implementing new vegetation / forestry metric: rumple

[Martin Isenburg]

Hello,

a few days ago I had another user request a metric called 'rumple' to be implemented in LAStools. This is not the first time and I have been toying with the idea since I learned about the metric which I believe was at the SilviLaser 2015 conference in La Grande Motte, France.

The definition of the rumple index is the surface area of the canopy divided by the surface area of the terrain. But where to measure this? After the DSM and DTM were rastered onto a grid? Or before when the  DSM and DTM are still in their original interpolation (like via a TIN).  When it's derived from a DSM and DTM raster it should become dependent on the cell grid cell size. When it's done on the TIN domain this effect should be smaller. But should we use a spike-free TIN or a pit-free raster? How the DSM is exactly computed (and there are many ways, see [1] and [2]) ought to have a huge impact on the metric.

Thoughts?

Martin @rapidlasso

[1] https://rapidlasso.com/2014/11/04/rasterizing-perfect-canopy-height-models-from-lidar/

[2] https://rapidlasso.com/2016/02/03/generating-spike-free-digital-surface-models-from-lidar/

[Andre Wise]

HI Martin,

This should be a useful metric that has recently come to our attention too. My thoughts are similar to yours, the calculation of the metric done as close to the point cloud data as possible i.e before DSM generation. Our initial thinking is that what we're after is a metric that describes the shape of the canopy of the individual trees as well as possible. The higher the surface area, the steeper the canopy and the opposite for a lower surface area. This means that we want to exclude spikes but identify as much of the real canopy shape as possible. Rumple should help with BA estimates as well as stem count adjustments (a higher surface area could be an indication of missing trees).

Thanks

Andre

[Steven F]

+1. I would also like to see rumple available in LAStools.

In some other software such as FUSION rumple is calculated from the canopy height model and digital terrain model rasters (surface area ratio in GridSurfaceStats). However, I agree with Andre that using the point cloud is likely to better represent canopy complexity than you can from the rasters. Considering some lastools can read rasters such as BIN files as point clouds perhaps you could let the user choose by offering options for the type of input (point cloud, raster or TIN). This could also grant the user some flexibility in choosing filtering options before calculating the statistic (e.g. run lasthin or spike-free chm, etc.).