because the Earth sciences require thinking over an incredible range of
time and space scales. I started my career worrying about how one atom
in a molecule bumps against another in seawater. Such chemical
reactions happen over time and space scales too short to properly
imagine... the best my brain can do is a grossly-blown up and slowed
down facsimile, akin to representing a blood pumping through our veins
as a glacier etching its way through a valley. Nowadays I'm working on
how a planet-spanning ocean is changing in response to patterns in human
behavior that emerged over hundreds of years of industrialization.
Earth's climate has been changing in response to greenhouse gases for as
long as I've been alive. This is still a difficult time scale for me to
imagine, but it is at least within reach. However, even if we stopped
emitting carbon and other greenhouse gases today, we'd expect Earth's
climate to continue to respond for many thousands of years to come. It
will likely be between hundreds of thousands to millions of years before
all of the gases we've emitted have been stripped out of the atmosphere
and sequestered back into various geological reservoirs by natural
cycles. My poor mind has no chance at imagining anything this long...
Timescales are on my mind partly because almost everything we do on the
boat is cyclical. Every 8-9 hours a rosette goes down to the ocean
depths. Then it comes back up; there's a flurry of communications
between technicians, scientists, and the crew piloting the boat; and the
samplers head out to collect their seawater. The samplers make their
measurements while we steam to the next station. The cycle begins anew.
Meals and shifts fly by. I pass out pieces of a chocolate bar once
each week, if only so we don't forget this vestigial unit of time
entirely while we are away from our lives on land. I keep a list of
stations to remind myself that the cycle is not endless. With each
iteration, another station is crossed off our list and we are 30-40
nautical miles closer to port... to planes that will take us home to
wives, husbands, children, friends, and pets.
I've also been thinking about timescales in the context of the science
we are doing. One of the great difficulties with using repeat
hydrographic measurements is making sense of changes which are happening
over a wide range of timescales. We measure our sections once every 10
years or so with the goal of examining climate oscillations happening
over ~10 years and climate change happening over longer timescales
still. However, tides, eddies, seasons, movements of fronts, and
internal waves (waves that move between seawater masses with different
densities much like a typical wave moves along the surface of the ocean)
are all capable of moving water around in the ocean--even the deeper
parts of the ocean--over timescales much shorter than 10 years. This
means we sometimes see big changes in the properties we are measuring
that are just due to our choice of which day or month of the year we
have chosen to make measurements!
So, how do we make sense of our measurements despite these fast
variations? One main way is through the use of the robotic floats that
we are so excited about. These floats will keep measuring the ocean
once every 10 days. This means they don't have to worry about these
short timescale processes nearly as much. Why? It isn't a problem if
they measure one profile in an eddy because they would also measure
another ~364 in the span of time it would take for repeat hydrography to
measure the ocean a second time, and the fast variations would average
out. (I should note that a single float would run out of batteries
before the full 10 years were up). My research involves a second way to
get around fast variability, which is to take advantage of the fact that
we are measuring such a large number of seawater properties all at once.
When a passing internal wave pushes the water we are measuring deeper
down in the ocean, all of the various things we are measuring tend to
tend to look as though we measured them just a little bit shallower.
With a bit (read: lot) of coding (read: hair pulling), we can figure out
how much an internal wave, eddy, or tide has changed the properties of
the water we are measuring by considering the measurements collectively.
We can then look at the changes in the measurements that cannot be
explained by these fast changes. This is another reason I'm so excited
about the prototype floats which are capable of measuring more
properties than the typical floats: the more distinct measurements we
can make, the more weapons we have in our arsenal to figure out exactly
what changed between our sets of measurements. Data assimilating
computer models are a final way to interpret repeat hydrographic
measurements in the context of long term changes. It would take a whole
post to explain these computer models though... perhaps next week. For
now I'll just say these models devour data in an effort to make a
climate model reflect what actually happened on Earth rather than just
what could plausibly have happened.
My Advisor has the dream of an "all of the above" approach to measuring
ocean climate changes (I do as well, but I credit him with convincing me
that the scientific community and the technology are ready). His vision
is of a network of profiling floats in the Southern Ocean outfitted with
the full array of the new sensors. The floats would be calibrated using
the high quality measurements made on the hydrographic cruises that
deploy them, and the data the floats collect would be passed into
data-assimilating computer models. The models would then do the brute
number-crunching on the huge volumes of data sent back to land by the
floats. Monitoring climate change in the ocean is daunting challenge
with monumentally high stakes. I'm not sure anything short of his
vision could get the job done. Many of us in the oceanographic
community have been working to refine the sensor technology and computer
modeling software we need to realize this vision. This P16S cruise is
case in point, as we have deployed a number of the floats with the new
sensors and software as part of a proof-of-concept for the larger study.
The Southern Ocean State Estimate computer model (see Veronica's figure
from the last post) is already assimilating similar float data from
older floats, and some of my research with colleagues at the Scripps
Institution of Oceanography will allow this model to assimilate the data
from the new sensors as well. The funding proposals to get the broader
network started are pending. Fingers crossed!
~2 weeks to go!
No comments:
Post a Comment