Globally, water is a critical and growing concern for both natural ecosystems and human activities due to increasing frequency, intensity, and duration of drought and depletion of reserves such as aquifers and reservoirs. Improved precision in the control of water inputs can provide significant increases in the profitability of all crops. However, to achieve optimal horticultural benefits, growers require physiologically relevant information on plant water status in the field. In-plant water potential influences plant growth, yield, quality and susceptibility to disease. Stem water potential (SWP) is acknowledged to be an integrator of water status across the soil-plant-atmosphere continuum (SPAC), but it remains difficult to measure. The manually operated Schölander pressure chamber (SPC), introduced over 60 years ago, remains a gold standard for SWP measurements. Limitations of available techniques have hindered the study of dynamic water stress in plants and the development of physiological models to rationalize the plant responses. In this dissertation, we provide an overview of the design and application of an in-situ, minimally destructive hygrometer, the micro-tensiometer (µTM), for accurate and continuous measurements of SWP. We first describe the application of the μTM in three important woody perennial crops: apple (2 months-NY), almond (4 months-CA), and grape (12 months-CA) with thorough benchmarking against the SPC. In all species, we observe significant night-time disequilibrium, an effect that is difficult to capture with the manual SPC. We also identify differences between the well-watered (apple) and water-stressed (almond) dynamics and the effects of seasonal phenology (grape). We then present the integration of the μTM in a multi-point sensing framework characterize the SWP dynamics in apple as a function of micro-climate drivers and management. We discover that the transient root response depends on both the plant stress level and the timing of irrigation. These phenomena suggest more efficient irrigation scheduling should be applied before midday or after sunset and when the plant is mildly stressed. To further interpret the set of continuous data, we turn to model development of the coupled dynamics of soil-plant atmosphere continuum (SPAC) using circuit analogy. A pair of hydraulic resistor and capacitor could effectively capture the rapid response of plant to high frequency fluctuations in the atmospheric demand in a wet environment with sufficient levels of soil water. The dynamics become more complicated when soils dry. The highly nonlinear responses of the soil compartment, with additional resistors and capacitors defined by the soil retention properties, is necessary to capture the observed dynamics of SWP under dry conditions. Using models as tools to elucidate the water stress dynamics, we vary the circuit architecture to examine major source of loss of conductance (vulnerability) along the SPAC. Our rationale of the central role played by the soil-root interface in defining plant drought responses could help address the observed persistence of disequilibrium between the soil and the plant during dehydration. This finding supports an existing but neglected hypothesis in the literature and provides new insight into the biophysical design of plant vasculature. The combination of the μTM and the matured circuit models open a new opportunity to unveil the full dynamics behind plant water relations, provide new insights to address the transient factors affecting plant physiological responses to soil dehydration, and open a new route to improve prediction of water stress dynamics to inform water management in irrigated crops.