Remote Sensing for Forest Inventories
The project "Innovative Methods for the Measurements of Forest Structures" was setup to evaluate the accuracy of airborne remote sensing sensors. For this purpose an experimental area in the national park “Bayerischer Wald” (Bavarian Forest) was chosen, presenting a number of different forest types and structures. By terrestrial inventory techniques data were acquired for the area as a reference for the results of interferometric radar, LIDAR (Light Detecting and Ranging Sensors) and conventional analog and digital aerial photographs. LIDAR- and digital cameras showed the best results.
The used LIDAR sensors were able to take 80.000 distance measurements per second This equals 10 to 20 measurements per square meter with an accuracy of +/- 15 cm. The fact that LIDAR radiation penetrates forest canopies opens up unknown possibilities. For instance, understory vegetation can be detected using LIDAR sensors. Even regeneration made up of conifers or hardwoods can be differentiated when measuring in summer and winter time.
Digital aerial photographs showed advantages in comparison with conventional analog ones. Besides their better resolution their advantage is the possibility to register important features automatically. For example, for an automated single tree detection the crowns can be identified by special algorithms. And with the help of mathematical functions diameters at breast height (dbh) and the volume of trees can be approximated.
The comparison of the results from airborne and terrestrial inventories showed, that airborne forest inventories can be operated with high accuracy. An essential advantage of airborne techniques is that large areas can be measured at relatively low costs.