Golodov M. Improvement of methods of measurement depths at the sea shelf

Dissertation work is focused on reduction of navigational vessel accidents by developing of effective methods for sea depths measurement in case of mixed distributions of navigational parameters measurement errors and factors affecting the process of depth measurement. The work covers the development of methods for determination of depth measurement accuracy dependence on vessel positioning determination errors in cases of positional errors distribution according to the Gaussian law, as well as in case of mixed laws for probability distribution of positional errors. Analytic expressions for distribution density of depth measurement errors have been formulated, that allows depth measurement errors of distribution density to be collected depending on two-dimensional distribution density of vessel positioning errors. An analysis has been performed and actuality of forecasting and determining of hydrological sea parameters have been defined in order to calculate the correction for sound velocity in water by using methods for determination of weighted average value, scheme of iterated-correlation type, vertical distribution forecasting of hydrological parameters together with integral models of active sea layer and inertial forecasting method. Direct methods for sound velocity forecasting have been suggested: an effective sound velocity calculation with regard to its vertical distribution and a method using two-dimensional polynomial for high precise calculation of sound velocity. The characteristics of influence of sea level and bottom soil structure on finite depth values have been determined. Formulas for generic bottom relief forms in the shape of undersea mounts and saddles were derived and errors modeling due to replacement of sea bottom surface by an osculating plane were performed. This has allowed calculating mean, Gaussian and principal curvatures, as well as normals tilting angle to sea bottom surface with positions values of regular network in the area of investigated underwater object or bottom relief section. It was developed a method for optimal interpolation of hydrological parameters using spatial and temporary autocorrelation mean-value function of predicted element allowing it to make extrapolation and estimate forecasting error. Vertical interpolation is performed by combination of weight parabolas and linear interpolation that is represented as an iterative scheme of successive approximations with weighting function. Generally, an algorithm developed in the work allowing improved accuracy in measuring depths, takes into account systematic errors in vessel positioning determination, makes it possible to estimate confidence intervals of particular depth measurement error for various probability distribution laws, as well as carries out an optimal interpolation of previously obtained hydrological sea parameters providing high-accuracy forecast of real-time sound velocity and in the area of depth measurement. Validation of theoretical results obtained was performed using the actual experiment on hydrological parameters forecasting, calculation of sound velocity in the water together with method offered in this work for the Black Sea in the vicinity of the Zmiinyi Island.