\section{Benthic Organism Detection}\label{benthic-organism-detection}

In repeated mapping, we are comparing two datasets collected by the same
instruments at the same location, varying the time between collection.
In the following example we describe a method in which differences are
observed between the two survey tools, over the course of a single
survey. In this comparison, the physical differences of the sensors
create two observations that are radically different.

Figure \ref{fig:surcrest}a shows the resolved topology of the area as
taken by the ROV using multibeam sonar. The 5-cm resolution survey
depicts the rocky, high-relief environment that is characteristic of
this type of ocean feature. In Figure \ref{fig:surcrest}b, a similar set
of data are shown in a similarly displayed topographic model. By
subtracting the multibeam data from the lidar data, we are presented
with a topographic map depicting large differences in Figure
\ref{fig:surcrest}c.

\begin{figure*}[htbp]
\centering
\includegraphics[width=0.98000\textwidth]{./img/surcrest.jpg}
\caption{Topographic representations of a section of the Sur Ridge
crest, as obtained by (a) multibeam sonar, (b) lidar, (c) subtraction of
lidar from sonar, and (d) photomosaic. \label{fig:surcrest}}
\end{figure*}

These differences are not noise in the system, but can be attributed to
``soft'' organisms. These animals have a similar acoustic impedance to
the surrounding water mass, which results in acoustic transparency to
the bottom detection algorithms of the 400-kHz multibeam sonar.
Optically, however, these organisms are not transparent at all. These
detections can be confirmed by inspecting the image mosaic collected
concurrently with the lidar and sonar data (Figure \ref{fig:surcrest}d).
Detailed inspection of targets in the photo-mosaic also can give insight
into the health of the sponge by examining its color. In Figure
\ref{fig:surcrest}d, healthy sponges are seen as white, and less healthy
or potentially dead sponges are seen to be brown.