Archive for July 5th, 2008

Dealing with Uncertainty

Saturday, July 5th, 2008

RU17 is about 20 km from the edge of our locally acquired satellite sea surface temperature imagery.  We have currents running generally to the east, and we are making about 12-15 km every 6 hour segment.  We will close most of this distance by the noon surfacing, so by about 6 pm today we will have left our satellite coverage.  From now on we will be relying on our firends for data and forecasts to help us navigate.

Today's Sea surface temperature image (above) does not help us much as we try to define the fronts and eddies of the Gulf Stream extension region.  We'll start using postings from the google earth viewer for the 4 km MODIS imagery we get from NASa Goddard's Ocean Color Web.  Checking the altimetry (below) from U. Colorado, we see RU17 should be approaching that high (red) in the sea surface height centered near 39.5 N, 44W, but we are not seeing any of the currents to the west that we would expect.  We are instead getting currents to the east.  Quite likely the sea surface high is still there, its just displaced to the north of where it appears in this composite data product.



So on to the models.  First a quick check of the weather (below).  I like this image because it has the ship observations combined with the pressure contours. The entire North Atlantic is dominated by a single high pressure system.  A classic summer image, Just like the text books.  Guess we save this one for fall classes. With this high pressure over the North Atlantic, we should not be hoping for much help from the SST.  The sunny weather of a high pressure system just heats up the surface and blurs the features.  One reason ocean feature analysts really like storms is because as soon as the storm clears the area, the ocean is mixed vertically and the horizontal fronts show up like the lights just got turned on. 


So then we move on to the Navy ocean forecast model.  The model results have a band of eastward velocities that are running from the location of RU17 to the northeast to about 40 N, 40 W.    The band is nearly a full degree wide.  RU17 is in this band, and it is heading east, just like the model. 

Nothing like a glider on the scene to help you deal with uncertainty.  We have three sources of information.  One says nothing is happening, the second says the currents are to the west, and the third says the currents are to the east.  The glider tells us which one to believe today.  We also know that it could change as fast as tomorrow, depending on when the altimeters pass overhead, when the winds stir up the ocean so the sea surface temperature looks like the deep field, and how the model's data assimilation scheme is pulling in the new data and dealing with it.  But today, its the Navy modeler's wizzard score that goes up by one.