Grid Creation Options - Specifics

ZFehr

Hello!

I am working on a bathymetric photogrammetric problem and am attempting to convert an extremely high-density las dataset to a 1-m DEM for comparisons to 1-m lidar DEM products generally available to the public and to explore change analyses across sensing technologies. The grid generation capabilities outlined in the 'Grid Generation Options' allow the user to use triangulation as a method of grid generation.

My question is this, in an event where there are many triangles in the TIN that fall within a 1 x 1m grid cell, what method does GM employ to generate the predicted value? Other software allow for linear and natural neighbor interpolation at the TIN level when determining the final raster value, in which a cell is determined from the planar value of the triangle or an assessment of nearest nodes to the cell center. Is GM looking at the underlying triangle relative to the position of the cell center or is it assessing all triangles within the user-defined resolution (and potentially no-data criteria)?

Alongside this, when determining the no data distance criteria, is this sliding scale in the same units as the current workspace projection? Meaning working with data in UTM would have sliding scale units of meters? If this is the case, is this no data distance criteria a radius about the output raster cell center? With a minimum value of 0.5, would this mean that las data in the corners of a 1 x 1m grid are not being considered in the final raster output? Or, does the no-data distance determine lengths for TIN generation?

Below is a screengrab showing a 1-m cell (blue regime) and the TIN as a vector layer to display the las density against the target grid resolution for context.

Thanks in advance for any information regarding these options,

Zach