Date:
2018/06/12

Time:
16:00

Room:
A2 Wivi


Quick LiDAR-based characterization of forest vertical structure to support forest management actions

(Oral)

Ihor Semko
,
Anna Kozlova
,
Mykhailo Popov
,
Ihor Kozak

SEE PEER REVIEW


Vertical arrangement of vegetation components is important for understanding forest environment, being both a product and driver of ecosystem processes and biological diversity. Features of stand level spatial distribution provides key information on forest ecosystems’ composition, functioning, and dynamics. Characterization of vertical structure and its complexity helps to assess forest stand response to various disturbances and clarify how current management affects biodiversity. Recent advantages in applying LiDAR data for estimating three dimensional forest stand structure across a range of scales gives novel informational tools for planning and assessing actions on forest management and conservation. Existing techniques often require prior information about stand characteristics or rely on pre-defined height and diameter thresholds.

In the study, we present an approach to characterize vertical forest structure based on echo distributions from airborne laser scanning (ALS). We developed a method using ALS data combined with a statistical estimation of vegetation components’ density to determinate canopy layers spatial distribution. The method allows us to obtain a morphometric characteristic of each layer, to determine the vertical complexity of a forest stand, and to derive canopy structure types. Since the developed method does not require ground reference data for calculations, the result can be obtained much faster than using other similar methods.

The proposed approach was tested at Zemborzyce study site (E22˚31'45" N51˚09'38"), located within a deciduous mixed forest in the south east of Lublin, Poland. These forests has a complex structure, up to four layers: mature dominant species and a slightly lower layer of mature trees, a shrub layer, and understory layer of grasses and other herbaceous plants. A LiDAR dataset flown in 2014 was obtained from The John Paul II Catholic University of Lublin. The characterization of forest vertical structure at the study site is resulted in a set of maps representing vegetation density at each canopy layer as well as fifteen canopy structure types describing the forest vertical complexity of the study site. Obtained maps were validated using forest inventory data and showed promising results. We conclude that our method gives a new robust and reliable approach to characterization of forest vertical structure and enables an efficient monitoring of canopy structure, making possible a quick respond to sudden disturbances such as fires and windfalls.


SEE PEER REVIEW