*4.2. Two-Dimensional Modelling*

The outputs for both the FEFLOW and PerTy3 models are consistent across a range of values of B from 0.1 to 10 (Figure 5a). The PerTy3 outputs are flatter and lower than those from FEFLOW. This is due to the representation of recharge external to the irrigated field (Figure 5b). The partitioning between recharge occurring outside and inside the irrigated field is consistent across the models, but PerTy3 shows initial delays in recharge occurring outside the field. Another modelling experiment (not shown here) shows even greater inconsistency, suggesting that assumptions were not adequate. The most likely cause is the lack of consideration of ponding external to the irrigated field. This would have the effect of delaying recharge increases.

The FEFLOW output for the perched head (Figure 5c) if for *x* = 0, while that for PerTy3 is an average across the irrigation field. While the FEFLOW output is higher than the PerTy3, the temporal distributions are approximately parallel. There is a large change in head for *B* = 0.1 (340 cm) to *B* = 10 (120 cm). The wetted width PerTy3 outputs (Figure 5d) shows a large variation from 90 m (*B* = 0.1) to 250 m (*B* = 10).

**Figure 5.** Comparison of outputs from two-dimensional semi-analytical and numerical modelling for new accession rate of 200 mm/year: (**a**) Transfer functions (Total) for *B* = 0.1, 1, 10; (**b**) Transfer functions (total, under irrigation field, external to irrigation field) for *B* = 10. (**c**) Perched heads for *B* = 0.1, 1.0, 10 for an increase in *IA* to 200 mm/year. The head for the FEFLOW model is for *x* = 0, while that for PerTy3 is an average across the irrigated field. (**d**) the width of the wetted zone outside the irrigation field for *B* = 0.1, 1.0, 10, and an increase in *IA* to 200 mm/year.

#### *4.3. Modelling of Brownfield Developments*

The effect of brownfield developments at a range of distances is shown in Figure 6. For low values of *B* (Figure 6a), there is almost no difference whether the new development is placed next to an existing field or at infinite distance (greenfield development). Even for *B* of 10, there is not much difference. For brownfield developments next to an existing development at equilibrium (Figure 6b), there is some earlier recharge and then later some delayed recharge. The earlier recharge is presumably due to pre-wetting by the existing development and the later delays due to the expansion on one side being constrained by the existing development.

(c) 

**Figure 6.** Functions for brownfield developments: (**a**) numerical outputs at varying distances from a pre-existing steady-state development and *B* = 0.1; (**b**) numerical outputs at varying distances from a pre-existing steady-state development and *B* = 10; (**c**) numerical outputs for the total development of a new development followed by another development, 250 m away for *B* = 1, 10 compared to superposition of numerical outputs for two independent developments, one 5 years after the other.

For brownfield developments occurring 5 years after a new development (Figure 6c) there is, minimal effect of separation.
