Sea level trends at locations of the United States with more than 100 years of recording

The paper presents the sea level rises (SLR) computed for the United States tide gauges with more than 100 years of recording. It is shown that the monthly sea levels oscillate about an almost linear longer-term trend with important multidecadal periodicities. The SLR time history is computed by linear fitting of 20, 30 and 60 years of data up to a given time (SLR20, SLR30 and SLR60) and is compared to the value obtained by considering all the data. It is shown that SLR60 has smaller oscillations, while SLR20 and SLR30 have much larger and frequent fluctuations. While SLR60 may oscillate ±10-30 % about the latest longer-term value, SLR30 may fluctuate ±50-100 % and SLR20 ± 100-200 %. The values obtained by considering all the data with a minimum of 60 years (SLRA) also fluctuate ±5-15 % about the latest longer-term value. This indicates the need to use the time history of SLR60 or SLRA when the record is longer than 60 years to assess the accelerating trend. For all the stations, the sea levels regularly oscillate about the linear longer-term trend, and if acceleration has to be computed, this is eventually negative, that is, the SLR is reducing