MODELLING SURFACE WATER-GROUNDWATER INTERACTION WITHIN
NIWA’S NATIONAL HYDROLOGIC MODELLING FRAMEWORK
Yang J., McMillan H.K., Zammit C.
National Institute of Water and Atmospheric Research
Aims
Over the past 13 years, NIWA has been developing an uncalibrated national hydrologic modelling
system to simulate the hydrologic cycle across New Zealand. The model is used to support
different water related decision making, ranging from flood prediction, water availability
estimation, to impact assessment of landuse and climate change. NIWA’s national hydrological
model (TopNet) (McMillan et al., 2013), is based on TopModel concepts, has been applied
successfully in most areas across New Zealand except some plains where strong interactions
between surface water and groundwater are present. This is mainly due to the fact that
groundwater process are too simply to represent correct hydrological processes in those areas.
One objective of the MBIE contestable Waterscape program is to enhance the capability of
TopNet model in simulating groundwater flow and interactions between surface water and
groundwater across all New Zealand landscape.
Method
An additional conceptual groundwater store was added to the current TopNet structure. This
allows TopNet to better represent groundwater processes in areas where hydrological processes
are not topographically driven. The main features of the updated model include:

Groundwater flow: from one groundwater store to another one or two stores (fluxes
and
in Fig. 1)

River-groundwater interaction: either losing or gaining rivers (flux
local information

Exchange with the external: gaining from/losing to the other watershed (flux
in Fig. 1) based on
in Fig. 1)
Figure 1. The Improved TopNet structure for
groundwater flow and river-groundwater interaction
(Left: Original TopNet structure; Right: Improved
structure. SS - shallow aquifer, SR - river, and SD – deep aquifer. Arrows are
fluxes)
Results
The updated model was applied to the Pareora catchment in the Canterbury region (Fig. 2). The
Pareroa catchment drains an area of 539 km2 with annual precipitation 700 mm. Previous studies
(e.g. Wilson and Aitchison-Earl, 2014) show there is a strong interaction between river and
groundwater in the lower Pareora catchment.
Flow stations:
Huts, and State Highway 1 (SH1)
Concurrent gaugings:
Holme Station Bridge, Brasells Bridge, and SH1
Losing reaches:
Huts to Brasells Bridge
Gaining reaches:
Brasells Bridge to SH1
Huts
Figure 2. Location of Pareora catchment
Application and results:
 Original TopNet parameters were calibrated at Huts; Groundwater parameters were
calibrated with concurrent gaugings in 2000 and 2001 (Fig. 3)
 Model was validated with spot gaugings in 2003, and flow at SH1 from 2009 to 2011
 Updated TopNet obtained a good match with observed flows
 There is a significant improvement in the low flow simulation, compared to original TopNet
Figure 3. Flow comparison among spot gaugings (red dot), original TopNet simulation (black line), and
improved TopNet simulation (blue line)
References
McMillan, H.K., Hreinsson, E.Ö., Clark, M.P., Singh, S.K., Zammit, C., & Uddstrom, M.J., 2013. Operational
hydrological data assimilation with the recursive ensemble Kalman filter. Hydrology and Earth System
Sciences, 17(1), 21-38.
Wilson N., Aitchison-Earl, P., 2014, Pareora groundwater flow system conceptualization (report draft)
Acknowledgement This work is Funded by New Zealand’s MBIE via the Waterscape programme
(C01X1006). We thank Philippa Aitchison-Earl from ECan for data and discussion.