How much streamflow is groundwater
discharge?
- A method for assessment
Vince Bidwell
Lincoln Environmental
-Pukemanga Working Group
(NIWA, GNS, LandcareResearch,
AgResearch, Lincoln Environmental)
Why do we want to know?
1. Water quality
Why do we want to know?
2. Water quantity
• Groundwater discharge is the source of “baseflow” in streams
• Baseflow is vital for stream ecology and human use
• Temporal distribution of baseflow (e.g., recession rate) is
determined by the dynamics of groundwater discharge
Model of groundwater discharge
Rainfall
Evaporation
Surface drainage
Soil-water balance model
Vadose zone response time T
Recharge ≤ Dmax
B
Aquifer:
conductivity k
storativity S
α
=
kB
2
SL
L
Groundwater
discharge to
surface water
Two questions to be addressed
and two model parameters to answer them
1. What proportion of catchment drainage becomes groundwater
discharge?
• Defined by the maximum recharge rate Dmax
•
•
•
What is the temporal distribution of the groundwater
discharge?
Defined by the parameter α of a dynamic groundwater model
that maintains a set of water storage states
These two parameters are optimised for best fit to the lower
range of streamflow values
Example 1: Toenepi Catchment, Waikato
• Area = 15.1 km2
• Slope: flat (89%), rolling (10%), steep (1%)
• Elevation: 40 – 130 m (amsl)
• Land use: dairy pasture
• Installed surface and subsurface drainage
Toenepi Catchment: hourly data 24/11/02 – 23/11/03
Rainfall = 1081 mm; Streamflow = 273 mm
0.60
Observed streamflow
0.50
Predicted groundwater discharge
Flow (mm/h)
0.40
0.30
74%
Dmax = 1.7 mm/h
0.20
0.10
0.00
Nov 2002
Jan 2003
Mar 2003
May 2003
Jul 2003
Sep 2003
Toenepi Catchment
Groundwater discharge dynamics
0.25
Flow (mm/h)
Observed streamflow
0.20
Predicted groundwater
discharge
0.15
Rapid response
Complex
recession
0.10
0.05
0.00
Jun 2003
Aug 2003
Oct 2003
Example 2: Pukemanga Catchment, Waikato
• Area = 3 ha
• Slope: hilly (17°- 20°) to steep (>30°)
• Elevation: 72 – 146 m (amsl)
• Land use: sheep pasture
• Headwater stream from a wetland
Pukemanga Catchment: hourly data 1/4/98 – 31/3/99
Rainfall = 1706 mm; Streamflow = 1013 mm; Area = 1.43/3.0 ha
18.0
Observed streamflow
16.0
Predicted groundwater
discharge
14.0
Flow (mm/h)
12.0
62%
Dmax = 2.0 mm/h
10.0
8.0
6.0
4.0
2.0
0.0
Apr 1998
Jun 1998
Aug 1998
Oct 1998
Dec 1998
Jan 1999
Pukemanga Catchment
Groundwater discharge dynamics
0.50
Observed streamflow
Predicted groundwater
discharge
Flow (mm/h)
0.40
0.30
0.20
0.10
0.00
Apr 1998
Jun 1998
Aug 1998
Oct 1998
Dec 1998
Jan 1999
Pukemanga Stream catchment is 48% of topographical area:
hillslope “stream function” model
Extent of groundwater system
For Pukemanga Stream
Wetland origin of
Pukemanga Stream
River
“Deep” groundwater system draining to river
Sources of the other 38% of Pukemanga streamflow?
- piezometric levels near the wetland (current research)
0
0.80
Flow (mm/h)
-50
0.60
0.40
-100
Observed streamflow
Predicted groundwater discharge
Piezometric level ~1 m depth
Piezometric level ~ 4.5 m depth
-150
-200
0.20
-250
0.00
Apr 2004
-300
May 2004
Jun 2004
Piezometric level (cm below ground)
50
1.00
Comparison of baseflow recession
in response to a unit pulse recharge
0.2
Relative groundwater discharge
0.18
0.16
Toenepi - 1510 ha
Pukemanga - 3 ha
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
0
10
20
30
40
50
60
Days since unit pulse recharge
70
80
90
100
Comparison of baseflow recession (log scale)
Relative groundwater discharge (log)
0.1
Toenepi - 1510 ha
Pukemanga - 3 ha
0.01
α = 0.0096 /d
kB
__
α=
SL2
0.001
α = 0.0278 /d
0.0001
0
10
20
30
40
50
60
70
Days since unit pulse recharge
80
90
100
Conclusions about quantity and quality
• For both catchments, during the year of observations, more
than 60% of the total annual streamflow is groundwater
discharge, from recharge at rates less than 2 mm/h
• Quality of this groundwater discharge would be determined by
leaching of the soil profile and exposure to geochemical
processes in the aquifer
• Much of the remaining streamflow may be from drainage of
saturated near-surface zones, with quality determined mainly
by leaching of the soil profile (current research)
• Contribution of overland flow to total streamflow is small
Conclusions about groundwater discharge dynamics
• Response of groundwater discharge to recharge entering the
groundwater surface is essentially instantaneous
• More of the streamflow during drainage events can be ascribed
to groundwater recharge than predicted by some alternative
baseflow separation methods
• The complex behaviour of streamflow recession is described
by a single-parameter model, based on the theoretical
dynamics of groundwater flow, which preserves water storage
states