Bridging the Gap Between Lidar, Thermal, and Multispectral Remote Sensing for Irrigation Scheduling Applications

Irrigation reduces the soil moisture deficit in crop production; however, the Canadian Prairies is a semi-arid landscape with limited water resource availability, requiring careful application of water management practices. This thesis studies methods to reduce irrigation water consumption in agricultural fields with complex soil texture dynamics using the water deficit index (WDI) to indirectly detect crop water stress and measure root zone soil moisture. This index is an extension of the crop water stress index (CWSI) that uses remotely sensed surface temperature (Ts) in addition to the fraction of vegetation (fc) to estimate the crop evaporative fraction through manipulation of the available energy balance equation. Seasonal and spatial relationships between WDI and volumetric water content (VWC) over a wheat and pea crop were observed at a study site with heterogeneous soil textures over two growing seasons; wheat was planted in the first growing season, and pea was planted in the second. Ten ground-based stations were used to observe average daily fluctuations in WDI by measuring Ts and deriving fc using the normalized difference vegetation index (NDVI). Results indicated that deep layers of sandier soils are more likely to cause high variations of WDI during dry-down events. Remotely sensed WDI agrees with measured eddy covariant energy fluxes at the beginning to middle of the growing season; however, NDVI is impacted by leaf senescence after seed fill for both crops, reducing the accuracy of WDI later in the growing season because of errors in fc. Light detection and ranging (lidar) is introduced as a more sophisticated approach to obtain fc and is used as a method to validate WDI obtained using NDVI canopy fraction using unpiloted aerial vehicle (UAV) imagery during the pea growing season. Canopy fraction obtained using NDVI UAV imagery produced WDI values that agreed with canopy fraction derived using lidar demonstrating that NDVI provides accurate fc for the calculation of WDI. A technical analysis was performed to assess the accuracy of crop height models obtained using Structure from Motion (SfM) photogrammetry techniques compared to lidar. Photogrammetry crop height models were obtained using high-quality red-green-blue (RGB) imagery with accurate real-time kinematic (RTK) positioning, or RGB, multispectral and thermal imagery georeferenced using 3D ground control points (3D-GCPs); thermal and RGB SfM crop height models georeferenced using 3D-GCPs were inaccurate when compared to lidar crop heights. Further analysis was performed on identifying the empirical relationship that existed between lidar-derived fc and crop height for wheat and pea crops. The ability to track seasonal and spatial relationships between WDI and VWC, and the ability to obtain crop height models using multispectral imagery provides exciting progress at bridging the gap between thermal, multispectral and lidar remote sensing for irrigation scheduling applications.

Name thesis:degree_name

Master of Science (M.Sc.)

Level thesis:degree_level

Masters

Discipline thesis:degree_discipline

Civil Engineering

Grantor

University of Saskatchewan

Year dc:date.issued

2024

Author dc:creator

• Cline, Emily G

Advisor dc:contributor.advisor

• Helgason, Warren D

Committee members dc:contributor.committeemember

• Bueckert, Rosalind A
• Noble, Scott D
• Derdall, Evan
• Smith, Laura

• Irrigation Scheduling
• Remote Sensing
• Evapotranspiration
• Soil Moisture

Language dc:language.iso

en

Handle dc:identifier.uri

https://hdl.handle.net/10388/15868

OAI identifier oai:identifier

oai:harvest.usask.ca:10388/15868