Nistad/Westfeld: Fusion of Measured and Synthetic Sound Speed Profiles ... 
a weighted mean of the measured profiles. The weights are customizable expo- 
nential functions that account for the distance (in both space and time) between 
each measured profile and synthetic profile. 
In order to quantify the depth accuracy of soundings ray-traced using syn- 
thetic profiles, a traditional ray-tracing technique following the so-called “layer 
cake model” was used. Given two sound speed profiles (measured or synthetic), 
a chosen reference depth and beam launch angle, the differing ray-traced solu- 
tions between the two profiles is measurable as a vertical and horizontal bias in 
the ray-tracing plane and is indicative of the relative impact of using one profile 
versus the other for the determination of the 3D sounding position. This assess- 
ment can be performed without or with measured soundings. In this investiga- 
tion, the 65° beam angle was used and the desired maximum vertical bias was 
fixed at 0.25 % of depth. 
3 Results and Discussion 
The accuracy of synthetic sound speed profiles derived from BSHcmod was eval- 
uated by comparing the ray-tracing solution of these profiles against the meas- 
ured sound speed profiles from the MVP-30. For two of the three investigated 
datasets, the distribution of vertical biases fully exceeds the 0.25 % of depth lim- 
it. These results indicate that the evaluated BSHcmod version is not adequate to 
be used in production for ray-tracing purposes. However, BSHcmod has been in 
operational use since the early 1990s and was in the process of being renewed as 
Depth Bias Distribution 
{Dataset B}Y 
5 
0.157 — 
0.10 | 
0.05 — 
D.0r " 
Single Survey Line Statistics 
(Dataset B) 
ST 
— Nearest-in-time 
= = Interpolation 
7 
3.05 - 
& 104 
we 4.0 0.2 
Percentage of Water Depth [%w.d.] 
a 0.4 
0.3 
2.2 
„1 
D.1 EN 
D0-— 
-20 y 
Beam Anagle [deg.ı 
1 
F 
Z 
zz 
50 
Fig. 2: Left: Histogram of depth biases incurred by ray-tracing at the 65° beam angle with 
a subset of the measured MVP-30 sound speed profiles with (in orange) and without 
(in green) interpolation. Right: Mean depth bias and standard deviation as a function 
of beam angle for a single survey line incurred by ray-tracing a subset of the measured 
MVP-30 sound speed profiles with (in blue) and without (in red) interpolation 
u MilsStier-Velrıa 
3and 102/207) a 
DVW-SCHRIFTENREIHE 
62