Seascape Projects

Invasion of the jellyfish!

By Andy Pershing on July 16, 2014 7:28 PM | No Comments

It seems like I'm the only one along the coast of Maine not to have seen a jellyfish this year. That's probably more a comment on my lack of contact with the ocean because, by all reports, there are tons of jellyfish along the coast of Maine this year.

Everyone on the coast is talking about jellies, and it seems that Nick and I are what pass for jellyfish experts. Perhaps someday we'll get a massive Calanus outbreak, but until then, it's really fun to have people talking about the ocean. Although we don't know a whole lot of what a normal jellyfish year looks like, it's pretty clear that this year is unusual. I think it's noteworthy that this summer is warm and that the reports started coming in when we had a big jump in temperature in early June. The other year with lots of jellyfish chatter was 2012. Still, lots of work is needed to really put this story together.

If you need your Nick and Andy fix, check us out in the Portland Press Herland and on the radio at MPBN. Until I get to the ocean to get some real underwater jellyfish pics, here's one of me dressed as a jellyfish:

photo by Petri Touhimaa, GMRI

Crash course in ABC

By Hillary Scannell on July 5, 2014 12:13 PM | No Comments

On July 3rd and 4th, the Australian Statistical Conference put on a free satellite workshop at the University of New South Wales, which introduced Approximate Bayesian Computation (ABC) to anyone interested. Having some familiarity with Bayesian statistics and a strong curiosity for ABC, I decided it was worth my 4th of July to attend.

The first day of the workshop provided a crash course in ABC methods, while the second day consisted of lectures on ABC applications to research. It turns out that approximate Bayesian analysis is useful when the likelihood function is computationally intractable and when likelihood-based inference models are unavailable. ABC is a non-parametric Bayesian method where pseudo data is generated by a candidate

\theta

parameter. The summary statistics from the two data sets are compared and if they are similar, then

\theta

is acceptable for the real data. We use this procedure to estimate the posterior distribution of the parameters of a model. This gets more complicated with dimensionality, but I'll leave that to the Bayesian experts for now. Wikipedia has a decent explanation on this for those who are interested.

ABC is a fairly new statistical frontier that dates back to only the 1980s. It started as a niche idea in population genetics and is now gaining momentum in mainstream statistics. Although ABC still feels a bit weird and awkward to me, my gut intuition is that these methods will become useful when modeling other natural systems too.

G'Day from Sydney

By Hillary Scannell on June 7, 2014 11:36 PM | No Comments

I'm reporting to you live from the middle of the equatorial Pacific, well sort of. Just a few hours ago I was sitting on a plane and realized that I was indeed flying above what might be a developing El Niño. You may have noticed El Niño receiving a lot of attention in the news lately. Scientists seem to be keyed in on the equatorial Pacific, and perhaps for a good reason. Sea surface temperatures appear to be warming in a classical El Niño-like fashion. The last significant El Niño occurred in 1997/98 and it's speculated that we're long overdue for another "big one."

You may also be wondering why I'm flying over the middle of the Pacific. My travels are owed to a NSF East-Asia and Pacific Summer Institutes fellowship, which takes me to Australia for 8-weeks to work alongside climatologists and oceanographers at the University of New South Wales, Climate Change Research Centre. From kangaroos to the animals trying to kill me, I will attempt to keep you updated with my latest discoveries in the land down under.

Cheers for now-

Hillary

I'd like to thank the academy...

By Andy Pershing on May 20, 2014 2:01 AM | No Comments

I'm in Portland, Jr. this week at the Joint Aquatic Sciences Meeting. I'm here to accept the Yentsch-Schindler Early Career Award from ASLO, the Association for the Sciences of Limnologists and Oceanographers. It's an amazing to receive this award, especially because my nomination was led by the intrepid and always well-coifed Fred Maps and my advisor Chuck Greene. Unfortunately the salinity of this meeting is so low that none of my close colleagues (including Fred & Chuck) are here. In lieu of that, I hope you'll indulge me (and I hope you'll check out Hillary's post, which I just bumped, on her recent cruise) in a quick round of acknowledgements:

First, I wouldn't be where I am without the support of my wonderful family. I can't believe that you put up with my increasingly whacky travel schedule and my even whackier notions that rants about climate change or reviews of the latest Matlab release are appropriate dinner conversation. You are an endless source of inspiration and support, and I love you all very much.

Second, thank you to the institutions that have supported me throughout my career. Cornell, UMaine, and GMRI has each influenced me in their own unique and sometimes nonlinear ways. Although I didn't include their logos, I'm extremely grateful to NSF, NOAA, NASA, and the Lenfest Ocean Program for funding my work. Please don't stop!

Finally, I have been very lucky to have always found colleagues who inspire and challenge me. Chuck and Bruce at Cornell helped me figure out that I could really do this. Then, Dan, Pete, Nick, and Fred (aka the Ecosystem Modeling Lab and Fraternity) helped establish the tone for the lab: geeky and creative, with a touch of irreverence. Walt, Sigrid, Kathy, and Carrie, and more recently Dom, Katie, and Elise brought a fishier vibe. Kathy and Hillary are now bringing in real work on climate change, and Karen is continuing the tradition of irreverent copepodology. Science is a tough business, with lots of ups and downs. Since we clearly don't do this for the money, I'm thankful to get to work with people who are fun and creative and who are passionate about this very weird and wonderful endeavor.

Making observations at sea

By Hillary Scannell on May 19, 2014 11:48 PM | No Comments

I've been an EML'er for exactly a year now, starting with my meager beginnings as an intern last May. This past year has been a fast pace transition. After graduating from the University of Maine in May of 2013 with a bachelor's in marine science, I essentially began working towards a master's in oceanography a week later. From that point on, I quickly evolved into a climate modeler and statistics geek. Now that the academic year is over, I can reflect back on my experiences as a first year graduate student and look forward to what the future may hold.

One such experience occurred during a recent (May 1-12) research cruise aboard the R/V Knorr. This cruise sailed between Woods Hole and Bermuda following a direct, linear path called Line W. The Line W field program is a long-term climate observing system put in place by the Woods Hole Oceanographic Institution (WHOI) in 2003. It consists of a moored array of 6 buoys extending over the shelf break. The goal of the program is to gather hydrographic data to better understand the baseline conditions of the Deep Western Boundary Current and Gulf Stream.

Line W is named after the late physical oceanographer, Valentine Worthington, who devoted much of his life and career to understanding the circulation and water mass formation in the North Atlantic Ocean. Like most great things in science, the Line W program recently came to an end due to unobtainable funds. I was fortunate enough to have been a part of its legacy by participating on the final leg of the program.

On this cruise, we took shipboard hydrography measurements at 26 stations along Line W. We lowered a CTD rosette to depths beyond 5,000 meters to measure temperature, salinity and oxygen. An upward and downward facing acoustic Doppler current profiler (ADCP) was attached to the CTD to measure the water current velocity. From the CTD control station within the main lab aboard the Knorr, we fired Niskin bottles at different depths to collect 10-liters of water. A trip down to 5,000 meters can take over 3 hours, so a lot of time was spent waiting in the lab. The CTD and ADCP measurements allowed us to obtain vertical profiles of the Deep Western Boundary Current and Gulf Stream along the location where they intersect Line W. Changes in the global thermohaline circulation are expected to occur due to climate changes at high latitudes, and the past successes of the Line W program have given scientists the ability to document these changes through observational studies.

Yes, observations! During my experience on the Line W cruise, I couldn't help but fixate on the apparent divide between climate modeling and observational oceanography. Models enable us to force a condition, test our theories and make predictions, but perhaps every modeler should connect their virtual reality with hydrographic observations at sea. For me, this is exactly what Line W did.

As I look forward to the future, I am reassured as a modeler that the oceans continue to be explored, measured and observed through long-term field programs like Line W.

On the Line W cruise, I not only gained a greater appreciation for the data I often take for granted, but I also gained a new admiration for the labor intensive field of observational oceanography.

Hillary Scannell, UMaine/GMRI

Photos provided by WHOI.

http://www.whoi.edu/science/PO/linew/

2014 Lobster Forecast--Update 6

By Andy Pershing on May 2, 2014 6:40 PM | No Comments

I think this will be my last update for awhile. While the skill of our forecast continues to increase through May, the value of the forecast decreases as well. I might check back on things at the end of the month, but we'll see.

Our final forecast has the start of the season exactly at the long-term average of 6/29:

There is a slightly greater chance that the start date will be earlier than 6/29, but it's pretty small.

While I'm pretty excited about this forecast, there is one thing that I find unsatisfying. Because we're defining the start date as the percent of the total landings for the year, we can't check the accuracy of the forecast until the year is complete. I have some ideas for how we might work around this problem (defining the start as the rate of change in landings, for example). Here are some other features we hope to add (assuming we can find some funding, of course):

move beyond landings to forecasts for hard/soft shell mix
forecasts for different lobster zones
improved lead time--we think we might be able to start issuing forecasts in November

Stay tuned!

2014 Lobster Forecast--Update 5

By Andy Pershing on April 27, 2014 7:23 AM | No Comments

The temperatures at 20m were slightly below average for a few days. This has caused our forecasted start of the high-catch period to shift slightly later, but we are still projecting a more-or-less normal year:

Note that we have changed the labeling on the left-hand axis slightly. The forecasts are based on the mean temperature over an eight day period. Originally, we were using the middle of this period as the "date when the forecast was made." This is obviously not true. The labels now indicate the last day of data used to make the forecast, as implied by the axis label.

Although the average temperatures over the last 8 days have been slightly cool, the last two days have been a bit warmer. When we update the forecast again (likely on Thursday), I would expect the projected start date to shift to the left by a day or so.

2014 Lobster Forecast--Update 4

By Andy Pershing on April 23, 2014 2:59 AM | No Comments

I'm on vacation this week, so this post will be very brief. Here's the latest update of our lobster forecast:

The 20m temperatures at Buoy E have stuck pretty close to the average over the last few days, and our forecasted start-date is pretty much where it was last week: a day or so later than the long-term mean date. I will update again this weekend and once more next week.

2014 Lobster Forecast--Update 3

By Andy Pershing on April 17, 2014 8:12 AM | No Comments

School vacation starts tomorrow in Maine, and I'll be taking some much needed vacation time. I will try to update the forecast at least once while I'm gone, but I wanted to make sure to get a fresh forecast up before I leave:

The forecast again has shifted to the left, and we're now essentially predicting a normal season. The best analog is 2008.

I'm intrigued by how the season is progressing in the water. My sense is that the ocean is warming up faster than on shore, especially below the surface. I think this might be a result of the series of clear, bright days we've had recently. The most important component to heating in the ocean is the penetration of light into the water. On a clear day, especially if the winds are low, there are a lot of photons going into the ocean. The temperature of the air above the water is a lesser factor, and can really just pull heat from the surface of the ocean. The forecast for the next few days is for sunny and calm conditions. My bet is that by the middle of next week, our forecast will be for an slightly early start.

2014 Lobster Forecast--Update 2

By Andy Pershing on April 15, 2014 2:57 PM | No Comments

The warming at Buoy E has continued and the temperatures there are now pretty much at the long-term average. This means that our forecast is for a season that is only delayed two days (now projected for July 1):

As before, the start dates from previous years are plotted at the top, with blue indicating cold years and red indicating warm years. The diamonds represent a forecast. The center of the diamond is the forecasted value and the width is the 95% confidence interval. Each diamond represents a forecast made using data centered on the date on the left.

Note that the width of the diamond is smaller than before. This is due to both higher skill for this week compared to previous and to the fact that our forecast is close to the longer-term average.

I don't have a good sense for weather the warming will continue or whether we will remain close to the long term average. NOAA's Climate Prediction Center is forecasting below average air temperatures over the next few weeks, so perhaps we'll hold steady with a slightly delayed season.

Seascape Projects

Seascape modeling project pages