State-dependent influence of ENSO and NAO on spring hydroclimatic variability at the watershed scale in the South-central United States (1950–2025)

Large-scale climate oscillations modulate spring precipitation across the south-central United States, but their hydroclimatic response at the watershed scale remains poorly quantified. We evaluate how the El Niño-Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO) are associated with spring (March–May) precipitation across three watersheds in Northwest Arkansas and Northeast Oklahoma from 1950 to 2025, accounting for antecedent winter (December–February) precipitation. Watershed-aggregated precipitation was analyzed using descriptive statistics and linear mixed-effects models. Sensitivity analyses evaluated temporal dependence and residual dependence among watersheds within the same year. Neutral ENSO conditions show the highest median precipitation across all three watersheds. Mean precipitation is higher during Positive than Negative NAO conditions in the pooled descriptive analysis. Although the original model indicates a negative La Niña x NAO interaction, this term is not significant with year-clustered standard errors (β = −66.10 mm per NAO index unit; p = 0.118). Adjusted ENSO coefficients and the positive temporal trend also lose significance under year-clustered inference. Antecedent winter precipitation is not a significant predictor of spring totals once ENSO, NAO, and time are accounted for. Fixed effects explain 10.9–13.7% of the variance in spring precipitation. The study characterizes watershed-scale precipitation contrasts among climate states and demonstrates the importance of accounting for shared annual variability when assessing teleconnection associations.