February 2010 Archives

OSM Day 7--Press Conference

| No Comments | No TrackBacks

Note:l my ambition to blog each day fizzled like a Portland (OR) rainstorm.  We'll try to add some additional reports from the meeting in the coming days. 


For me, the most memorable part of the meeting was being invited to do a press conference--something I've never done before.  I was invited to give a regular science talk in a session on the impact of climate change on marine ecosystems.  I thought I would use this as an opportunity to talk about some calculations I've done characterizing the carbon footprint of whaling (see this earlier post).  AGU, one of the societies that was running the meeting, thought the news media would be interested in this topic.


The hardest part was deciding to do it.  Since I hadn't presented my calculations to many other scientists, I was worried that there was something I was overlooking.  Visions of cold fusion were dancing in my head.  In the end, I decided to go for it.  To prepare, I organized a mock press conference at GMRI, with my colleagues acting as journalists.  This was extremely helpful.  At the conference, I spoke for about 15 min:

IMG_1935.jpeg

and then took questions.  In addition to the reporters in the room, there were a couple joining on the phone.  I then spoke with several reporters one-on-one, including the BBC:

IMG_1937.jpeg

The BBC story was online by 11PM (PST) last night, and by this morning, it had been translated into Hungarian, Slovenian, and Italian (I didn't know I was fluent in Italian).  Here are some links to a few of the stories, if you're interested in reading more.  All in all, a really fun experience.


BBC

Environmental Research Web

Discovery News

OSM Day 4--Fred's Talk

| No Comments | No TrackBacks

The official meeting started at 8AM this morning.  Meetings like this are the intellectual equivalent of drinking from a firehouse.  At any given time, there are 15 different sessions in progress.  Each session is organized around a particular theorem, and the themes at this meeting cover the full gamut of oceanography.  About a year ago, groups of scientists submitted proposals for sessions.  Once the sessions were selected, the oceanographic community was asked to submit abstracts.  An abstract is a brief (~one paragraph) description of a study, and when you submit an abstract, you select which session you think is most appropriate for your work.  Then, one of three things happens.  1. The session rejects your abstract, possibly passing to another session, 2. The session accepts your abstract and invites you to give a talk, or 3. The session accepts your abstract and asks to you prepare a poster.


Usually, breakfast is spent looking over the titles of the talks, and figuring out which ones you'll try to see.  One talk was easy to add.  Our very own Frederic Maps gave a talk at 8:30 in the morning on his work modeling copepods.  Depending on the talk and the session, you can have anywhere from a few people to more than 50 (remember, you're up against 14 other talks).  As you an see from the picture below, Fred's talk was quite popular:

FredTalkPic.jpg

OSM Day 2: Workshop

| No Comments | No TrackBacks

The meeting officially starts tomorrow (Sunday) night, so why did I fly out on Friday?     Science has always relied on communication and collaboration--hence, the need for conferences.  Oceanography is an inherently interdisciplinary field, but it is very hard to get truly interdisciplinary projects funded. One way to get some interdisciplinary work done is to organize a workshop.  The idea behind a workshop is to get a few very busy people to take a few days from their day-to-day work in order to work together on a common problem.  So, this is why I'm spending this weekend in a conference room instead of hiking with my family.


The point of this weekend's workshop is to develop a better understanding of how changes in the Arctic affect the North Atlantic.  I've stumbled into this line of research by uncovering a dramatic change in the Gulf of Maine plankton community that took place around 1990.  Turns out, lots of other things changed right around that time: the waters became less salty and began flowing faster, herring became more abundant and right whale calves became rarer.  Many of these changes were observed from New Jersey up to Newfoundland.  The best explanation so far is that these changes originated when the winds over the Arctic pushed a slug of fresh water and ice into the North Atlantic.  This created a pocket of fresher water that eventually moved down to the Gulf of Maine:



 The conditions that created this slug persisted through much of the 1990s.  The workshop, organized by my colleague (and former Ph. D. advisor) Chuck Greene, has brought together biologists like me, physical oceanographers, and Arctic climate specialists to try to get a better understanding of exactly what happened. 


OSM Day 1: I'm on a plane!

| No Comments | No TrackBacks

The Ecosystem Modeling Lab is headed to the other Portland for the biennial Ocean Sciences Meeting.  OSM is the main oceanographic conference.  Conferences like OSM are an important part of science.  They provide an opportunity to learn about the latest developments in the field, catch up with colleagues, and find collaborators, employers, students, and post docs.  This week, I'll try to give an inside view of  a scientific conference.


First up, getting there.  I'm writing you from a tiny little desk in the middle of O'Hare airport.  I find that I can often get a lot of work done while traveling.  Here's me working on my flight:


IMG_0619.JPG

On the flight, I managed to update a time series of temperatures from the Gulf of Maine and do some analysis of the relationship between the number of right whale calves and the amount of their food.  I walked off the plane feeling good, but with a battery at 50%.  I've become pretty good at finding power outlets in airports, but O'Hare seems to be doing a good job hiding them.  They do provide some very tiny desks with power outlets.  Between the small desk, uncomfortable stools, and exposure in the middle of the concourse, I think I'll be moving on.  

Ecological forecasting system -- the concept

| 4 Comments | No TrackBacks
Early numerical/computational weather forecasts could not compete with traditional forecasting methods.  Simple lore-based indicators--like "red sky at night, sailors delight"--could, in many cases, out-perform computational models.  In fact, simply predicting that tomorrow's weather will be just like today's was much more reliable than the cyclone-ridden numerical predictions put forth by scientists.

Over the past 50 years, computational power, theory, and data collection have improved, leading to a revolution in weather forecasting.  What began as an inferior  practice quickly overtook other methods.  Computational predictions have now replaced traditional subjective predictions, and detailed weather forecasts are integral parts of our everyday lives.  The figure below (adapted from Shuman 1989) shows how computational weather forecasting models have progressed since the 1950s.

Shuman1989edit.jpg
Figure adapted from Shuman 1989.  Error in numerical weather forecasts has decreased substantially since the 1950s due to improvements in the models and the computers.

In contrast to weather forecasts, there are very few ecosystem forecasts available.  Those that are available are at the cutting edge of science, and are often released only after the passage of the events they are trying to predict.  Yet I would argue that we could be on the cusp of a similar revolution in ecological forecasting.

Many lessons from the past 50 years of weather prediction can be carried over to the development of systems designed for ecological forecasting.  There are three key components to a good forecasting system, all of which we can begin to implement and/or take advantage of right away.  These are: monitoring (steady streams of data), adaptive computational algorithms, and forecast availability.  The common service provided by all of these components, and the crux, I would argue, to a revolution in ecosystem forecasting, is feedback.

I've sketched out below a concept map of an ecological forecasting system, highlighting the mechanisms of feedback.  Briefly, data is input to the algorithm, which produces a forecast.  This forecast then becomes available to a number of feedback avenues, including the designer, users, and the algorithm itself.  (Side note: I used the "Concept Map Builder" designed by the Center for Ocean Sciences Education Excellence, which is a tool under development, worth checking out.)
ForecastingConceptMap.jpg
Monitoring is, of course, a critical component as it provides input data.  In this system, however, data plays another role.  Suppose we are producing weekly forecasts.  Then each week, we want to know how well we did with our forecast.  At first, our forecasts might be little better than a random guess.  That's okay.  As long as we have the data coming in to tell us how well we're doing.  When we're not doing well, we want to know so that we can adjust.

That brings us to the second component: adaptive computational algorithms.  Using techniques borrowed from computer sciences, such as the genetic algorithm, a system can learn based on its past successes and failures.  It's important for the system to be able to assimilate a steady stream of data and to adapt based on that.  Ideally, a system should be flexible enough to incorporate new theoretical information as well.

The third component, forecast availability, is one that has been missing.  In my opinion, this has been holding back progress in ecosystem forecasting.  I suggest that, like numerical weather forecasters, we should be putting our forecasts out there, even while they are little better than subjective forecasts.  It's okay if many forecasts fail.  With an ensemble of forecasts available, and with a steady stream of observations coming in, our computational systems can begin to evaluate, learn, and adapt.  Over time, the failures will presumably be fewer and fewer, and the forecasts will be more and more useful.

There are a number of subtleties that I've brushed over here, for the sake of simplicity and generality, as well as some practical considerations that are nontrivial.  Nevertheless, these are the basic ingredients, as I see it, to a revolution in ecosystem forecasting.  I'll be presenting these ideas at the Ocean Sciences meeting in Portland, Oregon next week.  If this sort of thing is your bag, look me up.


What does a model look like?

| No Comments | No TrackBacks
This intriguing question was posed to us by a high school student.  My initial reaction to this question is to post a link to Heidi Klum.  More productively, I'd like to try to present some ideas for thinking about models and what they look like.

To a scientist, a model is a way of representing an idea about how the world (or some part of it) works.  In many ways, a model is just a way of expressing an hypothesis or a set of hypotheses.  How a particular scientist thinks or the audience they're addressing to will dictate what the model looks like.

For me, I like to start with a conceptual model, usually represented as a drawing.  For example, here's a diagram I use to explain how temperature and chlorophyll influence copepod growth and reproduction:
copepoddiag.jpg
The circle at the left represents an egg.  The red arrows show the path that the egg takes to become an adult copepod.  The arrows are colored red to suggest that how long it takes to go through these stages depends on temperature.  The long arrow at the top represents reproduction (adults making eggs).  The number of eggs produced depends on the amount of food available.  Since this particular copepod eats mostly phytoplankton, the arrow is colored green.  These graphical models are very useful for helping think through a problem.  My notebooks are filled with less attractive versions of these, and most days, there is some version written on my whiteboard with colored markers.

While conceptual models and diagrams are the most common models in science, when most scientists speak of models, they mean a mathematical models.  The advantage of mathematical models is that they force the modeler to be very precise about how the components fit together.  They also can be used to make predictions that can be compared to data.  The disadvantage is that they require mathematical training to understand.  Some mathematical models are relatively simple and can be written on a few sheets of paper.  Other models are more complicated, and this is where computers come in.   Here's a snapshot of some computer code that represents copepod growth and reproduction:
codefragment.jpg
This code is written in a language called "C".  The code is then given to a computer program called a compiler that turns the code into the language of 1's and 0's that the computer recognizes.  We then push a button and wait while the program runs.  The program produces a series of output files.  To view the results, we have to load these files into Matlab and plot them in various ways.  
SEASCAPEapr15.jpg
This is probably my favorite step--part science, part engineering, part art.  Pretty, in it's own way, but no Heidi Klum.

Red tide photogrammetry in Mexico

| No Comments | No TrackBacks

Just a quick note on our sea surface monitoring project. We are working with a group in Ensenada, Mexico to apply our camera system (designed for oil spill mitigation) to a red tide monitoring project. The images below show a dry run, so there is no red tide present, but stay tuned. If this project gets off the ground, it would be a neat application of our system.


Ensenada1.jpg
Original photo

Ensenada2.jpg

Georectified photo

The ground control points (x's and o's) are just eyeballed in this rectification, so there is noticeable error, particularly with the middle point.  This is something I hope to improve upon.  Also, we hope to cover more area with multiple cameras.

About this Archive

This page is an archive of entries from February 2010 listed from newest to oldest.

January 2010 is the previous archive.

March 2010 is the next archive.

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