A water balance model for rainfed lowland rice fields emphasising lateral water movement within a toposequence

Abstract

In the Mekong region, most rice paddies lie in gently sloping lowlands. An accurate estimate of paddy water availability is crucial for modelling rice ( Oryza sativa L.) productivity in such environments. However, modelling of water balance in the sloping lowlands can be difficult due to problems in estimating lateral water movement from high to low positions in the toposequence. This paper describes a semi-empirical model for estimating net lateral water flow along a toposequence of rice fields. Net lateral water flow is separated into three sub-components: (i) lateral seepage through the bunds from field to field; (ii) surface runoff over the bunds from field to field; (iii) water run-on from the catchment above the toposequence. The lateral seepage is estimated using the Dupuit equation for steady unconfined flow. Surface runoff over the bund is calculated as excess water depth above the bund height, while run-on from the catchment is calculated using a rainfall–runoff relationship. The other water balance components used in the model are rainfall (measured), paddy evapotranspiration (ET), and downward water flow (within the field and under the bund). Paddy ET is estimated using the FAO crop ET model. The downward water flow is empirically determined for soil texture groups. Simulation results showed that field water levels computed by the model varied between the bund height (maximum) and the hardpan level (minimum). The daily change in the water level was satisfactorily validated using data collected from a field experiment in a toposequence in Cambodia. The component governing the water balance in the higher toposequence positions was the downward water flow (45% of seasonal rainfall at the top of the toposequence). However, the water balance in the lower positions was affected largely by the net lateral water flow (21% of seasonal rainfall at the bottom of the toposequence). Seasonal ET was 67–69% of seasonal rainfall.