Abstract Management of uncertainty in model predictions of long-term coastal change begins by admitting uncertainty. In the case of geometric mass-balance models, the first step is to relax restrictive assumptions to allow for open sediment budgets, time-dependent morphology, effects of mixed sediment sizes, and variable resistance in substrate material. These refinements introduce new uncertainty regarding the choice of parameter values. The next step is to actively manage uncertainty using techniques readily available from information science. The final step requires a shift in coastal management culture to accept decision making based on risk-management protocols. Stochastic simulation was applied to manage predictive uncertainty in cases involving complications resulting from open sediment budgets, rock reefs, and seawalls. In these examples, the respective effects caused between 20% and 60% difference from conventional predictions based solely on equilibrium assumptions and substrates comprised entirely of sand. Stochastic simulation makes it possible to establish confidence limits and determine the statistical significance of differences caused by varying effects such as substrate resistance and shoreface geometry. It also enables the likelihood of critical impacts to be specified in terms of probability. Moreover, probabilistic forecasts provide a transparent basis for coastal management decisions by revealing the consequences if quantitative estimates prove to be wrong.

Using a conservation of sand approach, the effects of a rising sea surface are quantified and separated from other causes of shore retreat. Sitespecific data important in predicting shoreline changes are: (1) Initial shoreface and backbeach profile; (2) subsequent backbeach profile; (3) relative sea level rise; (4) grain size distribution of sediment landward of the shoreface; and (5) net quantity of sand‐sized material that enters or leaves a specified coastal reach. A key element of the approach, Bruun's assumption of a shoreface is dynamic equilibrium with the sea surface, was evaluated and found to be reasonably accurate. Field application of the method shows that sea level rise accounts for about 53% of the total shore retreat of 5.5 m/yr measured at Smith Island, Virginia, and for about 88% of the measured, 1.7‐m/yr retreat of the barrier island south of Oregon Inlet, North Carolina. Net sand losses account for the remainder. Because shoreface adjustments are required to maintain an equilibrium profile, sand replenishment is probably the most realistic method to stabilize a shore against the effects of relative sea level rise. Conversely, a negative sediment budget may also be mitigated by structures which hinder the movement of sand away from a problem beach and enhance its deposition there.

: The saltwater intake structure (groin) built in 1982 by the United States will not cause sediment to accumulate and cover either this or nearby Sultanate Oman Air Force (SOAF) intake structure. The intake groin will not cause erosion but will create a slightly more stable beach. If a proposed pier is built over the U. S.-built groin, or as near south of it as possible, pier length will be reduced to a minimum; if a compound pier--consisting of a rubble-fill section extending to the end of the existing groin and a pile-supported section extending to the pier's end--is constructed, sedimentation problems at the U. S.-built and SOAF saltwater intakes will be minimized and effects on nearby beaches will be acceptable. In addition to presenting these conclusions, the report provides wave and current data for use in design of the proposed pier and presents a means of using current speed and wave height to predict percents of time the pier can be used for vessels with specific handling characteristics. Appendices A-E provide background, describe data collection and analysis procedures used, and present more specifically the findings on which the report's conclusions are based.

: Time-sequence aerial photos often constitute the only source of data to determine past shoreline changes. This report presents a method for obtaining shoreline change data from base maps constructed from time-sequence sets of aerial photos, with the image of the aerial photos superimposed at the constant scale of each base map. Shoreline position and other features of interest, such as vegetation line location and orientation of breaking wave crests near the coast, etc., are then traced on each base map. A comparison of each base map from the different sets of aerial photos will provide shoreline change data through time. (Author)

Sedimentation may be an important problem when quantities of suspended material are carried into an enclosed harbor on a flooding tide. In order to forecast future maintenance costs, two methods for predicting the sedimentation rate prior to harbor construction are proposed: 1) a sedimentation tank to be placed at the proposed harbor site, and 2) a mathematical model which uses sediment and hydraulic data collected at the harbor site. Certain considerations in the design phase of a project may effect a reduction in harbor sedimentation. If feasible, the harbor may be sited in a region where suspended sediment concentrations are low and sediment sizes (settling velocities) are small. Proximity to river sediment sources may be a factor. Conversely, a harbor site in a clear-water river adjacent to a sediment-laden estuary may be desirable if bedload transport during freshets would not be a problem. Settlement of suspended material may occur in the channel which connects an enclosed harbor basin with navigable waters. This material may subsequently be resuspended and carried into the basin thereby increasing the sedimentation rate. To reduce that rate the channel should be designed as short as possible. A sill in the channel may also be used to reduce initial excavation costs and the sedimentation rate. Flotation for vessels in the basin will be provided at all times, but movement into and out of the harbor will be reduced to times of higher water. In high latitude areas where harbor use is limited to periods when ice cover is absent, the sedimentation rate may be reduced using a channel closure structure during non-use periods. Winter sedimentation rates can be predicted using the mathematical model for summer conditions, and when ice thickness is known.

: Sand volume changes above mean sea level (MSL) and shoreline position changes at MSL were obtained from 4400 beach profiles acquired over a 10-year period along three New Jersey barrier islands. The results provide insight into the behavioral characteristics of sandy ocean beaches. Storm changes were highly variable between islands, and between profile lines on the same island. Often changes on profile lines less than 0.8 km apart were opposite in sign, suggesting a closer profile line spacing is required to obtain an accurate picture of storm changes. On two islands a definite seasonal change was found when 10-year data were averaged. The maximum sand volume and most seaward shoreline position occurred in August and the least in the January-April period. A year-to-year comparison of surveys would be best using data collected from January through April because changes from month to month were least then. Large variations in beach changes were measured from one year to the next, and on one of the three islands 10-year data did not appear sufficient to establish a long term trend in beach behavior. (Author)

A beach monitoring program between 1962 and 1972 at Atlantic City, New Jersey was designed to observe the response of beaches to waves and tides of specific intensity and duration as a first step in developing a storm warning system for low-lying coastal communities. As a by-product of that study the behavior of beach sand following two beach replenishment projects in 1963, and again in 1970, was determined. Monitoring was done using repetitive beach surveys above mean sea level (MSL) at seven profile lines. Survey results show that following replenishment, losses of the fill material above MSL were between nine and twelve times the losses measured in adjacent non-fill areas. Loss rates were largest at the updrift end of the fill region. About two and one-half times more material appeared to move in a seasonal on-offshore direction than moved permanently alongshore and above MSL to the southwest. For each meter of beach retreat, 5 to 6 m /lineal meter of fill were lost.