Regional Report Cards

U.S. East Coast
Sea level is rising faster north of Cape Hatteras.

Although sea level is rising everywhere along the U.S. East Coast, the rate is significantly higher north of Cape Hatteras, as exemplified by comparison of tidal records from Norfolk, Virginia and Wilmington, North Carolina. These sites make a good pair for comparison because they have similar record lengths, unlike other NOAA tide stations near Cape Hatteras.

The cause for the different rise rates appears to be oceanographic behavior associated with the Gulf Stream, which heads offshore north of Cape Hatteras.

In this, like many other analytical efforts, statistical confidence intervals are a way of expressing confidence that the linear trend difference is real and not simply a matter of random chance stemming from the sizable variation seen in the monthly sea-level values observed at both sites. But be aware that the confidence interval for a sea level trend in mm/year should not be confused with a confidence interval on sea level height in meters. While the upward slopings solid lines in the chart below depict rising trends, they are actually a plot of predicted height versus time. The dotted lines above and below each trend (LHi and LLo) show the uncertainty about the predicted mean height in any one year but say nothing about the trend itself.

 

A further word about confidence intervals: The ones NOAA uses here are based on the standard deviation of the predicted mean height which is smallest in the middle year of the series increasing in either direction away from the middle in a classic ‘bow-tie’. If the diverging black lines in the NOAA graphs were extended forward to the year 2050 they would show the range of uncertainty in the predicted mean RMSL height for that year. A different statistic was used in the VIMS non-linear sea level graphic shown previously for Norfolk (Graph 1). The height interval between the dashed red lines in that graph is based on the standard deviation of the individual monthly observations and includes approximately 95% of those observations. When extended, the same dashed red lines project the range within which 95% of all future values of monthly RMSL height are likely to be found, including those expected in 2050.

An obviously important question is whether the current trends are likely to change as more data become available at tide stations like those at Norfolk and Wilmington during our proposed Anthropocene Epoch. For linear trends, the answer is seen in a pattern of generally increasing rise rates as series length increases. Starting with a trend fitted to monthly observations for the years 1969-2004 (36-year series length) and following up with biannual additions through 1969-2016 (48-year series length), linear trend rise rates for Norfolk and Wilmington show the following change with increasing series length (Graph 4):   

As seen in Graph 4, linear sea level rise rates for more than a decade have consistently been higher at Norfolk compared to Wilmington although Wilmington has shown a surprising increase in rise rates in the last four years. At Norfolk, an upward bump in rise rates took place about seven years ago but has since leveled out at slightly more than 5 mm/year.

Quadratic trends for Norfolk and Wilmington (Graph 5), though more irregular, help to explain what is happening to the linear trends; higher rates of sea level acceleration at Norfolk have declined over the last decade while statistically insignificant acceleration at Wilmington during most of the decade has recently become highly significant, surpassing the rate of acceleration at Norfolk in a ‘crossover’ at the beginning of 2016 – a year and a half ago as this is being written.

Graph 5

Are Norfolk and Wilmington representative of sea level trends their respective regions? In general, yes. Most U.S. NE coast tide stations presently have significantly higher linear rise rates than their neighbors to the south, Norfolk being the highest and Wilmington one of the lowest for the 1969-2016 period. Wilmington’s recent acceleration uptick is seen at all SE stations with 1969-2016 records but at none of those to the north. These are trends worth watching.