Analysing Deep Percolation Dynamics: A Lysimeter‐Based Study in a Cold Environment

ABSTRACT Understanding deep percolation dynamics is crucial for effective groundwater management. This is particularly important in cold regions significantly affected by soil freezing and snow. The scarcity of direct measurements and the insufficient focus on cold regions has limited previous research. This study addresses these gaps by measuring deep percolation by placing four drainage lysimeters inside three large experimental plots in Québec, Canada. The study classified the 67 measured percolation events into two categories: those starting in the cold (36) and warm (31) seasons. Cold‐season events were shorter but had considerably higher peak and average intensities than warm‐season events. Cold‐season events were mostly triggered by rain‐on‐snow and significantly influenced by soil freezing. The study conducts partial correlation analyses to examine the effects of four variables on the volume of percolation events. The variables included the volume and intensity of input water (rainfall and snowmelt), antecedent soil moisture, and soil temperature. The results demonstrate that input water, influenced predominantly by seasonal precipitation and snowmelt, is the principal driver of variations in percolation volumes, with partial correlation coefficients ( r p ) of 0.90 and 0.59 concerning the warm‐season and cold‐season events, respectively. The results also show that the volume of warm‐season percolation events was negatively correlated to rainfall intensity, with r p = −0.52. The findings have important implications for improving hydrological modelling and water management in cold regions. They emphasise the need to consider the dominant roles of rainfall characteristics (volume and intensity) and snowmelt in driving the variation in deep percolation, particularly in the face of changing climatic patterns.