Assignment LH-09
Lt.Arya Meiko W.
20699/P
Shoalest Depth True Position
This surface stores the shallowest depth found within each node of the depth layer. It can be used
to generate a true position surface, with optional layers for horizontal and vertical uncertainty if such data
is available.
If uncertainty information is not present for the selected lines, the surface can still be created. Later, if
uncertainty data becomes available, the surface can be recalculated and updated accordingly.
CUBE (Combined Uncertainty and Bathymetry Estimator)
There are two versions of the CUBE algorithm: version 1 and version 2. Depending on the process
used, they are identified differently. For instance, the New Regular Gridded Surface process supports both
versions. Version 1 is simply called “CUBE,” while version 2 is referred to as “CUBE v2.”
In contrast, all other processes ncluding the New Variable Resolution Surface and any workflow within
CARIS Bathy DataBASE Editor use only version 2 of the algorithm, and they refer to it as “CUBE.”
The CUBE (Combined Uncertainty and Bathymetry Estimator) algorithm creates a surface
composed of multiple depth hypotheses that represent possible variations in seafloor elevation.
These hypotheses can be reviewed in the Subset Editor to check their validity and replaced if necessary.
When a surface format that retains metadata is used, this information remains available even after
generating a new CUBE surface. This allows CUBE to be applied iteratively, helping achieve a surface
model that best represents the actual surveyed area.
.
Swath Angle Weighting
The influence of a sounding on the final surface depends on the grazing angle it makes with the
seafloor. This weighting is especially useful in areas where survey lines overlap or run closely together. The
swath angle weighting gives more importance to soundings from the center of the swath compared to those
near the outer edges from adjacent lines.
In the graph, soundings with grazing angles between 90° and 75° are assigned a full weight of 1.0.
As the grazing angle becomes smaller, the weight gradually decreases—down to 0.01 when the angle
reaches 15°.
Uncertainty Weight
The influence a sounding has on a node is determined by how certain the depth measurement is—
less uncertainty means a stronger influence. In other words, a sounding with lower predicted depth
uncertainty contributes more weight to a node.
This uncertainty is calculated based on how far the sounding is from the node, taking into account
both vertical (depth) and horizontal uncertainty. Therefore, both types of uncertainty influence how much a
sounding affects a node.
The area of influence defines how many nodes can be affected by a single sounding. This is
calculated using the selected IHO S-44 survey order's uncertainty formula:
Once this uncertainty exceeds the limit set by the chosen S-44 survey order, the sounding no longer
influences additional nodes—this marks the limit of its influence.
All calculations are made using ground coordinates, not just simplified distances. Every node within
the area of influence is considered precisely—no approximate or simplified matrix is used.
New Variable Resolution Surface
Source data often contains areas with uneven point density due to changes in depth, survey paths,
or data coverage. Fixed-resolution grids can struggle to handle this, since choosing a single resolution may
either lose detail or waste resources.
Variable Resolution (VR) surfaces allow different areas of the dataset to have different resolutions
while still ensuring overall consistency. Rather than assigning a single resolution value, resolution is
computed per region using specific algorithms.
To create a VR surface, one must first decide how to divide the area—commonly using track lines,
raster surfaces, or point clouds—and determine the required resolution per tile or region. A parameter (such
as data density) can be used to guide this. If multiple sources are involved, the system treats them as a
single combined source.
Note: When source data overlaps, all points are considered during processing, regardless of whether values
differ.
The area is then divided into regular tiles, often using a quad-tree structure (binary spatial
partitioning), until each tile has enough data.
For each tile, the resolution is calculated using one of the following three methods:
Estimation Method
1. CARIS Density
Estimates resolution from source point density using a binning-based algorithm. Avoids skewed
estimates from clusters, making it useful for multibeam/single beam datasets with highly variable
depth.
2. Calder-Rice Density
Estimates resolution based on point density over a set tile size. Designed for raw multibeam data
with relatively small depth variation. Runs faster by using more predefined parameters.
3. Ranges
Assigns resolution using statistical measures (min, max, mean) and lookup tables. Three
statistical options:
• Mean: Uses the average value in a tile.
• Mode: Uses the most frequent value.
• Percentile: Based on a percentile between deepest and shallowest points (0 = deepest,
100 = shallowest). For non-elevation values, 0 = highest, 100 = lowest.
Population Method
1. Inverse Distance Weighting (IDW)
• Depth is calculated as a weighted average of all nearby samples, where weights are
based on the inverse of the distance between the sample and the node.
• Closer points have more influence.
2. CUBE
• Multiple depth hypotheses are calculated based on depth and uncertainty. The strongest
hypothesis is selected.
• Useful for automated processing that handles outliers and uncertainty.
3. Uncertainty
• Depth is the average of all samples in the resolution bin, weighted by both distance and
sample uncertainty.
4. Mean
• Uses the average depth value of all samples within the tile or bin.
5. Swath Angle
• Considers beam angle and footprint radius to define the maximum area of influence for
each sounding point.
6. Selected Value
• A single value is selected for each node (e.g., minimum, maximum, or median),
depending on the criteria set in the options.
New Regular Grid Surface
Shoalest Depth 1m Cube 1m
Swath Angle 1m Uncertainty 1m
New Variable Resolution Surface – Calder Rice Estimation
CalderRiceEstimation_CubePopulation CalderRiceEstimation_IDWPopulation
CalderRiceEstimation_MeanPopulation CalderRiceEstimation_SelectPopulation
CalderRiceEstimation_SwathPopulation CalderRiceEstimation_UncertaintyPopulation
New Variable Resolution Surface – Caris Density Estimation
CarisDensityEstimation_CubePopulation CarisDensityEstimation_IDWPopulation
CarisDensityEstimation_MeanPopulation CarisDensityEstimation_SelectPopulation
CarisDensityEstimation_SwathPopulation CarisDensityEstimation_UncertaintyPopulation
New Variable Resolution Surface – Ranges Estimation
RangeEstimation_CubePopulation RangeEstimation_IDWPopulation
RangeEstimation_MeanPopulation RangeEstimation_SelectPopulation
RangeEstimation_SwathPopulation RangeEstimation_UncertaintyPopulation
Depth Compare New Regular Grid Surface Depth Compare Calder Rice Estimation
Surface
Depth Compare Caris Density Estimation Depth Compare Caris Ranges Estimation
