- 62 views
- 130 downloads
Soil Moisture and Hydraulic Parameter Estimation and Remote Sensing for Precision Irrigation
-
- Author / Creator
- Orouskhani, Erfan
-
Water scarcities are becoming serious issues worldwide primarily due to population growth, climate change, and increasing pollution. Since a large portion of freshwater is consumed in agricultural activities, with the main consumer being irrigation, increasing the water-use efficiency in irrigation through precision irrigation is a critical step toward reducing the freshwater shortage. Precision irrigation can be achieved by forming a closed-loop irrigation system that closes the irrigation decision loop. To implement the closed-loop irrigation, soil moisture information that is required for feedback control must be provided. In order to obtain the soil moisture information of the entire field, the soil moisture estimation techniques based on the measurements of a small number of sensors are proposed. While soil moisture estimation in agricultural fields is possible, a few main challenges yet exist.
Firstly, because of the limited number of available sensors in the agricultural fields, a major challenge is to identify the optimal location of the sensors in the soil so that improved state estimation can be achieved.
Additionally, accurate quantification of soil hydraulic parameters, which are crucial for developing an agro-hydrological model and affect its accuracy, is essential for estimating soil moisture. Another challenge associated with soil moisture estimation is converting the remote sensing data into soil moisture that can provide the soil moisture measurements for a large region of the agricultural field.
The purpose of this thesis is to find solutions to the aforementioned challenges, resulting in a comprehensive soil moisture estimation method that can be used for closed-loop irrigation. Firstly, we describe an actual agricultural field in Lethbridge, Canada studied in this thesis.
A three-dimensional agro-hydrological framework is then developed to model the actual field. We specifically use the cylindrical coordinates version of the Richards equation to model a field equipped with a center pivot irrigation system. The heterogeneous distribution of the soil parameters is considered in the Richards model.
The modal degree of observability is then applied to the 3D field model to determine the optimal sensor locations in the actual field.
The extended Kalman filter is also employed to estimate the soil moisture content of the actual field using the real measurements obtained from the point sensors.
The estimation results are then analyzed to investigate the effects of sensor placement on the performance of soil moisture estimation in the actual applications. To address the second challenge, we propose a systematic estimation approach to simultaneously estimate the soil moisture and soil hydraulic parameters in the 3D agro-hydrological systems with spatially heterogeneous soil parameters.
In this part, microwave remote sensors that are mounted on the center pivots are considered to provide the rotating measurements.
The sensitivity analysis is employed to determine the most important subset of soil hydraulic parameters for estimation.
Another feature of the proposed method is using the Kriging interpolation method for updating the rest parameters that are not estimated.
The proposed method is applied to two different simulated three-dimensional fields, and the simulation results of the considered agro-hydrological systems illustrate the applicability and effectiveness of the proposed method on the performance of soil moisture estimation.
Further, the algorithm for surface soil moisture estimation using the thermal and optical remote sensing images is proposed. The machine learning-based Multilayer Perceptron (MLP) model is developed to convert the multispectral images to soil moisture.
The developed model is applied to the real agricultural field in Lethbridge and is trained using the experimental data collected in the summer of 2019.
The results demonstrate a strong agreement between the measured soil moisture and predicted soil moisture from the MLP model.
Throughout this thesis, we demonstrate how the proposed solutions can be used to effectively address the challenges discussed earlier. -
- Graduation date
- Fall 2022
-
- Type of Item
- Thesis
-
- Degree
- Master of Science
-
- License
- This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.