Detecting Changes in Grass Physiological Properties under Drought Stress Using Hyperspectral Remote Sensing

Authors: Phuong D. Dao*, University of Toronto, Department of Geography, Yuhong He, University of Toronto, Department of Geography, Bing Lu, University of Toronto, Department of Geography
Topics: Remote Sensing, Physical Geography, Hazards, Risks, and Disasters
Keywords: Plant physiology, chlorophyll content, water content, chlorophyll fluorescence, grassland drought, hyperspectral remote sensing
Session Type: Paper
Day: 4/10/2018
Start / End Time: 8:00 AM / 9:40 AM
Room: Zulu, Sheraton, 8th Floor
Presentation File: No File Uploaded


This study examines the change in grass physiological properties under drought stress at an old field located at Koffler Scientific Reserve, a biological site of the University of Toronto. We create artificial drought stress on grasses in field sites and greenhouse experiments and explore how grasses respond to drought with different intensities including early, moderate, severe, and extreme droughts. For the field experiment, three 4m x 4m rain-exclusion shelters, equipped with real-time soil moisture sensors, were built along an elevation gradient to create an artificial drought in summer, 2017. In parallel, a greenhouse drought experiment was also conducted on grass samples collected from the same study area. A total of 12 grass patches were grown under four water treatments with 100%, 75%, 50%, and 25% of field precipitation, with three replicates in each treatment. A number of air-borne, ground-based, and laboratory measurements were taken including airborne hyperspectral images, soil moisture, leaf area index, species composition, biomass, leaf and field reflectance, chlorophyll and water content, and chlorophyll fluorescence to examine drought effects. Hyperspectral remote sensing with hundreds of contiguous narrow spectral bands, which provides insight into how vegetation reflectance changes in response to drought, was used to investigate the change of grass physiological properties at the leaf, canopy, and landscape levels. The results from this work indicate strong correlations between hyperspectral indices and chlorophyll content, water content, and chlorophyll fluorescence, demonstrating the ability of hyperspectral data for detecting the subtle changes in these grass physiological properties under different drought stress intensities

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login