(484a) Simultaneous Estimation of Soil Moisture and Hydraulic Parameters for Precision Agriculture: A Real Case Study

Agriculture consumes about 70% of all freshwater withdrawals, the vast majority of which are used for irrigation purposes [1]. As is generally known, the water scarcity crisis is worsening, due to increased stress on freshwater supplies stemming from rapid population growth and climate change. Since agriculture consumes a large portion of freshwater, ensuring the optimal use of water through increased water-use efficiency in agricultural irrigation represents an effective means to mitigate the current water supply crisis. Closed-loop irrigation is a promising approach to improve the water-use efficiency in irrigation and consequently reduce the wastage of water resources. To implement closed-loop irrigation systems, various soil moisture techniques must be employed to provide the required soil moisture information for feedback control. In this regard, microwave remote sensors have been demonstrated to have the ability to measure the near-surface soil moisture quantitatively. Microwave radiometers mounted on center pivot irrigation systems provide a feasible approach to quantitatively measure the soil water information of fields that are equipped with center pivots.

Measurements provided by soil moisture sensing technologies have one key limitation. It is practically infeasible to ensure continuous and long-term observation of soil moisture, thus soil moisture measurements have spatio-temporal gaps. It is necessary to fill in the gaps in soil moisture information provided by these measurements: (i) to make the soil moisture information more complete, and hence more useful; and (ii) to provide soil moisture information at a regular scale to facilitate continuous feedback and control. State estimation techniques can be used to fill in the information gaps present in the soil moisture measurements. At the same time, the models that are employed in soil moisture estimation require soil hydraulic parameters, which are important inputs that affect the model's accuracy in simulating soil moisture content. Hence, an accurate quantification of these parameters is crucial for soil moisture estimation.

Before simultaneous state and parameter estimation of a given model can be performed, it is important to know a priori whether values for the considered set of model parameters can be uniquely identified. Typically, estimation techniques will find it challenging to estimate the parameters of a non-identifiable system and thus will not be able to find reliable estimates for the unidentifiable parameters [2]. In an instance where the system is non-identifiable, it is prudent to determine the elements of the considered parameters that are most relevant to the model predictions for estimation.

In this study, we investigate the dual estimation of soil moisture and soil hydraulic parameters using remotely sensed microwave sensor measurements obtained from an agricultural field equipped with a center pivot irrigation system. In this work, the agricultural field under study is modeled with the cylindrical coordinate version of the Richards equation. Before carrying out the estimation, the sensitivity analysis method is first employed to assess the observability of the augmented state and parameter system. The orthogonal projection method is then used to select a subset of the augmented system that are most relevant to the model predictions for estimation, using the information obtained from the sensitivity analysis method. The extended Kalman filter is then employed to estimate the selected elements of the augmented system. Cross-validation results obtained from this study indicate that, the proposed soil moisture and soil hydraulic parameter estimation method can provide reliable soil moisture information for the implementation of closed-loop irrigation.

References

[1] United Nations World Water Assessment Programme, “The United Nations World Water Development Report 2018”, Nature-Based Solutions for Water, 2018.

[2] O.-T. Chis, J. R. Banga, and E. Balsa-Canto, “Structural identifiability of systems biology models: a critical comparison of methods,” PloS one, vol. 6, no. 11, p. e27755, 2011.