2014 Lobster Forecast--Update 1

| No Comments
Seascape has a long tradition of trying to develop ecosystem forecasting.  It started with copepods and whales, and last year, we made our first stab at forecasting lobsters, specifically, the date when the Maine lobster landings begin their rapid summer increase.  Our  forecasts were motivated by the extremely warm year in 2012 and the havoc that it caused in the Maine lobster fishery.  

Here is our first attempt at a forecast for the 2014 fishing season:

Forecast2014_04.07.jpg

Just a reminder that this is the forecast for the start of the "high catch period" when landings really begin to increase.  They are based on the temperature at 20m at NERACOOS Buoy E operated by the University of Maine Physical Oceanography Group.  Full details are at the end of the post.

There's a lot on this figure (we're still experimenting with the best way to present these forecasts).  The blue diamonds are the forecasts, with the width of the diamond representing the 95% confidence interval.  Each diamond is the forecast generated on a specific date.  The forecast at the top is the most current, the one on the bottom was made using data in early February.  Our current forecast is for the landings to start picking up around the Fourth of July, about 5 days later than usual (usual is indicated by the black line).

The sticks and the text at the top indicate the start date from the past.  Our current forecast is for the timing of the 2014 season to look a lot like 2003, which was a pretty cold year (indicated by its blue color).  Years that were warmer than normal are indicated in red, years that are near the average temperatures are in black.  Gray indicates that buoy temperatures were not available in those years.

You'll notice that we were forecasting an even later start of the season (about 8 days delayed) back in the end of March.  This coincided with the coldest temperature anomalies. The warming over the last few weeks has been greater than normal, and water temperatures are almost to the 2002-2011 average for this time of year.  This has caused the forecasts to move back towards the center.  I would not be surprised if this trend continues and that our forecast at the end of April is close to the middle.  

We will try to update these forecasts over the next few weeks.  We also have a proposal pending at NASA that would allow us to improve and expand these forecasts in terms of what we forecast (timing and volume of landings by zone, hard/soft-shell mix) and how early we issue them.

Methods
We downloaded temperature data from the 20m sensor on NERACOOS buoy E.  The best place to look at the current conditions at E is through the NERACOOS Climatology Tool, but if you want the data, UMaine provides netCDF files.  We created a time series of 8-day average temperatures.  The dates on the left side of the figure are the middle of the 8d period used to generate the forecasts.

For each 8d period, we used historical temperature and landings data to build a simple linear model relating temperature on that date to the date when landings begin to increase.  We generated the lobster landings "start date" using monthly landings for Maine. For each year, we divide the monthly landings by the total landings for the year, and then use linear interpolation to find the point where normalized landings increase to 0.08.  These are the dates indicated by the sticks at the top.

Leave a comment

About this Entry

This page contains a single entry by Andy Pershing published on April 12, 2014 11:49 AM.

Another warming-Maine map was the previous entry in this blog.

2014 Lobster Forecast--Update 2 is the next entry in this blog.

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