Groundwater recharge interface and nitrate discharge: Central Canterbury, New Zealand Vince Bidwell Lincoln Environmental Research Lincoln Ventures Ltd Central Canterbury 40 km Central Canterbury C. R. Hanson (Environment Canterbury, 2002) • •• • .,.. • '.. • ·t. -'~ • •• •• • • Darneld • • • • • ,",orolata • • • • • • • • s: . ' • • • • ~ ., ·'.:..• • • • 1 1.3-36 8.5- 11.2 5.7 - 8.4 3.0 - 5.6 1.0 - 2.9 0.0 - 0.9 , .' ~ ". . ..'.- • .~~",,",... . .. • ~ h •• • • .,., . • • ~ .J#ftit ~ou.etton e/~ • ...... • It J { O) k-it-....•• , ElJ rnhanf -/ _ :.. . . ,..... ..... • A e . • •• . . ... • • ;':'tm:si'n11.t ~ 1 . , -. , I • •• .-,a ... ' . ... ... • • • . ' • • • • • . .•'"- ~~'J• • • Me dian nitrate nitrogen concentration (mg/L) • • , Lake Ellesmere • • o 10 20 Kilometers ;..............~iiiiiiiiiiiiiiiiiiiiiiiii 0 Nitrate-N mg/L 5 10 15 20 40 MAV = 11.3 mg/L Depth below water table (m) 60 80 100 From ~ 600 wells in Central Canterbury (adapted from Burden, 1984) Sample of groundwater nitrate-N concentration for Central Canterbury (M. Stewart, 2003) Nitrate-N (mg/L) 0 5 10 15 0 Depth below water table (m) 20 40 60 80 100 120 140 MAV = 11.3 mg/L 160 180 200 Groundwater age derived from tritium data for Central Canterbury (M. Stewart, 2003) Groundwater age (y) 0 Depth below water table (m) 0 40 80 120 160 200 10 20 30 40 50 60 70 80 Groundwater recharge interface • Observed variation of nitrate concentration with depth is due to a: • mixing interface between groundwaters from contaminated land surface recharge and from uncontaminated river recharge (Hypothesis 1) or • descending “cloud” of contamination from many years of agricultural land use (Hypothesis 2) or both Relevance • Existence of a stable interface between groundwaters from contaminated and uncontaminated recharge implies continued access to high quality groundwater for human consumption • Pumped abstraction of this high quality groundwater is likely to cause increased nitrate contamination of groundwater discharge to surface waters Method 1. 2. 3. 4. Identify principal sources of groundwater recharge by means of a steady-state groundwater flow model and separation of piezometric effects Model vertical distribution of groundwater flow by stream function analysis Assign nitrate concentration values (from literature) to recharge sources Model contaminant transport in groundwater with a mixing-cell model Principal sources of groundwater recharge Waimakariri River mean flow = 120 m3/s River recharge unobservable Land surface recharge 200 mm/y (water bal.) 13 m3/s from 2020 km2 Rakaia River mean flow = 210 m3/s Separating the recharge effects on groundwater levels 300 Observed groundwater level Predicted LSR effect River recharge effect Land surface recharge (LSR) 50 250 200 150 45 100 50 40 1972 0 1976 1980 1984 1988 1992 1996 2000 Land surface recharge (mm/mth) GWL (m) above mean sea level 55 The separation of recharge effects was applied to 12 observation well records from Central Canterbury L35/0163 L35/0163 M35/1080 M35/1080 M35/1878 M35/1878 L36/0063 L36/0063 L36/0092 L36/0092 M36/0255 M36/0255 L36/0023 L36/0023 L36/0124 L36/0124 M36/0592 M36/0592 M36/0424 L36/0258 M36/0424 L36/0258 M37/0010 M37/0010 The results show the dominant piezometric effect of river recharge, relative to sea level. 100 Land surface recharge River recharge 80 60 40 20 59 2 M 36 /0 01 0 M 37 /0 87 8 M 35 /1 42 4 36 M /0 36 Observation wells /0 25 5 8 M /0 L3 6 /1 35 25 08 0 2 M /0 09 3 L3 6 /0 06 4 L3 6 L3 6 /0 12 3 02 /0 L3 6 /0 16 3 0 L3 5 Piezometric effect (m) amsl 120 A 4-parameter, steady-state, groundwater flow model was calibrated for prediction of river recharge effect… 100 Observed Predicted 80 60 40 20 Observation wells 59 2 M 36 /0 01 0 M 37 /0 87 8 M 35 /1 42 4 /0 36 M M 36 /0 25 5 58 /0 2 L3 6 M 35 /1 08 0 92 /0 0 L3 6 24 /0 1 L3 6 63 /0 0 L3 6 23 /0 0 L3 6 /0 1 63 0 L3 5 River piezometric effect (m) amsl 120 Observation wells M M 36 37 35 36 36 59 2 /0 01 0 /0 87 8 /1 42 4 /0 25 5 /0 8 25 80 M M M 2 08 0 /1 /0 35 L3 6 M 4 09 /0 L3 6 12 /0 L3 6 3 3 06 /0 L3 6 3 02 /0 L3 6 16 /0 L3 5 Total piezometric effect (m) amsl … and total recharge effect. 120 100 Observed Predicted 60 40 20 0 Groundwater model provides flow paths and recharge inputs for analysis of vertical flow distribution Groundwater flow path Stream function analysis of vertical flow distribution along the example flow path shows the pattern of recharge contributions Groundwater discharge to streams and lake Rakaia River From land surface recharge ~ 300 m From river recharge 50 km Recharge and nitrate discharge for each land use along the flow path are used as inputs to a 2-D vertical groundwater transport model. Combines stream functions and dispersive mixing. Recharge (mm/y) Nitrate-N (g/m3) 1m3/s/km 0 Dairy 250 12 Forest 100 1 Sheep 150 3 Crops 200 15 Land use Rakaia River Nitrate distribution for an example of typical land use Rakaia River recharge Discharge to surface waters Nitrate-N = 2.9 mg/L Sheep Dairy Forest Sheep Crops 300 m 50 km 0.00-1.00 5.00-6.00 10.00-11.00 1.00-2.00 6.00-7.00 11.00-12.00 2.00-3.00 7.00-8.00 12.00-13.00 3.00-4.00 8.00-9.00 13.00-14.00 Nitrate-N concentration (mg/L) 4.00-5.00 9.00-10.00 14.00-15.00 Nitrate v depth, from sampling of model results at 5 km intervals along the flow path Nitrate-N (mg/L) 0 5 10 15 0 Depth below water table (m) 20 40 60 80 100 120 140 MAV = 11.3 mg/L 160 180 200 Comparison of model example with Stewart (2003) Nitrate-N (mg/L) 0 5 10 15 0 Depth below water table (m) 20 40 60 80 100 120 140 160 180 200 Model Stewart (2003) MAV = 11.3 mg/L Conclusions • It is likely that the variation of nitrate concentration with depth in the Central Canterbury aquifers is due to the presence of a dispersive mixing interface between groundwaters from river recharge and from land surface recharge • This interface, although stable, may be in a transient position • The contribution of uncontaminated river recharge is important for maintaining access to high quality drinking water and for reducing the contamination of groundwater discharge to surface waters