Paper No. 4
Presentation Time: 8:50 AM

GLOBAL SEA LEVEL RISE-20TH CENTURY AND RECENT


DEAN, Robert G., Civil and Coastal Engineering, Univ of Florida, P. O. Box 116590, Gainesville, FL 32611-6590 and HOUSTON, James R., U. S. Army Corps of Engineers, Vicksburg, MS 39180, dean@coastal.ufl.edu

Sea level change characteristics over the past century are generally based on tide gauge measurements with the recent addition of satellite measurements commencing in December 1992 and which provide for the first time, global coverage of the World’s oceans. In general, sea level rise can be characterized by a trend and an acceleration. The overall trend and acceleration for the 20th Century are quite well established at approximately 1.7 mm/yr and 0.0 mm/yr2, respectively as demonstrated by reference to a number of journal papers.

Short term global sea level change characteristics are more difficult to quantify with accuracy. The satellite data document that the current sea level trend is on the order of 3 mm/year, substantially higher that the 20th Century rate. Thus, there has been an acceleration which Ray and Douglas (2012) determine to have commenced in the mid 1980’s. The satellite data also document a significant deceleration, thus indicating oscillatory characteristics. Two approaches are applied to examine whether the recent trend increase is part of a long-term change or part of a cycle. First, reconstruction results of global sea level change developed by Church and White (2011) and Ray and Douglas (2012) through application of empirical orthogonal function methods are examined to see if similar past oscillations have occurred. Secondly, the satellite results are examined by fitting an equation which contains both a trend and an oscillatory term. The best fit to this equation establishes the trend and the period, phase and amplitude of the oscillating term.

It is found that the reconstructed global results of both Church and White (2011) and Ray and Douglas (2012) include oscillations comparable to those in the satellite data. Secondly, the fit by the oscillation equation provides a better representation of the satellite data than a linear fit (with the number of degrees in each representation taken into consideration). The root-mean square differences for the linear and oscillation fits are 4.51 and 4.01 mm, respectively.