Generation of a ground-level DEM in a dense equatorial forest zone by merging airborne laser data and a top-of-canopy DEM

(i) selecting ground points from the airborne data and deducing a pseudo canopy height for these points, (ii) characterizing the canopy height from a statistical and geostatistical standpoint, (iii) kriging the canopy height and subtracting the resulting model from the top-of-canopy DEM to obtain a DEM corresponding to a ground-level DEM, and (iv) validation. Validation consists of comparing the results with topographic maps and a local heliborne DEM and studying the relationship between the configuration of the airborne trials, in terms of flight-line spacing, and accuracy of the resulting kriged DEM. The results show that the accuracy of the kriged ground-level DEM is significantly better than that of the initial radargrammetric DEM. The standard deviation of elevation errors is reduced from 21.2 to 11.9 m or from 25.3 to 11.1 m, depending on the validation source adopted (French National Geographic Institute spot heights and local heliborne DEM, respectively). In addition, the relationship between flight-line spacing and accuracy of the resulting kriged DEM helps estimate what flight-line spacing is needed to obtain a given accuracy.