Utilizing multispectral lidar in the detection of declined trees

Tutkimustuotos: Doctoral ThesisCollection of Articles

Abstrakti

The World’s forests are facing novel stress due to climate change. Pest insects and pathogens are shifting towards new latitudes and heat stress is resulting in increased tree mortality and more frequent forest fires globally. Uncertainty in estimating the magnitude of climate change induced forest and tree decline requires new methods for unbiased estimation of tree decline. The development of remote sensing methods to detect early tree decline has been a major challenge due to the subtle nature of the early changes caused by different stressors. Multispectral lidar technology has the potential of detecting early tree decline by providing accurate three-dimensional and spectral information of tree structure simultaneously. The main objective of this thesis was to investigate the capabilities of multispectral terrestrial lidar in the detection and assessment of tree decline caused by different stressors. This was done by investigating the estimation of a remotely detectable indicator of tree decline, leaf water content (LWC). Specifically, new methods for measuring LWC using multispectral lidar intensity were developed from the leaf to the canopy scale in various environments and the relationship between LWC and tree decline induced by various stressors was investigated. Furthermore, the developed methods were tested in a forest environment to assess the applicability of multispectral lidar in the detection of bark beetle infestation in the field. Studies I-III focused on investigating the relationship between LWC and lidar intensity at multiple wavelengths. First, a hyperspectral lidar instrument was used to detect significant changes between fresh and drought-treated Scots pine (Pinus sylvestris L.) and Norway spruce (Picea abies L.) trees (study I). Then, a leaf-scale study (II) with Scots pine, Norway spruce, Small-leaved lime (Tilia cordata L.), Norway maple (Acer platanoides L.) and Silver birch (Betula pendula L.) was conducted and a strong relationship (R2=0.93) between a normalized difference index (NDI) calculated from 1550 nm and 690 nm wavelengths and LWC was found. This was followed by a study (III) where LWC estimation and pathogen- and drought-induced variation in LWC was studied with Norway spruce seedlings. Blue-stain fungi (Endoconidiophora polonica) inoculated seedlings expressed a rapid decrease in LWC while drought-treated seedlings showed more stable LWC until a very severe drought. LWC of the seedlings was predicted with an R2 of 0.89 using an NDI with 1550 nm and 905 nm wavelengths. In study IV, the developed method and the relationship between LWC and tree decline was investigated in the field with European spruce bark beetle (Ips typographus L.) infested trees. It was found that of the LWC metrics studied, gravimetric water content showed significant differences in the early stages of infestation and was more sensitive to bark beetle induced tree decline than equivalent water thickness (i.e. amount of water per leaf area). Linear discriminant models that were developed between infestation severity and lidar intensity metrics from 1550 nm and 905 nm wavelengths showed that green attack stage of the infestation could classified with an overall accuracy of 90%. This dissertation contributes both to the development of an objective and automatable method for detecting and measuring tree decline in the field, and to the understanding of the relationship between LWC and tree decline with implications to remote sensing. The dissertation will be published and popularized as a music video here: http://bit.ly/idanproffa.
AlkuperäiskieliEnglanti
PätevyysTohtorintutkinto
Myöntävä instituutio
  • University of Helsinki
Kustantaja
Painoksen ISBN978-951-651-645-8
Sähköinen ISBN978-951-651-644-1
TilaJulkaistu - 31 toukokuuta 2019
OKM-julkaisutyyppiG5 Tohtorinväitöskirja (artikkeli)

Sormenjälki

Sukella tutkimusaiheisiin 'Utilizing multispectral lidar in the detection of declined trees'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä