Significant wave height prediction by using a spatial model

Spatial assessment of variables in a considered region saves its significance for engineering applications. Branches in ocean engineering need the results of this assessment like radius of influence of stations that records the wave measurements and various meteorological variables values. Classical approaches like Kriging do not provide radius of influence for the concerned station. On the other hand, prediction of these measurements from other surrounding stations in the region is a basic requirement. In this paper, it is aimed to predict significant wave height records in a specific region by using trigonometric point cumulative semivariogram (TPCSV) concept. The main difference of this approach from the point cumulative semivariogram (PCSV) approach is the determination of influence radius. More accurate results can be obtained by TPCSV. The spatial correlation and weightings are also obtained through the TPCSV where the distance between two sites is known. The proposed method yields the least prediction error compared with other objective methodologies. The implementation of this methodology is presented for a set of offshore locations distributed along the west United States coastline.