ESTIMATIONS OF A GLOBAL SEA-LEVEL TREND - LIMITATIONS FROM THE STRUCTURE OF THE PSMSL GLOBAL SEA-LEVEL DATA SET

Among the possible impacts on environmental conditions of a global war ming expected as a consequence of the increasing release of CO2 and va rious other greenhouse gases into the atmosphere, a predicted rise in global sea level is considered to be of high importance. Thus, quite a number of recent studies have focused on detecting the ''global sea l evel rise'' or even an acceleration of this trend. A brief review of t hese studies is presented, showing, however, that the results are not conclusive, though most of the studies have been based on a single glo bal data set of coastal tide gauge data provided by the Permanent Serv ice for Mean Sea Level (PSMSL). A detailed discussion of a thoroughly revised subset reveals that the PSMSL data set suffers from three seve re limitations: (1) the geographical distribution of reliable tide gau ge stations is rather uneven with pronounced concentrations in some ar eas of the northern hemisphere (Europe, North America, Japan), and muc h fewer stations on the southern hemisphere where particularly few sta tions are located in Africa and in Antarctica; (2) the number of stati ons recording simultaneously at any time is far less than the total nu mber of stations with the maximum within the interval between 1958 and 1988; (3) the number of long records is extremely small and almost al l of them originate from a few regions of the northern hemisphere. The sensitivity of the median of the local trends to these temporal and s patial limitations is discussed by restricting the data set in both th e spatial and temporal distribution. It is shown that the data base is insufficient for determining an integral value of the global rise in relative sea level. The effect of polar motion on sea level is modelle d and it turns out to be locally of the order of 0.5 mm/yr, affecting regional trends to an order of 0.1 mm/yr. Thus, this effect can be neg lected on time scale of decades to a hundred years. Though the data se t is insufficient for determining an integral sea level rise, the data , nevertheless, are providing information concerning regional trends i n sea level and even temporal variations in these regional patterns. T hus, the most comprehensive subset of the data reveals the existence o f a significant interdecadal east-west fluctuation of the oceans possi bly indicating some kind of teleconnection. Besides the limitations of the data set, the physical nature of the long term changes in relativ e sea level, which by no means are globally uniform, might render any globally averaged value of relative sea level rise useless for detecti ng climatic impacts or validating climate models.