In this study methods were developed for the prediction of wave conditions that are hazardous to navigation at river entrances, with emphasis on applicability to the Columbia River. There are two basic components to the prediction system: (1) a semi-automated spectral method for forecasting the significant height and average period of waves in deep water, and (2) an index of navigation hazard at river entrances that depends on the significant height and average period in deep water and on the mean current and water depth at the entrance. The computerized, deep water forecast method is a hybrid scheme that combines the spectral principles of the Pierson-Neumann-James method with the graphical input techniques of Wilson and the fetch limited spectrum of Liu. The significant heights generated by the method are well verified by winter wave measurements at Newport, Oregon. The hazard index is based on the probability of wave-breaking in water of arbitrary depth and current. The breaking probability is derived under the assumption that wave heights and squared periods are statistically independent and distributed according to a Rayleigh probability density function. The breaking-wave probability and the hazard index depend on the wave steepness in deep slack water and on the depth (relative to the wave period squared) and current (relative to the period) at the river entrance. The dependence on depth and current is achieved in two ways: (1) the limiting steepness (breaking index) k found as a function of relative depth and relative current, and (2) the wave spectrum in water of arbitrary depth and current is found by transformation of the spectrum in deep slack water. The transformation is performed by requiring that the rate of wave energy propagation remain constant. The hazard index is closely related to the probability of breaking swell. At water depths that are typical of river entrances, the hazard index depends strongly on the significant wave height, mean current and depth, but only weakly on the mean wave period (since the breaking height of swell at such depths is only weakly dependent on period). Hindcasts of deep water significant wave heights and hazard indices compared reasonably well with measured heights and Columbia River bar closure periods. Forecasts based on accurate prognostic weather charts should provide similar results.

Waves critically affect man in coastal regions, including the open coasts and adjacent continental shelves. Preventing beach erosion, designing and building structures, designing and operating ships, providing marine forecasts, and coastal planning are but a few examples of projects for which extensive information about wave conditions is critical. Scientific studies, especially those in volving coastal processes and the development of better wave prediction models, also require wave condition information. How ever, wave conditions along and off the coasts of the United States have not been adequately determined. The main categories of available wave data are visual estimates of wave conditions made from ships at sea, scientific measurements of waves made for short time periods at specific locations, and a small number of long-term measurements made from piers or offshore platforms. With these considerations in mind, the National Ocean Survey of the National Oceanic and Atmospheric Administration sponsored the Ocean Wave Climate Symposium at Herndon, Virginia, July 12-14, 1977. This volume contains papers presented at this symposium. A goal of the symposium was to establish the foundations for a com prehensive and far-sighted wave measurement and analysis program to fully describe the coastal wave climate of the United States. Emphasis was placed on ocean engineering and scientific uses of wave data, existing wave monitoring programs, and modern measure ment techniques which may provide currently needed data.