Scripps Institution of Oceanography

SIO Refrence 62-13. A collection of 3,465 detailed reports of sightings of surface vessels from aircraft of the U. S. Coast Guard are analyzed by pro-bit analysis to determine visual thresholds and measures of variance of the thresholds. Each of seventeen conditions affecting the sighting range is studied separately to determine its effect. Empirical functions are developed to describe the threshold effects of e~ch of the eleven following oonditions in decreasing order of importance: (1) meteorological visibility, (2) altitude of aircraft, (3) ship size, (4) height of major swells, (5) cloud cover, (6) wind velocity, (7) relative bearing of target, (8) sun altitude, (9) relative bearing of sun, (10) wake size, and (11) wind azimuth. Thresholds are developed for the six following discrete conditions in decreasing order of importance: (12) visual aid, (13) range determination method, (H) type of observing unit, (15) time of day, (16) observer, (17) station. In each case, tables of probit results and graphs are inoluded. Only wind azimuth is found to have an insignificant effect. Classification of data ia made by mechanical card sorter, probit analysis by electronic computer and the remainder of the calculations by desk calculator. Measures of precision are listed and2 X , F and t tests are used at the 0.05 level. A table of factors for each condition is included to allow forecasting of sighting thresholds and explanation of use for any probability level. A random selection of sightings is made in order to supply conditions for use of these factor tables as a demonstration of forecasting and as a test of the reliability of the data and of the forecasting method. Suggestions for further study are made.