Association of Atmospheric River Frequency over the Western Himalayas with Large scale Modes of Climate Variability

Sheikh Imran Fayaz, Munir Ahmad Nayak, Mohammad Humza Kanth

Winter atmospheric rivers (ARs) contribute substantially to precipitation in the Western Himalaya, but how large-scale climate modes jointly influence their frequency remains unclear. Using 19 climate modes and their interactions over 39 winters (1980-2018), we combine correlation, phase-composite analysis of integrated vapor transport (IVT), and linear, generalized linear (GLM) and generalized additive (GAM) models. The Pacific/North American (PNA) and East Atlantic/Western Russia (EA-WR) patterns are strongly associated with winter AR frequency and show consistent IVT composite signals, whereas East Pacific-North Pacific (EPNP) acts mainly through its interaction with PNA. Notably, several modes show significant IVT changes in the phase analysis like Arctic Oscillation and PNA, but are not prominently retained in models with least error metrics. No single mode explains AR variability, and models with interaction terms outperform a simple linear model. The GLM and GAM show comparable predictive skill, but the GAM utilizes three climate modes against eight for the GLM. EA-WR and the EP-NP × PNA interaction are retained in both models, and EP-NP also enters the GAM as a nonlinear main effect. These results indicate that winter AR frequency depends on the combined and partly nonlinear effects of several climate modes, providing a statistical basis for predicting seasonal AR frequency and water resource planning in this data-limited region.