@ARTICLE\{IMM2016-06946,
    author       = "P. L. Svendsen and O. B. Andersen and A. A. Nielsen",
    title        = "Stable reconstruction of Arctic sea level for the 1950-2010 period",
    year         = "2016",
    month        = "aug",
    pages        = "5697–5710",
    journal      = "Journal of Geophysical Research: Oceans",
    volume       = "121",
    editor       = "",
    number       = "8",
    publisher    = "American Geophysical Union, {AGU}",
    url          = "http://www2.compute.dtu.dk/pubdb/pubs/6946-full.html",
    abstract     = "Reconstruction of historical Arctic sea level is generally difficult due to the limited coverage and quality of both tide gauge and altimetry data in the area. Here a strategy to achieve a stable and plausible reconstruction of Arctic sea level from 1950 to today is presented. This work is based on the combination of tide gauge records and a new 20-year reprocessed satellite altimetry derived sea level pattern. Hence the study is limited to the area covered by satellite altimetry (68ºN and 82ºN). It is found that timestep cumulative reconstruction as suggested by Church and White (2000) may yield widely variable results and is difficult to stabilize due to the many gaps in both tide gauge and satellite data. A more robust sea level reconstruction approach is to use datum adjustment of the tide gauges in combination with satellite altimetry, as described by (Ray and Douglas, 2011). In this approach, a datum-fit of each tide gauges is used and the method takes into account the entirety of each tide gauge record. This makes the Arctic sea level reconstruction much less prone to drifting.

From our reconstruction, we found that the Arctic mean sea level trend is around 1.5 mm +/- 0.3 mm/y for the period 1950 to 2010, between 68ºN and 82ºN. This value is in good agreement with the global mean trend of 1.8 +/- 0.3 mm/y over the same period as found by Church and White (2004).",
    isbn_issn    = "DOI: 10.1002/2016JC011685"
}