Lidar DEMs for Forestry

DEMs from the lidar survey provide a significant resource for foresty.  The analysis can use either a DSM or a CHM for difference maps; for measurements, the CHM will be easier since it has already removed the DTM bare earth elevations.

There has been a lot of work to extract trees, and most of that work has used the grid rather than directly use the point cloud.  Much of the work uses standard image segmentation techniques.

Example 1, Coniferous forest post fire survey

DSM from summer survey in New Mexico.
DSM from late spring survey, two years after the survey above, after a fire.

Differences in the two maps appear visible.  The trees in portions of the area appear smaller and more open.
Differnce maps, PostFire-PreFire.

Negative numbers show a loss of vegetation.  There are very few large positive numbers apparent at this scale.
Histogram of the difference grid.  There are some positive values (vegetation growth in the two year period is a likely explanation), but many more negative values from the fire.



Example 2, Deciduous forest after leaf growth


April survey, before leaf growth.

Power line on the right side.

Several large buildings in clearings.
July survey, after leaf growth.
Difference map.  Most of the difference is large positive numbers, particularly around clearings.

Many large negative values on the east side of the map, where the power lines apparently moved slightly between the two surveys.
Histogram of the difference map, showing many more positive values.

last revision 12/7/2017