does to figure out the projection.
  5. I then created the movie using the application "Time Lapse" (www.microprojects.ca)
  6. The second movie was uploaded straight to YouTube.
  7. I brought the first movie into iMovie and added the title and credit.  I also added the first and last images at the beginning and end.  Hopefully, this allows the viewer to ease into the movie and gives you a chance to stop the movie before YouTube shows you a list of suggested kitten videos.  All part of the quality user experience we aim to deliver here at SeascapeModeling!

2012 Sea Surface Temperature Patterns

| No Comments
My colleagues know that I am obsessed with 2012.  Every talk I've given in the last six months has featured some element of "2012 was amazingly weird," and I even manage to work it into most dinner conversations at home (probably why the family developed a sudden obsession with watching "Lost" with dinner). My latest 2012 indulgence was to assemble an animation of the daily SST anomalies over the year (minus about two weeks):

There's a lot going on in this animation. The first time you watch, it's hard not to notice the bright blue (colder than average) and red (warmer than average) blobs that swirl around in the the area east of North Carolina and south of the Gulf of Maine and Gulf of St. Lawrence.  These are rings and eddies formed by the Gulf Stream.  Although less swirly than the Gulf Stream, the dominant feature in the animation is the persistent warm area in the northwest Atlantic.  At the beginning of the year, the anomaly is mostly confined the Gulf of Maine region, but by early summer, it has expanded to include the Labrador Sea and even extends to Iceland.  By early September, the majority of the North Atlantic is warmer than usual.

For you data nerds, the images were produced from NCDC's "AVHRR_OI" data.  This product uses an objective interpolation procedure to fill in pixels that were obscured by clouds.  You can download the AVHRR_OI data along with lots of other SST products from the GHRSSThttp://www.nodc.noaa.gov/sog/ghrsst/accessdata.html site.  I computed the climatological mean and standard deviation for each day over the period 1982-2011.  I then subtracted each day's image from the climatological day.  I then tried to damp some of splotchy points, especially in the Gulf Stream area, by multiplying the anomaly by the square-root of the probability that the value is not zero.  Fully saturated blue and red values are -3°C and 3°C anomalies, respectively. The images were plotted using routines from Matlab's "mapping toolbox."  If you email me, I can send you the plotting function.

Dirty Words in Science

| No Comments
Every language has its dirty words.  I feel like a have thorough grasp of how to use the main English ones and have even picked up a few in French, German, and Norwegian.  Like most responsible adults, I think I've learned when not to use them, as well, although when the reviews hit the fan, this becomes difficult.
Science is also kind of its own language, or at least its own culture, and we have our own set if dirty words.  The two that make me cringe like my grandma at a Tarantino movie are "proof" and "belief."

It drives me absolutely batty when my students or NPR talk about how a study "proved" something.  Proving something means to establish with absolute certainty that it is true.  Absolute certainly is possible in math, were you have a small set of operations and a few axioms that are 100% true.  The real world is much, much messier.  Proof doesn't even apply in physics.  Newton's laws seemed like axioms until Einstein came along--there is always some element of doubt or uncertainty, even if it is very, very small.

But, the word that sends me into unprintable (at least on this blog) fits of screaming every time NPR says it is "belief."  Belief actually has a defined meaning in probability--Bayesians talk about belief as an estimate of the likelihood that a statement is true.  Unlike proof, the problem is not so much with the definition in science, but in the way this word is used outside science.

For example, the Pew Foundation conducts a regular survey on attitudes about climate change.  The main question they ask participants is "Do you believe that the earth is getting warmer?"  When a scientist answers "yes," I hope they mean, "based in the available evidence, my conclusion is there is a high likelihood that the earth is warming."  However, for many people, belief is also wrapped up with hope.  

I always think back to the end of the musical Peter Pan (formative experience of my youth), when the audience is asked to clap to show they believe in fairies, to keep poor Tinkerbell from expiring (apparently, fairies, possibly including the "Magic Climate Fairy," metabolize hope instead of carbohydrates).  Belief in this case is not based on any assessment of the likelihood of the existence of hope-consuming winged girls, rather, it is based on our desire for Tinkerbell to exist and survive.  My suspicion is that many people approach climate change in this way: if enough people don't clap to show their belief, perhaps it won't happen, and we can keep fracking.  Precision is an important goal of science, and belief is an inherently imprecise word that should be left at home.

About this Archive

This page is an archive of entries from February 2013 listed from newest to oldest.

November 2012 is the previous archive.

March 2013 is the next archive.

Find recent content on the main index or look in the archives to find all content.