Applications of in-situ soil moisture observations to better characterize field-level water dynamics

Abstract

Water is a relevant input in crop production and represents the second major consumer of freshwater withdraws in United States. In the state of Kansas, a total of 9 million hectares are used for crop production annually, with 1 million hectares being managed in irrigation conditions. In a current scenario facing challenges associated with water shortages and concerns about water use in agriculture, accurate monitoring of rootzone soil moisture has become relevant towards a more efficient use of water in agricultural systems. In this thesis we present and discuss two research questions aimed at improving water use based on soil moisture monitoring in rainfed and irrigated agricultural fields. The first research question that we addressed is: How can we define the number and optimal deployment location of soil moisture sensors in agricultural fields under rainfed and irrigated conditions based on soil moisture-based management zones? This study involved the intensive collection of in situ surface soil moisture observations and soil physical properties across multiple fields. Delineation of soil moisture-based management zones was compared to common proxy variables used to characterize management zones such as soil texture and elevation. A new method to characterize soil moisture-based management zone is proposed in order to objectively define the optimal number and tentative location of soil moisture sensors in crop fields. The second research question that we addressed is: Is it possible to estimate rootzone soil water storage solely based on surface soil moisture observations? In this study we tested a widely used semi-empirical exponential filter model to estimate rootzone soil moisture in agricultural fields using observations from a single soil moisture sensor located near the soil surface. Different rootzone depths were tested and the accuracy of the model was calculated in order to evaluate the feasibility of this model to guide irrigation management. As a general result we propose an objective methodology to guide the deployment of a limited number of soil moisture sensors across crop production fields as well as a method to delineate in-field soil moisture management zones based in soil moisture observations. In addition, we provided useful insights to estimate rootzone soil moisture from near-surface soil moisture observations in agricultural fields.