Last weekend, the Portland Press Herald
ran a story on how lobstermen in Maine are worried about a repeat of last year's warm temperatures and the havoc that the warming caused on the lobster fishery (early inshore migration and ramp-up of high catch period, glut of soft-shells, record catch, collapse in price, etc.). The paper used the NERACOOS "Ocean and Weather Climate Display" tool
to conclude that temperatures at 50 m (locations not specified) are running 1-2° above the average. My colleague Rick Wahle was quoted "We aren't likely to see as an extreme of an event as last year, but I wouldn't be surprised if we see an earlier-than-normal shed." This prompted me to wonder whether conditions in April are a good predictor of temperatures in the rest of the year?
I grabbed the data from NERACOOS Buoy E
. This buoy has been in the water since 2001, and I think it's location makes it a good indicator of conditions along the coast of Maine. I then averaged the 0, 20, and 50m temperatures into 8 day bins. Then, I looked at the correlation between the bin starting on April 15 with the temperatures later in the year:
Not surprisingly, at all depths, the correlation with April 15 decreases the further you go on the year. For the surface temperatures, the correlations decline quite rapidly, but at 20 and 50m, the correlation is quite strong, even into December. The rapid decline in the correlations at the end of the year reflects the influence of winter mixing and cooling. My hunch is this continues to decline until late March, when the minimum temperature is reached.
The strong correlation between April and summer temperatures can be used to forecast conditions this summer. Unfortunately, the 20m and 50m instruments at Buoy E seem to be offline. However, nearby Buoy B is running 1°C above its mean temperature, so we'll use that. Note: the 1-2° increase over the mean reported by the PPH seems to be driven a layer of very warm, salty water in Jordan Basin. This is potentially an intrusion of temperate, perhaps even Gulf Stream, water. Worth keeping an eye on this, but my suspicion is that it is an isolated pocket of water that will not impact the more coastal areas we're interested in. So, if I plug a 1°C anomaly into the equations relating April temperature to the rest of the year, I can predict the temperatures for this summer:
The blue line is the average temperature at 20m at Buoy E, and the red dashed line is the prediction. The shaded area indicates the uncertainty around the prediction. Based on the uncertainties, there's a 75% chance that bottom temps will be warmer than average this summer and a 50% chance that they'll be more than 0.5 degrees warmer. There is only a 1 in 10 chance that we'll get temperatures as warm as last summer.
It's worth noting that this prediction assumes that atmospheric conditions will be normal. During my temperature-cod analysis
, I was able to estimate 20m and 50m temperatures using temperatures from the surface. This suggests that while the 20 m waters might be separated from the surface, the isolation is not complete. Weather conditions will have a weak effect below 20m, even in the summer, and likely, the deeper layers accumulate the changes coming from the atmosphere. If the next few months are warmer (or cooler) than average, then I would expect the 20m predictions to run slightly warmer (or slightly cooler). Right now, NOAA's Climate Prediction Center
says there is a 33% chance that New England will be warmer than the 1981-2010 average:
If this plays out, I would expect the summer 20m temperatures to be even warmer than I predict, but a repeat of 2012 is still unlikely.