Water Management Proceedings of the Institution of Civil Engineers

Volume 166 Issue WM5 Water Management 166 May 2013 Issue WM5
Pages 231–241 http://dx.doi.org/10.1680/wama.12.00064
Effect of drought on groundwater in a Paper 1200064
Chilean irrigated valley Received 29/05/2012 Accepted 02/01/2013
Keywords: drainage & irrigation/groundwater/hydrology & water resource
Arumı́, Rivera, Holzapfel and Muñoz

ICE Publishing: All rights reserved

Effect of drought on
groundwater in a Chilean
irrigated valley
j
1 José Luis Arumı́ PhD j
3 Eduardo Holzapfel PhD
Professor, Water Resources Department, Agricultural Engineering Professor, Water Resources Department, Agricultural Engineering
College, University of Concepcion, Chillan, Chile College, University of Concepcion, Chillan, Chile
j
2 Diego Rivera PhD j
4 Enrique Muñoz PhD
Associate Professor, Water Resources Department, Agricultural Assistant Professor, Civil Engineering Department, Universidad
Engineering College, University of Concepcion, Chillan, Chile Católica de la Santı́sima Concepción, Concepción, Chile

j
1 j
2 j
3 j
4

Chile is strongly impacted by climate variability driven by the El Niño southern oscillation and the Pacific decadal
oscillation. These produce multi-year droughts that affect Chilean agriculture, which depends heavily on available water
resources. This paper is based on the results of a 6-year study that was carried in the Peumo area of south central Chile,
where there is intensive agriculture based on vineyards, orchards and annual crops. In the Peumo area, irrigation has a
strong impact on groundwater recharge (22% from irrigation losses, 52% from canal seepage and 26% from rainfall).
This research analysed how water management decisions can affect the groundwater system and how agricultural
production within the valley would be affected. It was found that decisions such as intensive use of wells and the lining
of canals will produce depletion of the water table, affecting low-income farmers who depend on shallow wells.

1. Introduction the aquifers and eventually forcing the Chilean water authority
Irrigated agriculture has long been a major economic activity in (Dirección General de Aguas (DGA)) to restrict further well
the Central Valley of Chile (Oyarzún et al., 2007), but in recent construction.
decades agriculture has undergone a re-invention: advantage is
being taken of the Mediterranean climate of the area, the demand As a mitigation strategy against climate variability, farmers keep
in the northern hemisphere for fresh agricultural products in their groundwater rights as a reserve for dry years, preferring the
winter and the growing market for Chilean wines. As a conse- use of surface water due to its lower energy cost. In this way, the
quence, there is increasing production of vegetables, fruits and hydrological systems in the Central Valley are kept in equilibrium
wine grapes in the Central Valley, which are heavily dependent conditions as described by Rivera et al. (2007), Uribe (2008) and
on available water resources. Arumı́ et al. (2009). However, a recent long period of below-
average rainfall years (2007–2012) and the uncertainties asso-
The strong dependence of agriculture on water resources makes ciated with climate variability led the authors to evaluate the
this sector highly vulnerable. The Central Valley of Chile is changes in water use patterns as well as the corresponding
strongly impacted by climate variability caused by the El Niño changes in the vulnerability of the system. The evaluation was
southern oscillation (ENSO) phenomenon, which results in based on the results of a 6-year study developed in the
irregular fluctuations of above-average annual rainfall (,30%) agricultural valley of Peumo, where canal management and
during the El Niño (warm) episodes and below-average rainfall irrigation practices were found to deeply affect the hydrological
during La Niña (cold) periods (Garreaud, 2009). Traditionally, system of the valley (Arumı́ et al., 2009; Orrego, 2003; Rivera et
farmers based their production of annual crops on surface water al., 2007).
using river-fed irrigation canals. However, droughts associated
with La Niña years (e.g. 1990, 1999, 2010) produced a deple- 2. Study site
tion of river flows, forcing farmers to use groundwater to The Peumo area (160 km2 ) is located at 34.38S, 71.38W, about
irrigate their orchards and vineyards. This situation led to 100 km south of Santiago, Chile, and has an altitude ranging
intensive use of groundwater, surpassing the safe yield limit of from 112 to 192 m a.s.l. (Figure 1). The study area encompasses

231
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

the northern half of the lower part of the Cachapoal River basin, j The period from February to March is the dry season,
between the town of Peumo (34.418S, 71.148W, 192 m a.s.l) and which is characterised by very low rainfall but high
the Rapel reservoir (34.268S, 71.428W, 112 m a.s.l). The head- evapotranspiration. The smaller rivers run dry and the flow
waters of the Cachapoal River are in the Andes mountains, and of the Cachapoal River is highly reduced in large sections
its watershed covers approximately 6500 km2 : After crossing the due to diversions for irrigation and the reduction of
Central Valley, the river bisects the Coastal Range, where the snowmelt.
valley bottom forms the narrow southwest to northwest-oriented
Peumo area. The river flow reaches a maximum when the rainfall Figure 2(b) shows the annual rainfall for the period 1960–2006
is greatest in the southern hemisphere winter and the effect at Rapel, located at the reservoir downstream of the study area.
of the spring snowmelt is felt in late spring and summer The mean rainfall for the period (ì) is 564 mm, with a standard
(Figure 2(a)). deviation (c) of 236 mm. The ratio of standard deviation to mean
(c/ì) at Rapel is 0.42, which is indicative of a high interannual
The hydrologic regime of the Cachapoal River follows the same variability (Garreaud, 2009). As observed, the time series exhibits
tendency of all rivers located in central Chile (Nuñez and Soto, interannual fluctuations that are mainly caused in the study area
2010) that receive water from the Andes, which can be divided by ENSO (Montecinos and Aceituno, 2003; Montecinos et al.,
into three main periods. 2000) and the Pacific decadal oscillation. Thus, the variability of
the climate depends on two superimposed timescales – interann-
j The period from April to September is the rainy season, with ual and decadal –which lead to decades of enhanced warm–wet/
abundant surface runoff in both the river and the streams that cold–dry periods. This variability is also shown on a monthly
drain the slopes of the Coastal Range mountains. Irrigation basis in Figures 2(c) and 2(d), where outliers and larger c/ ì ratios
canals are not operational during this period to avoid flood occur during spring and summer due to extreme rainfall events
damage. The delay between the peak monthly cumulative for the period. However, it is worth noting that, in those periods,
rainfall in June and the peak monthly average flow rate in the amount of rainfall is less than 10% of the annual value
August is normal for rivers in central Chile and is due to the (Figure 2(d)).
fact that soils are not fully saturated until the end of Chilean
autumn. The Peumo area (160 km2 ) is the lower section of the Cachapoal
j The period from October to January is characterised by watershed, and is limited to the north and south by the Coastal
occasional rain, with less surface runoff than the rainy period. Range mountains. The upper limit of the area is defined by the
Some of the smaller rivers dry up, while the Cachapoal River Cachapoal River, which presents a bend where it turns around a
mostly receives snowmelt inputs from the Andean mountains. section of the Coastal Mountains to enter a narrow 30 km valley.
Between the middle and the end of September, irrigation The area is bounded downstream by the Rapel reservoir
diversions are opened. (695 million m3 ), constructed in 1970. The predominant agricul-

Cachapoal
River
Rapel
reservoir

Peumo
area

0 15 30 km

Figure 1. Location of Peumo area and Cachapoal River

232
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

Average monthly flow: m3/s

Annual precipitation: mm
160 250 1200
140 1000
precipitation: mm
Average monthly

Precipitation
200
120 Flow 800 ⫹c
100 150
80 600 μ
60 100
400 ⫺c
40 50
20 200
0 Dec 0 0
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005
(a) Year
(b)
500 160 4
Monthly precipitation: mm

Monthly precipitation: mm
c /μ 92%
140
400
120 3
300 100

c /μ
80 2
200
60
100 40 6% 1
2%
20
0
0 0
Dec

Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov

Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
(c) (d)

Figure 2. (a) Average rainfall over the valley and the flow
measured at the Cachapoal River upstream of the Rapel reservoir.
(b) Annual amount of rainfall, mean annual rainfall and deviations
from the mean for the period 1960–2006 at Rapel. (c) Monthly
rainfall variability. (d) Annual distribution of rainfall in an average
year showing the ratio of standard deviation to mean (c/ ì)

tural activity is fruit production, which requires irrigation from earthen canals, most of which are more than 100 years old
water that comes mostly from the river and to a lesser extent (Figure 3).
from groundwater wells within the valley. The irrigation canal
network covers the entire valley and is composed of unlined The groundwater system in the area consists of an unconfined

Rapel Coastal Mountains

N reservoir

Cachapoal
River

Peumo

0 2·5 5·0 7·5 10·0 km

Figure 3. The Peumo area is limited by the Coastal Mountains,

the bend in the Cachapoal River shown (lower right-hand corner)
and the Rapel reservoir (upper left). The area is covered by an
extensive network of irrigation canals showed by the white lines

233
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

aquifer. The analysis of 15 well logs, distributed throughout the j four wells in an avocado orchard, also at the upstream end of
area, identified the existence of three layers the area
j seven wells in a table grape field at the lower end of the
j an upper layer, which is an aquifer composed of sand and valley.
gravel from fluvial deposits with an average thickness of
50 m To characterise shallow groundwater flow and study the influence
j a thick sand and silt layer of an average of 5 m thickness of irrigation canal seepage on the groundwater table, additional
j a sandy and gravel layer of average thickness 20 m. groundwater monitoring locations were sampled throughout the
area (Figure 4). Water levels were also measured at surface water
locations throughout the valley – at upper and lower valley
The saturated hydraulic conductivities range from 30 to 100 m/day locations on the river, in five irrigation canals and at four
(DGA, 2005; Hauser, 1990; Rodrı́guez et al., 2002). The highly locations in irrigation drainage canals.
permeable layers are traversed by the irrigation canal network,
which allows for the possibility of groundwater recharge by During the execution of the 6-year project, several irrigation
seepage from leaky canals (Rivera et al., 2005). experiments were conducted at the three study fields (Holzapfel
et al., 2011; Jara et al., 2011). An extensive irrigation practices
3. Experimental method survey was conducted for the whole area (9600 ha) to estimate
A monitoring network was installed to obtain groundwater, the irrigation efficiency for the different crops, fruits and the
surface and irrigation water data. Data were collected at experi- irrigation practices used in the zone. From the information
mental fields and additional control points in the area (Figure 4). provided by the survey it was found that 51% of the irrigated area
The measuring period was from 2003 to 2006. The data were is dedicated to orchards, 41% to annual crops and the remainder
taken monthly and more frequently from October to April during to prairies and other uses. The irrigation level is good: 43.6% of
the irrigation season. the irrigated area has drip irrigation, 43.3% furrow and 13.1%
sprinkler systems (Araneda, 2004).
Observation wells were installed to study the effect of irrigation
systems on the water table at the field scale. Wells had a 5 cm inside 3.1 Groundwater model
diameter steel pipe screened over the bottom and driven manually, A groundwater model was developed using geological and
so that well bottoms were approximately 1.5 m below the lowest hydrogeological information obtained from previous studies
seasonal water table, thereby covering the normal fluctuation of the (DGA, 2005; Hauser, 1990; Rodrı́guez et al., 2002). The
water table. Wells were installed at three study fields numerical model was built using PMWIN 5.1 software (Chiang
and Kinzelbach, 2001), which is based on the ModFlow simula-
j six wells in a vineyard at the upstream end of the valley tion model of McDonald and Harbaugh (1988). The study area

Grape field

Observation wells
used for model La Rosa
calibration
Water table control
points
Vineyard
Experimental field

Avocado
field

0 2·5 5·0 7·5 10·0 km

Figure 4. Monitoring network installed to obtain groundwater,

surface water and irrigation water data at field and area scales

234
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

was discretised by three layers of a uniform quadratic grid with a Groundwater levels from two years of measurements (2003–
cell size of 250 m by 250 m. The hydraulic properties of each 2004) from six observation wells located along the study area
layer are presented in Table 1. The boundary conditions for the (Figure 5(b)) were used for the calibration. Additionally, the
model were as follows. model was validated by a consistency analysis of the water
balance. The first model approximation neglected the effect of
j The Coastal Mountains, located along the north and south irrigation canals on groundwater, but field data, field observations
border of the valley, were designed as no-flow boundaries for and modelling showed that this was incorrect. The main irrigation
the three layers. The mountains are a granitic formation from canals were thus incorporated into the groundwater model during
the Paleozoic period and have negligible flow across them the irrigation season. Considering the limited data available, it
(Rodrı́guez et al., 2002). was decided to use the general head boundaries package as a
j Cachapoal River and Taguilla Creek were used as general simpler approach. Thereby, the flow between the canal and the
head boundaries for the upper layer (Anderson and Woessner, aquifer was calculated by multiplying the difference of the
1992). hydraulic head value by the conductance of the canal bed, which
j Rapel reservoir was used as a downstream head boundary for is a function of canal width, longitude along a model block, canal
the three layers (Figure 5(a)). bed thickness and hydraulic conductivity (Anderson and Woess-
ner, 1992). This approach is limited by the condition that water
The last boundary condition was considered a simplification of table levels do not drop below the bottom of the canal, a
the main geographical feature that can be identified at the bottom condition that is fulfilled because groundwater levels are shallow
of the Peumo area. Differences in the groundwater model results in the study area.
for different reservoir levels (full capacity, average value and
minimum operational level) were evaluated. After comparing the 3.2 Modelling scenarios
simulated groundwater levels for each condition, the differences After the drought of 1998–1999, farmers in the area constructed
were never found to propagate more than 5000 m upstream of the depth wells to ensure irrigation with groundwater resources.
reservoir, due to the topographic slope of the area. The average However, if available, they prefer to use low-cost surface water,
head of the reservoir was used, therefore, as a constant head value keeping wells as a water source reserve for future droughts.
as the main study area was located more than 7000 m upstream Following that trend, it is possible to infer that after a long or
of the reservoir (Figure 4). severe drought period, farmers will change their water manage-
ment practices in two main ways – by reducing seepage losses
Groundwater recharge was estimated using a monthly hydrologi- lining irrigation channels and/or by increasing groundwater
cal model developed by Orrego (2003), which includes rainfall pumping at the existing wells. As discussed earlier, the current
and losses from irrigation. The contribution to recharge from shallow groundwater levels of the valley are due mainly to
irrigation losses was estimated using average values of irrigation recharge from irrigation canal seepage, so the reduction of such
efficiency for the different crops and irrigation practices existing recharge and an increase of pumping from the wells leads to a
in the area (i.e. 90% for drip irrigation, 40% for furrow and 80% depletion of groundwater levels in the valley. The question then
for sprinkler systems). The data for the topography were obtained arises as to who will be affected by these conditions.
from SRTM (Shuttle radar topography mission) data from NASA
(http://www2.jpl.nasa.gov/srtm/), 90 m resolution). To explore the effect of such scenarios, different simulations of
the groundwater model were run with a reduction of the rain
The groundwater model was calibrated for semi-steady-state recharge of 50% (representing the reduction of rain during a
conditions using two seasonal average conditions drought period) and

j winter without irrigation and with rain recharge

j an increase of groundwater pumping up to 100% of actual
j summer with irrigation and active irrigation canals.
water rights (approximately 1600 l/s distributed over the
whole valley)
Layer Description Thickness: Hydraulic Porosity j an increase of groundwater pumping plus a reduction of
m conductivity: groundwater recharge caused by lining irrigation canals.
m/day
The model scenarios were run in unsteady-state conditions,
1 Sand and gravel 50 30 0.25 considering two seasons
2 Sand and silt layer 5 3 0.35
3 Sand and gravel 20 30 0.25
j the winter season with the wells stopped, the canals non-
Table 1. Hydraulic properties of the three layers considered for operational and with recharge by rain
the groundwater model j the irrigation season with the wells and canals operating and
recharge by irrigation.

235
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

6 210 000 Rapel reservoir

Coastal Mountains used as

north no-flow boundary condition
6 205 000
North: m

Taguilla Creek
6 200 000
Simulated
valley
Cachapoal
6 195 000 River

Coastal Mountains used as

south no-flow boundary condition
6 190 000

275 000 280 000 285 000 290 000 295 000 300 000
East: m
(a)

0·5
Simulated water table depth: m

1·0

1·5

2·0

2·5

3·0

3·5
⫺3·5 ⫺2·5 ⫺1·5 ⫺0·5
Observed water table depth: m
(b)

Figure 5. (a) Boundary conditions used for the groundwater

model. (b) Comparison between two years (2003 and 2004)
average observed water table depth in six observation wells
distributed along the valley and the simulated values with the
groundwater model

Lining of irrigation canals was simulated assuming that the canal is fully replenished, but in a dry year it does not have the chance
is impermeable by setting the canal conductance to zero. to be recharged by deep percolation. However, the water table
during the irrigation season is always similar, which indicates that
4. Results recharge during the irrigation season is constant and depends on
Measurements of the water table at the different observation wells irrigation practices.
allowed observations to be made of some patterns of groundwater
behaviour. At the annual scale, water levels closely follow the Measurements of two wells located in an upper valley vineyard
climate conditions (Figure 6(a)). In a rainy year, the upper aquifer (Concha y Toro) and at La Rosa Farm (located in the centre of

236
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

Wet Irrigation Dry Irrigation Average

winter season winter season winter
0
0·5
Water table depth: m

1·0
1·5
2·0
2·5
3·0
3·5
05-11-2001 24-05-2002 10-12-2002 28-06-2008 14-01-2004 01-08-2004 17-02-2005
Date: day-month-year
(a)

⫺1·5 97·5

97·2 Opening of
Water table depth: m

⫺2·0 Water table depth: m irrigation canals

Vine grape 97·0
⫺2·5 Well 1
La Rosa farm 96·7
Opening of canal Well 2
⫺3·0
96·5
⫺3·5
96·2
⫺4·0 96·0
Jan. Apr. Aug. Nov. Feb. May. 15 Sep 2005 30 Sep 2005 15 Oct 2005 30 Oct 2005
(b) (c)

10
24 hours rainfall: mm

8
6
4
2
0
15 Sep 2005 30 Sep 2005 15 Oct 2005 30 Oct 2005
(d)

Figure 6. (a) Observed groundwater levels at observation well

located at the upper valley vineyard from February 2002 to
December 2004. (b) Observed groundwater levels in the vineyard
and at La Rosa Farm, from April 2004 to May 2005. (c) Water
table levels recorded hourly at two observation wells located in
the upper valley vineyards. (d) Rainfall between 15 September
and 30 October 2005

the area) show that the groundwater level dropped during the from the main irrigation canals. The central valley observation
winter of 2003 (from April to August), and then the water table well is located near a major drainage creek from which the water
rose by 40 cm three to four weeks after the beginning of level is stabilised.
irrigation canal operations in September 2003 (Figure 5(b)).
Around late spring (November) and early summer (January– More detailed information on groundwater response to irrigation
February), water levels dropped again at La Rosa Farm because seepage was observed with the automated sampling locations in
the well is located close to the Taguilla creek, which acts as a the upper valley vineyard (Figure 6(c)). Both wells showed a
drain. The different water table depth (Figure 6(b)) can be water table rise of 35 cm after the opening of irrigation diversions
explained by the locations of the wells. The upper valley vineyard and the start of irrigation (20 September 2003). Between
is located in an area where groundwater is recharged by seepage 15 September and 30 October of 2005 there was 2.2 cm of

237
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

rainfall (Figure 6(d)); the rise of the water table is thus due to locations due to evapotranspiration demands. Further discussion
irrigation. Well 1 was located 110 m from the irrigation canal and of these results is presented by Rivera et al. (2005).
it shows a continuous increase in the water table with a peak
observed 22 days after the opening of irrigation diversions. Well In natural conditions, the recharge processes in the Peumo area
2 presents a sharp rise of the water table at the beginning of the and similar zones in the region would involve winter surface and
irrigation season but the peak was observed 28 days after the subsurface runoff from the Coastal Range uplands into the valley
opening of irrigation diversions. Both wells were located on the floor. This runoff would recharge groundwater in winter, leading
same line perpendicular to the irrigation canal. Well 2 received to higher water tables. In the summer dry season, plant evapo-
the effect of irrigation seepage before well 1 because it was transpiration demand and lack of rainfall or upland runoff would
located at the centre of the irrigated field; however, it received lead to lowered water tables. However, in the Peumo area, where
the effect of canal seepage later because it is located at 390 m there is now a highly developed irrigation canal network, the
from the irrigation canal. Although the peaks (Figure 6(c)) large main irrigation canals, which are empty in winter and
represent different travel times of seepage pulses, the water table located at the piedmont, intercept upland runoff and transport the
decline corresponds to exactly the same period at different water to the river at the lower end of the zone (Figure 7(a)). This

Rain

Runoff

Piedmont artificial canals

Groundwater recharge by rainfall

Coastal Mountains Cachapoal River

Groundwater system

(a)

Piedmont artificial canals

Irrigated crop land

Coastal Mountains

Cachapoal River

Groundwater system

(b)

Figure 7. Peumo Valley conceptual model including interaction

between irrigation canals, groundwater and the river in (a) winter
condition with empty canals that intercept upland runoff and
groundwater recharge by rain and (b) irrigation season condition
with groundwater recharge by seepage from irrigation canals and
irrigation losses

238
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

runoff interception by canals considerably reduces the valley controlled by the river head maintaining equilibrium with the
aquifer recharge from upland runoff. In spring and summer, once water table. From the location of the avocado orchard to the end
the irrigation canals are full of water, seepage of excess irrigation of the valley (Figure 4) the river receives seepage from ground-
recharge (Figure 7(b)) leads to higher water tables in late spring water (Figure 8(a)). Seepage from the main irrigation canals is
(Oyarzún et al., 2007). the main source of groundwater recharge in the upper section of
the area. This produces high water table levels in the upper third
5. Model results of the valley (for instance at the avocado orchard of Figure 4),
The model results show that groundwater flow is strongly affected forcing farmers to install drainage systems to lower the water
by the valley longitudinal gradient (0.3%). At the upper part of table to be able to produce crops. Meanwhile, at the lower end of
the valley and north of the river, the groundwater level is the Peumo area, where the water table is above the stream level,

6 210 000
Irrigation canal
Estero Taguilla
6 205 000 Cachapoal River
Groundwater flow
direction
5
12
6 200 000
North: m

150
6 195 000 175

6 190 000

275 000 280 000 285 000 290000 295 000 300 000
East: m
(a) 1 2 3 4 5 6 7

6 210 000 6 210 000

6 205 000 6 205 000

1
North: m

North: m

6 200 000 6 200 000

1
1

6 195 000 6 195 000

1

6
1

6 190 000 6 190 000

275 000 280 000 285 000 290 000 295 000 300 000 275 000 280 000 285 000 290 000 295 000 300 000
East: m East: m
(b) (c)

Figure 8. (a) Steady-state simulated water table and

groundwater flow direction for the Peumo Valley. (b) Water table
drawdowns with respect to the modelled steady-state conditions
if wells (black dots) are pumped at full capacity. (c) Water table
with respect to the modelled steady-state conditions if the main
irrigation canals are lined and wells (black dots) are pumped at
full capacity

239
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

several riverbank springs were observed. From an evaluation of water levels, affecting small farmers who depend on shallow
water balance at the model domain, it was found that, on average, wells.
the irrigation canals interact with the aquifer and contribute more
than 50% of groundwater recharge, while deep percolations from Water management decisions must be taken based on a sound
irrigation fields and rainfall infiltration each contribute 25%. knowledge of the hydrology of the system that will be affected.
Poor knowledge of the hydrological connections may produce
5.1 Analysis of drought scenarios unforeseen effects that could affect third parties such as the small
If farmers use the full capacity of their wells, the impact of such farmers of the Peumo Valley.
extractions is limited to the sectors of the study area, where wells
are allowed to pump larger volumes of water (Figure 8(b)). Acknowledgements
According to Chilean water law, the basic unit for groundwater The authors wish acknowledge the support given by the Chilean
management is a section of an aquifer, in this case the whole government through the project FONDEF D02I-1146 and the
study area. In a section of an aquifer, groundwater rights in the National Sciences Foundation of the United States through the
past were given in historical order, so the first farmer who asked project Acequia Water Systems Linking Culture and Nature:
for groundwater rights was able to ask for as much water as his Integrated Analysis of Community Resilience to Climate and
well was able to pump when it was constructed. There was no Land-Use Changes. NSF-Proposal 1010516.
restriction of this rule until a point was reached where the total of
the existing water rights were equivalent to the estimated REFERENCES
recharge on the section of the aquifer and, after that moment in Anderson MP and Woessner WW (1992) Applied Groundwater
that section of the aquifer, no more water rights would be given. Modeling. Academic Press, San Diego, CA, USA.
Thereafter, the construction of new wells in the valley was not Araneda L (2004) Land Use Assessment of the Peumo Area,
allowed and there was no possibility of more wells in the area. Using GIS. Agricultural Engineering final project report,
Due to this previous legal practice, the imbalance between local Universidad de Concepción, Chillan, Chile.
recharge and well pumping is not a surprise. Arumı́ JL, Rivera D, Holzapfel E et al. (2009) Effect of irrigation
canal network on surface and groundwater connections in the
If the main irrigation canals are lined, recharge from canal lower valley of the Cachapoal River, Chile. Chilean Journal
seepage will be reduced. That condition will produce a depletion of Agricultural Research 69(1): 12–20.
of water table levels in the areas where canal seepage feeds the Chiang WH and Kinzelbach W (2001) 3D-Groundwater Modeling
aquifer (Figure 8(c)). The reduction of recharge and full pumping with Pmwin: A Simulation System for Modeling Groundwater
of the existing wells will surely affect local small farmers who Flow and Pollution. Springer, New York, NY, USA.
support their production with shallow wells. DGA (2005) Evaluación de los recursos hı́dricos subterráneos de
la VI Región: modelación hidrogeológica de los Valles de
It proves paradoxical that lining, which is considered a logical Alhué, Cachapoal y Tinguiririca: informe técnico. Ministerio
measure to improve irrigation efficiency, could affect the long- de Obras Públicas, Dirección General de Aguas, Santiago,
term equilibrium of the hydrological system. In that sense, it is a Chile (in Spanish).
good example of the potential effects of management decisions Garreaud RD (2009) The Andes climate and weather. Advances
made on the basis of a poor understanding of the hydrological in Geosciences 7(1): 1–9, www.adv-geosci.net/7/1/2009/.
system. Hauser A (1990) Carta hidrogeológica de la VI Región. Servicio
Nacional de Geologı́a y Minerı́a (SERNAGEOMIN),
6. Conclusions Santiago, Chile.
The Peumo area irrigation canal network has a strong influence Holzapfel E, Jara J and Martinez G (2011) Evaluation of a micro-
on the hydrologic system of the valley, defining the spatial and sprinkler irrigation system at an apple farm in the Central
temporal patterns of groundwater recharge. Once the irrigation Valley of Chile. Acta Horticulturae 889(1): 499–506.
diversions are opened, they produce sustained elevated water Jara J, Holzapfel E, Quiñónez A and Soto M (2011) Evaluation of
tables due to irrigation canal seepage, contributing to more than drip irrigation systems for blueberry orchards in South
half of the groundwater recharge. Central Chile. Acta Horticulturae 889(1): 489–497.
McDonald MG and Harbaugh AW (1988) MODFLOW, A
For land and water managers planning to line irrigation canals, it Modular Three-dimensional Finite Difference Ground-water
is important to know how the reduction or removal of canal Flow Model. US Geological Survey, Reston, VA, USA,
seepage affects the interactions between surface water and Open-file report 83-875.
groundwater. The current shallow groundwater levels of the Montecinos A and Aceituno P (2003) Seasonality of the ENSO-
valley depend mainly on recharge from irrigation canal seepage. related rainfall variability in central Chile and associated
Modifications to the canals, such as lining them with imperme- circulation anomalies. Journal of Climate 16(2): 281–296.
able materials, could lead to a reduction of groundwater recharge Montecinos A, Dı́az A and Aceituno P (2000) Seasonal diagnostic
and changes in crop production patterns due to lowered ground- and predictability of rainfall in subtropical South America

240
Water Management Effect of drought on groundwater in a
Volume 166 Issue WM5 Chilean irrigated valley
Arumı́, Rivera, Holzapfel and Muñoz

based on tropical Pacific SST. Journal of Climate 13(4): actividad agrı́cola en el valle de Peumo, Chile. Revista
746–758. Gestión Ambiental 11(1): 59–80 (in Spanish).
Nuñez CJ and Soto G (2010) Recursos Hı́dricos en Chile. Centro Rivera D, Arumı́ JL and Holzapfel E (2007) Efecto de la red de
del Agua para America Latina y el Caribe, La Serena, Chile canales y sistemas de riego en la hidrologı́a del valle de
(in Spanish). Peumo, Chile. Revista Ingenierı́a Hidráulica en México
Orrego X (2003) Caracterización del sistema hidrológico en el XXII(4): 115–119 (in Spanish).
valle del Rı́o Cachapoal, sector Peumo-las Cabras, usando Rodrı́guez M, Alfaro G, Rojas C, Canales A and Fajardo J
SIG. 52 p. Memoria para optar al tı́tulo de Ingeniero Civil (2002) Evaluación Geológica del Metalotecto ubicado
Agrı́cola, Universidad de Concepción, Chillan, Chile (in entre el Estero Las Palmas y San Vicente de Tagua-
Spanish). Tagua, VI Región, Chile. Universidad de Concepción-
Oyarzún R, Arumı́ JL, Alvarez P and Rivera D (2007) Water use in Secretarı́a Regional de Minerı́a VI Región, Rancagua,
the Chilean agriculture: current situation and areas for Chile (in Spanish).
research development. In Agricultural Water Management Uribe H (2008) Efecto de la Agricultura Intensiva de Riego Sobre
Research Trends (Columbus F (ed.)). Nova Science el Régimen Hidrológico Superficial en Cuencas del Valle
Publishers, Hauppauge, NY, USA. Central de la VII Region, Chile. PhD thesis, Universidad de
Rivera D, Arumı́ JL and Holzapfel H (2005) Impacto de la Concepción, Chillan, Chile (in Spanish).

WHAT DO YOU THINK?

To discuss this paper, please email up to 500 words to the
editor at [email protected]. Your contribution will be
forwarded to the author(s) for a reply and, if considered
appropriate by the editorial panel, will be published as a
discussion in a future issue of the journal.
Proceedings journals rely entirely on contributions sent in
by civil engineering professionals, academics and students.
Papers should be 2000–5000 words long (briefing papers
should be 1000–2000 words long), with adequate illustra-
tions and references. You can submit your paper online via
www.icevirtuallibrary.com/content/journals, where you
will also find detailed author guidelines.

241