Using Field Instrumentation to Validate Numerical Modelling.

Numerical models, such as Hydrus 1D, can be used to estimate the time lag (t_u) of nutrients to groundwater. Such models frequently use soil hydraulic properties measured under laboratory conditions or inferred, via pedotransfer functions, as input parameters. Consequently, in order to comment on their accuracy, models should be compared to real-life equivalents of the scenarios they purport to simulate.

In situ field tracer tests are currently being conducted in two vulnerable agricultural watersheds (arable and grassland) in the south of Ireland, to validate estimates of t_u made using Hydrus 1D. At each site, two soil pits (one mid-slope and one approximate to the receptor) were excavated until bedrock was reached. Pits were sampled and described in accordance with national guidelines, which will provide soil characteristic data for the numerical model. Each horizon was equipped with pore water samplers, matric potential probes, TDR or electrical conductivity probes, and temperature sensors. These devices were installed in early May 2014, and will be monitored until the beginning of the recharge period in early September to establish background readings. A potassium bromide tracer will then be applied and breakthrough at each horizon will be observed. Synoptic weather stations are present at each location, and will provide meteorological records to be used as model inputs. These results will indicate the suitability of this model for estimating t_u.