Water content estimation of conifer needles using leaf-level hyperspectral data

Area of Science:

• Plant physiology
• Remote sensing
• Forestry

Background:

• Plant water content is vital for growth and stress monitoring.
• Accurate estimation aids in understanding vegetation health.
• Conifers exhibit diverse needle morphologies impacting water dynamics.

Purpose of the Study:

• To construct water loss curves for three conifer species.
• To evaluate common water indices for conifer needle water content estimation.
• To explore hyperspectral data for estimating leaf water content in conifers with varying needle morphology.

Main Methods:

• Water loss curves were generated for three conifer species.
• Twelve water indices were assessed for their applicability.
• Partial Least Squares Regression (PLSR) modeling was employed using hyperspectral data.

Main Results:

• Olgan larch exhibited significantly higher water loss rates compared to Chinese fir and Korean pine.
• Reflectance changes were most pronounced in the Near-Infrared (NIR) and Short-Wave Infrared (SWIR) regions.
• SWIR bands were identified as most sensitive to conifer needle water content.
• Water indices were effective for single species but not universally applicable.
• The PLSR model accurately estimated water content across all tested conifer morphologies.

Conclusions:

• Hyperspectral data combined with PLSR is a robust method for estimating conifer needle water content.
• This approach holds promise for monitoring vegetation water status in diverse conifer species.
• Understanding species-specific water loss dynamics is crucial for effective vegetation management.