Appendix F 



Evaluation of Shipboard Visual Estimation 



of Iceberg Size 



LCDR Walter E. Hanson, USCG 



INTRODUCTION 



During 1987, approximately 21 per 

 cent of all sightings entered into 

 the International Ice Patrol (IIP) 

 computer were from visual ship 

 observations. Shipping has each 

 year contributed a significant 

 number of visual sightings. Many 

 studies have assessed the ability 

 of ships to detect icebergs, 

 primarily by radar (Budinger, 1960; 

 Ryan et al, 1985; and Harvey et al, 

 1986). However, little information 

 is known about the sizing accu- 

 racy of visual sightings. Conse- 

 quently, the 1987 iceberg IIP 

 iceberg deterioration study, 

 described in Appendix E, pre- 

 sented an opportunity to evaluate 

 the ability of shipboard observers 

 to visually estimate iceberg size. 



This study evaluates visual sizing 

 efforts which had neither the aid of 

 visual cues (i.e. an object in close 

 proximity for size comparison) nor 

 the aid of stadimeter or sextant. 

 This sizing technique may mirror 

 that of the shipping community. 

 The icebergs studied by IIP were 

 primarily medium-sized, non- 

 tabular shaped. This category of 

 iceberg seems to be the most 

 often sighted by shipping. 



BACKGROUND 



The IIP uses iceberg size and 

 shape in sighting reports to predict 

 their drift and deterioration. For 

 operational purposes, only seven 

 different categories of icebergs are 

 modelled (Mountain, 1980). They 

 are: 



" Growler 



° Small, Non-Tabular 



° Small, Tabular 



° Medium, Non-Tabular 



° Medium, Tabular 



° Large, Non-Tabular 



° Large, Tabular 



Drift and deterioration predictions 

 are computed twice daily using 

 computerized models. Iceberg 

 size is used differently by the two 

 models. 



In the drift model, the size and 

 shape parameters together select 

 one of seven profiles. Each profile 

 is a different cross-sectional 

 representation of above-surface 

 and sub-surface area. The profile 

 represents the average dimen- 

 sions for icebergs in that size and 

 shape category. The iceberg is 

 drifted based on the forces acting 

 upon the profile. The profile is not 

 changed until a new size and/or 

 shape is specified. 



In the deterioration model, the size 

 and shape parameters together 

 select one of seven waterline 

 lengths. The model calculates 

 "melt" in terms of length instead of 



mass. Each of the seven lengths 

 is assumed to be the maximum 

 value for the particular size and 

 shape category. Environmental 

 conditions and waterline length 

 are then used as inputs for the 

 daily "melt" of the iceberg. 



DATA COLLECTION 



The USCGC TAMAROA, a 68 m 

 (205 ft) U. S. Coast Guard cutter, 

 was used for the 6.3 day study. 

 Visual obsen/ations were made 

 from the bridge wing; height of eye 

 was 10.8 m. Two IIP ice observ- 

 ers, who each had at least two 

 years of aerial iceberg reconnais- 

 sance experience, made the 

 observations. 



Iceberg above the waterline 

 dimensions were measured during 

 daylight (from 0800Z to 2400Z) in 

 al! weather and light conditions. 

 Table F-1 shows the hours when 

 iceberg sizing occurred. Iceberg 

 shape and size were both esti- 

 mated by the ice observers and 

 calculated from photographic 

 images scaled according to 

 rangefinder measurements. This 

 required a 360 degree look at 

 each iceberg; measuring and 

 photographing all prominent faces. 

 Measurements were accurate to 

 +/- 8% of the observed dimen- 

 sions; see Table F-1. 



The cutter circled each iceberg 

 twice; once to identify the promi- 

 nent faces; and during the second 

 pass, to make measurements. 

 When perpendicular to each face 

 the true bearing and laser-derived 

 distance to it were recorded, a 



129 



