Long-term trends and regime shifts in sea surface temperature on the continental shelf of the northeast United States

We investigated sea surface temperature (SST) variability over large spatial and temporal scales for the continental shelf region located off the northeast coast of the United States between Cape Hatteras, North Carolina, and the Gulf of Maine using the extended reconstruction sea surface temperature (ERSST) dataset. The ERSST dataset consists of 2°×2° (latitude and longitude) monthly mean values computed from in situ data derived from the International Comprehensive Ocean Atmosphere Data Set (ICOADS). Nineteen 2°×2° bins were chosen that cover the shelf region of interest between the years of 1854 and 2005. Mean annual and range of SST were examined using dynamic factor analysis to estimate trends in both parameters, while chronological clustering was used to determine temporal SST patterns and breakpoints in the time series that are believed to signal regime shifts in SST. Both SST and SST trend analysis show that interannual variability of SST fluctuations shows strong coherence between bins, with declining SST at the beginning of the last century, followed by increasing SST through 1950, and then rapidly decreasing between 1950 and mid-1960s, with somewhat warmer SST thereafter to present. Annual SST range decreases in a seaward direction for all bins, with strong coherence for interannual variability of range fluctuations between bins. The trend in SST range shows a decreasing range at the beginning of the last century followed by an increase in range from 1920 to the late-1980s, remaining high through present with some spatial variability. A more detailed spatial analysis was conducted by grouping the data into 7 regions using principal component analysis. We analyzed regional trends in mean annual SST, seasonal SST range (summer SST−winter SST), and normalized SST minima and maxima. Both the summer and winter seasons were also analyzed using the length of each season and amplitude of the warming and cooling season, respectively, along with the spring warming and fall cooling rates. Trends in all of the parameters were examined after low-pass filtering using a 10-point convolution filter (n=10 years) and regime shifts were identified using the sequential t-test analysis of regime shifts (STARS) method. The analysis shows some difference between regions in the timing of minimum SST with minima being reached 1 month earlier in the south (February) relative to more northern regions (March). Regional annual SST range decreased in a seaward direction. Amplitude of summer warming and the length of summer have shown fluctuations with recent years showing stronger warming and longer summers but generally not exceeding past levels. Overall, the difference in SST range, with recent larger values may be the most significant finding of this work. SST range changes have the potential to disrupt species important to local fisheries due to combinations of differing temperature tolerances, changes in reproduction potential, and changes in the distributional range of species.