18 



including the noise present in the records. Enhancement of the existing 

 data base through new measurements will most likely occur through satellite 

 altimetry once this is proven to centimeter accuracy over the open ocean. 

 Additionally, in some cases much can be learned locally about 

 anthropogenically generated compaction in areas of tide gages through the 

 installation of rather simple compaction measurement devices. One feature 

 of new data is the length of time that will be required for such data to 

 "mature" to yield significant meaningful information. 



2.4.1 Use of Existing Data 



Analysis in light of the physics of RSL change appears to be the most 

 effective and productive use of existing data. In particular, accounting 

 for the contribution of long period waves as explored by Sturges (1987) 

 would allow interpretation and removal of a major portion of the noise in 

 the RSL measurements . 



A second productive area is a more thorough analysis than presented 

 previously of the contribution of post-glacial adjustment of the earth 

 following the last ice age. As noted previously, Lambeck and Nakiboglu 

 (1984) have inferred from viscous models of the earth that the actual 

 eustatic rise is roughly one -half to two -thirds the value determined from 

 analysis of records based only on areas of relative stability. Improved 

 estimates of eustatic sea level rise could be based on either a more 

 inclusive data set with or without the use of a viscous earth model. 

 Obviously more meaningful results could be obtained with the combined 

 approaches simultaneously. The approach envisioned here is in general the 

 same as applied in "physical principles" with the addition that the global 

 viscous model would be employed for interpretation, guidance and 

 confirmation of the results obtained. 



Most approaches of direct analysis attempt to reduce the noise in a 

 record on a station-by- station basis through determining some sort of RSL 

 estimate through fitting to the data. Unfortunately, the noise in 

 individual records is such that at least 20 to 40 years of data must be 

 available at the individual gages before these results can be considered 

 meaningful. An approach that would make these results meaningful early 

 after their availability is the weighted averaging of many stations along a 



