| dc.description.abstract | Subalpine wetlands in the Canadian Rocky Mountains function as buffers for snowmelt runoff towards downstream systems and communities. However, hydrological regimes are changing rapidly in these ecosystems due to climate warming in the region causing earlier snowmelt patterns and extended vegetation growth periods. Response of vegetation health and composition to temperature and precipitation increases can have a significant influence on evapotranspiration (ET), the main component of wetland water balances. Subalpine plant communities are especially sensitive to shading mechanisms over the growing season, which limits ET flux. However, as the composition of plant communities are expected to migrate with climate, it is important to monitor changes in primary water sinks such as ET in vulnerable ecosystems, such as subalpine wetlands. Due to difficult accessibility, few studies have been conducted to monitor these ecosystems. Recent technological advances in unmanned aerial vehicles (UAV) provide opportunities to monitor these ecosystems at a high spatial resolution.
This study aims to quantify plant community scale ET, assess the spatial variability and sensitivity of this ET to climate and vegetation health, and evaluate the Mapping Evapotranspiration with Internalized Calibration (METRIC) model for ET estimation in a subalpine wetland. ET was measured in-situ using a dynamic closed chamber method for the plant community scale at Fortress Mountain in Kananaskis, Alberta. Vegetation health, water content, and plant water stress was derived from spectral signatures using vegetation indices. High-resolution imagery with multispectral, thermal, and LiDAR sensors were collected during ground measurements to capture the spatial variability of ET throughout the wetland using the METRIC model. Modelled ET was compared with chamber ET measurements to assess the accuracy and applicability of the METRIC model using UAV imagery in a subalpine wetland.
Net radiation and plant community type were the dominant controls on ET at the community scale. Variability in physiological differences between plant communities, such as depth of stomatal openings, cuticle thickness, leaf surface area to volume ratio, and root water uptake rates affect plant response of ET to radiation and temperature. Plant physiology as well as volumetric water content, proximity to surface water, and groundwater connections, also influenced spatial ET trends.
METRIC model results had high estimation accuracy when compared to chamber results. METRIC ET had strong relationship with hourly (R2=0.79) and daily (R2=0.82) chamber ET. Taller vegetation (trees and shrubs) had higher estimation accuracy than lower-lying vegetation (ground vegetation and moss). Spatial variability of ET using the local indicators of spatial association (LISA) with METRIC results showed clusters of high ET in the Southern and Western sections of the meadow and low ET in the Northern and Eastern sections of the meadow.
The results of this study demonstrate that as plant communities are expected to migrate with changing climate conditions in subalpine ecosystems, METRIC model applications using UAV imagery could be an effective solution to monitoring plant community ET at a high spatial resolution in vulnerable and inaccessible areas. | en |
