Data-Driven and Process-Based Modeling Approaches for Advancing Irrigation Water and Nitrogen Management in Humid Cropping Systems

Abstract

Humid agricultural regions face frequent shifts between too little and too much soil water. Since irrigation decisions depend on the current soil water status, this variability makes it harder to know the right timing and amount of irrigation. The same swings also influence how nitrogen moves through the soil. Climate change is expected to make rainfall more variable and temperatures higher, further complicating both irrigation scheduling and nitrogen (N) management. In these settings, managers need accurate, field-specific soil water parameters and clear evidence on which irrigation–nitrogen strategies perform best under current and future weather. Chapter 2 addresses why accurate soil hydraulic parameters matter in humid systems and how soil-moisture data can provide them. Using high-frequency volumetric water content from corn and cotton fields under non-irrigated, full, and precision irrigation, an event-based hydrologic signature was used to estimate field capacity (FC) and plant extraction limit (PEL) from identifiable periods of infiltration, gravity drainage, evapotranspiration, and rewetting. Compared with two existing signature methods, the event-based approach achieved higher FC classification accuracy (75.0%–79.2%) with lower FC estimation error (1.69%–1.87%). PEL classification accuracy was also higher except at deeper sensors (>30 cm), where stable moisture made events harder to detect. These results show that soil-moisture time series can supply field-specific thresholds for irrigation scheduling in humid regions. Chapter 3 evaluates irrigation and N management strategies under changing humid climates using a calibrated SWAP-WOFOST model and observed soil-moisture, yield, nitrogen uptake and nitrate leaching data. Rainfed (Rainfed-1N), calendar (Calendar-1N), and precision irrigation with single (Precision-1N) or split (Precision-2N) N applications were compared across multiple climate scenarios for yield, N uptake, nitrate (NO₃) leaching, and irrigation water productivity (WP). Precision-2N consistently outperformed Calendar-1N, with higher yields and N uptake and significantly better WP; benefits were largest when daily rainfall variability was higher and smallest when total precipitation increased. Across years, precision advantages were strongest during periods with frequent extreme temperature events. Irrigation strategy had a larger effect than N timing: Precision-1N performed similarly to Precision-2N, while Calendar-1N matched or exceeded precision-treatment yields in about 22% of years but required much more irrigation and produced lower WP. Yield and NO₃ differences were not always statistically significant, but WP improvements were significant in all but one scenario. Chapter 4 examines how modeled outcomes change between definitions of the upper limit of plant-available water (PAW). In many applications, PAW's upper limit is set with a fixed pressure head (e.g., −330 cm or −100 cm), while flux-based definitions use a drainage threshold (qFC) derived from internal drainage behavior. These choices affect how much water the model treats as available to roots during and after drainage and, in turn, influence simulated transpiration, yield, and NO₃ leaching. This chapter implements upper-limit definitions within SWAP-WOFOST using both pressure-head-based and flux-based PAW and quantifies their impacts on maize yield and NO₃ leaching under irrigated and rainfed conditions across historical and projected climates. It also tests whether irrigation reduces or amplifies the differences among PAW definitions. Together, these chapters provide (1) a data-driven way to obtain field-specific FC and PEL for scheduling irrigation in humid fields, (2) model-based evidence that precision irrigation improves yield, water productivity and can reduce NO₃ losses under challenging climate, and (3) a clear evaluation of how PAW upper-limit thresholds propagate into agronomic outcomes under climate variability.