System dynamics modelling as discussion support tool for upscaling subsurface irrigation

Fresh water is needed worldwide for agricultural and economic sectors, but also for nature. However, the water demand continues to increase due to economic growth, urbanization and increased food production, while water availability decreases. Thereby, weather extremes are expected to increase and occur more frequently. As a result, there is an increased mismatch between water demand and water supply. Historically, Dutch agricultural fields were drained to remove water in wet periods. Nowadays, drainage systems are increasingly being modified to controlled drainage with subirrigation (CDSI) systems to i) discharge water when needed and ii) retain water and iii) recharge water when possible. However, the implementation of CDSI on a local scale alters several water balance components. CDSI positively affects transpiration for crop growth, increases drainage to the surface water and increases downward seepage, i.e. groundwater recharge. However, CDSI also requires surface water, which is not infinitely available. It is, therefore, important for regional water management authorities to understand how the field scale measure CDSI propagates through the regional water system in order to estimate if sufficient surface water is available to scale up CDSI to other fields.

A system dynamics model (SDM) approach is used to get insight into the hydrological effects of upscaling CDSI. SDM’s are widely used to understand non-linear behavior of complex systems with feedback-driven components in order to make policy decisions for example. Our SDM is a simple, but comprehensive model based on four field experiments conducted in the Netherlands and a detailed calibrated Soil, Water, Atmosphere and Plant (SWAP) model. The results show that the SDM takes account of different feedback loops that determine the possibilities of upscaling CDSI. This includes an increase in drainage to the ditch, but at the same time, subirrigation lowers the ditch level, which in turn reduces drainage to the ditch. The results further show 3 CDSI possibilities: i) sufficient surface water is available to scale up CDSI to 20 % of the area, ii) sufficient surface water is available, but surface water levels decline when scaling up CDSI between 20 – 30 %, iii) insufficient surface water is available to scale up CDSI between 30 – 100 % as the surface water runs dry. In the latter case, hydrological characteristics (regional surface water inflow, regional weir height and minimal surface water level) can be adapted to increase the regional water availability and therefore allow further CDSI upscaling. We show that a SDM is a useful method for an initial design of how a local measure affects the regional water management, which gives the regional management authority insight in the hydrological effects of upscaling measures and therefore supports the conversations between policy makers and stakeholders (e.g. farmers).