THE EFFECT OF DATA DENSITY ON THE ACCURACY OF FOOT-LINE DETERMINATIONTHROUGH MAXIMUM CURVATURE SURFACE BY AUTOMATIC RIDGE-TRACING ALGORITHM

The influence of data density on the accuracy of foot-line determinati on by the automatic ridge-tracing algorithm is investigated through a set of simulated bathymetric surfaces. The results show that no obviou s differences exist for different sea bottom morphology, different dep ths and gradients of continental slope when data are dense enough, i.e ., when the data interval is smaller than some 12.5 km. When the data are very sparse, the accuracy of the foot-line determination becomes w orse as the surfaces become more complicated or the gradients become s teep. Approximately, the root mean square error of foot-line determina tion is equal to one third of the data interval and the areal error is 23km(2)/100km foot-line length per one kilometre of the data interval .