Deep Percolation under Irrigated Water-Intensive Crops

Abstract

Deep percolation from an irrigated field has received little attention in research studies although it contributes to significant loss of fresh water in a given region. Deep percolation is often estimated as a residual in a water balance equation. Several methods have been used to estimate deep percolation from a cropped area ranging from empirical relations to physically based models. However, there are limited studies available concerning deep percolation from water-intensive crop fields such as rice and berseem under unpuddled field conditions. This study deals with the estimation and field verification of deep percolation from water-intensive crop fields. Deep percolation was estimated using both water balance and physically based models. Field observation of deep percolation was carried out using drainage-type lysimeters. Field experiments showed that deep percolation accounted for approximately 85% of the input water in Rice Season 1 and 79% of input water in Rice Season 2. Conversely, about 65% of input water in Berseem Season 1 and nearly 49% of input water in Berseem Season 2 was accounted as deep percolation. The vertical fluxes simulated using the physically based Hydrus-1D model matched the observed data well on a daily time basis, unlike the simple water balance model. Both models performed better on lumped time steps. A simple water balance model could be a preferable choice for lumped time steps, while a process-based model may be advantageous for shorter temporal resolution. Comparatively, the performance of the physically based model is better in the wet season than the dry season, which is attributed to the pronounced development of macropores in the dry season.