Preprocessing and Analysis Strategies 
for Hydrographic Measurements at 
Very Shallow Water Depths 
Bahareh Mohammadivojdan 
Robert Weiss | Thomas Artz 
| Frederic Hake | Hamza Alkhatib | Ingo Neumann 
1 Introduction 
Digital Terrain Models (DTMs) related to underwater or water exchange zones 
are an important source of information for various applications. The develop- 
ment of different methodologies including range based acoustic and optical 
scanning techniques, enable the possibility of attaining high-resolution maps of 
these areas. In particular, the surveying of shallow water areas is a challenge even 
today. Deep and shallow water area requires different measuring systems and 
their results must be fused in a meaning way. In order to improve the processes 
for the determination of DTMs in very shallow areas of waterways, a project 
was initiated by the Federal Institute of Hydrology (Bundesanstalt für Gewäs- 
serkunde, BfG) and Geodetic Institute of the Leibniz University of Hannover 
(GIH). The very shallow or water exchange areas are highly sensitive since the 
classical measurement methods including Airborne Laser Scanning (ALS), ba- 
thymetry (ALB), and ship-borne echo sounder measurements are not usually 
suitable to sample the surface of interest due to coordinative, technical, and safe- 
ty-relevant aspects. 
Although the mapped areas are usually of high resolution, they are only a 
discrete sample from the region of interest derived from a point cloud. These 
captured point clouds, especially due to the rough environment, are typically 
corrupted with noise and outliers. In such data sets, it should be considered that 
the sampled point clouds might be unevenly distributed over the domain. Also, 
due to the surface geometry and possible measurement failures, there are areas 
with less point density or no measurements at all (data gaps). In water exchange 
areas these limitations are even more pronounced, due to the technical limita- 
tions, causing significant data gaps, which are usually compensated by interpola- 
tion to derive comprehensive models. 
The solution to overcoming the aforementioned challenges and merging data 
from different platforms and sensor techniques is to find a unique mathematical 
model to represent such point clouds. Achieving a continuous surface from this 
U WilSter-Velıa 
3and 102/207) a 
DVW-SCHRIFTENREIHE 
5% 
3