Selection of Standard Range for LiDAR Intensity Range Correction

Question

When correcting LiDAR intensity values for the effects of range (distance from sensor to target) many methods, including the one supplied by the lidR package (https://rdrr.io/cran/lidR/man/lasrangecorrection.html), involve supplying a standard range. Is there any rationale behind choosing a standard range value? Should I supply the mean flight height? Or should I choose something simple like 1000 meters?

Answer 1

The method relies an a model driven correction. Making some assumptions on how energy is propagated, attenuated in atmosphere and back-scattered the intensity correction is:

I' = I*(R/Rs)^f

Where I' is the corrected intensity, I the raw intensity, R the range, f an empirical value close to 2 and Rs a range of reference.

The range of reference is needed because you want to normalize relatively to something. Indeed there is no absolute intensity values it depends on the altitude of the sensor. So you can normalize as if the sensor was at 3000 m with Rs = 3000.

But the best is indeed to provide a reference that is close to the actual sensor elevation. The mean flight height (such as the one given by the data provider) is good because your intensity values will stay in the same range of values than the raw values.

Also you must understand that this correction is relative to the raw data. It is not an absolute correction. It corrects the intensities for variation of distance between the sensor and the targets (because of topography mainly but also flying altitude variations). But it does not correct intensities for emitted energy, sensor sensibility, sensor gain and so on. If two datasets are sampled at the same altitude but with different energy emitted, the intensities will be different but the range correction does not account for that. Usually we know nothing about energy emitted, gain of the sensor and so on so we can't account for that with model-driven methods.