AGU Fall Meeting Day 3 (Wednesday)
Day 3 was an easier day for me with fewer talks that I wanted to see so I was able to spend a bit of time thinking about work matters and cogitating a bit more on the talks I was listening to. I also spent a large chunk of time in one of the poster sessions, which was very enlightening; it was like an informal talk session with authors presenting their posters ad hoc and groups discussions taking place about the results. (As I write this I have just come from my own poster session and had the same experience; far better than many poster sessions at other meetings.)
The first talk I attended was on an initiative to map high latitude lakes and provide a data product for scientists. The mapping involved a new algorithm that appeared to perform far better when faced with satellite images of lakes during partial ice-out. The author mentioned that a new product for Alaska had been released a week or so ago and this will be very useful for a project proposal I am currently involved in writing.
The next talk, by a student of Stephen Carpenter and Michael Pace, was something I have been interested in for some time; identifying early warning indicators in ecological communities prior to regime shift. Several early warning indicators are invoked from theory of tipping points and regime shifts
- increased autocorrelation
- increased variance
- increased skewness
- conditional heteroskedasticty
The latter can be thought of as “autocorrelation” in the variance of the system. This indicator was demonstrated to be found in data from experimentally manipulated lakes where the fish community was altered. I got plenty of ideas from this talk, with a view to doing similar analyses on palaeoecological data.
I later spent a bit of time in a session focussed on the marine nitrogen cycle over the past few billion years. Not really my area of interest, but plenty of interesting ideas with regard to compound specific N isotope measurements, including information on compound specific pigment isotopes.
I went to a poster session on palaeo records of change in lake ecosystems. I met up with Dave Porinchu, someone I spent a week with in Helsinki at a Cladocera workshop 10 years or so ago. Consumption of a bottle of Talisker comes to mind. I met several other people here too; Yarrow Axford had an interesting poster on chironomid temperature reconstructions in a West Greenland lake for example and I chatted to her for a while about palaeoecological transfer functions, many of which I just don’t believe even though I teach the theory every year to plenty of palaeoecologists! I also bumped into Simon Goring, whom I’d had some email correspondence with a year ago regarding a new transfer function methodology he’d come up with. I joined him and some colleagues for a nice lunch in a local Thai restaurant.
Having had a two-course lunch, I didn’t need further sustenance and after an email session I headed back to my hotel and set to work on my river water temperature models, squashing an annoying data-processing bug. I set the laptop off data crunching through the multivariate smooth model fitting whilst I slept and now I have something to work with on the flight home so I can finish of my additive modelling paper for river water temperature trend modelling.
And that was day 3…
