Mohammadivojdan et al.: Preprocessing and Analysis Strategies ... 
"u 
Ka 
Sb 
5 
0 
‚OS 
a 
ET 
fol 
WC 
(el 
‚06250 
1, 
A 
By 
SEE 
STE 
HE 
Ye 
A 
Fig. 1: Approximated sample data set of Kiel Canal. The largest deviations occur at the gaps. 
means that there is not a significant systematic deviation between the derived 
model and the observations. As an error measure, the Root Mean Squared Error 
(RMSE) of the deviations is calculated, which is 0.045 m. 
2.3 Evaluation 
In order to evaluate the adapted methodologies, derive an uncertainty measure 
and describe the approximated model, a simulation is designed. In this simulat- 
ed environment, the real measured data are imitated and the associated uncer- 
tainties are calculated. To achieve this, the physical and mathematical models 
of the sensors must be implemented in the environment. It is then possible to 
generate the measuring beams from the sensor and intersect them with a math- 
ematically defined surface. The result is an error-free, point-wise representation 
of the mathematical surface. Fig. 2 shows an example of one profile of a multi- 
beam echo sounder with the measurement beams in blue and the mathematical 
surface in brown. 
3 Conclusion 
In this project, a pipeline of methods is proposed for modelling the underwater 
observations. The main idea is to improve the quality of the final model as well 
as the efficiency of the final algorithm. Also, the final methodology is claimed to 
be fully automated, which helps to reduce the manual work of users to a min- 
imum. Consequently, a unique mathematical model of the measurements can 
U Wilstier-Velıa 
3and 102/2022 @® DVW-SCHRIFTENREIHE 
5