Home Search Collections Journals About Contact us My IOPscience

Integrated crop water management might sustainably halve the global food gap

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2016 Environ. Res. Lett. 11 025002

(http://iopscience.iop.org/1748-9326/11/2/025002)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 182.73.193.34
This content was downloaded on 16/02/2016 at 11:30

Please note that terms and conditions apply.

 Environ. Res. Lett. 11 (2016) 025002 doi:10.1088/1748-9326/11/2/025002

LETTER

Integrated crop water management might sustainably halve the

OPEN ACCESS
global food gap
RECEIVED
30 October 2015
J Jägermeyr1,2, D Gerten1,2, S Schaphoff1, J Heinke1,3,4, W Lucht1,2 and J Rockström5
REVISED
15 January 2016
1
Research Domain Earth System Analysis, Potsdam Institute for Climate Impact Research (PIK), Telegraphenberg A62, 14473 Potsdam,
Germany
ACCEPTED FOR PUBLICATION 2
Geography Department, Humboldt-Universität zu Berlin, Unter den Linden 6, 10099 Berlin, Germany
25 January 2016 3
International Livestock Research Institute (ILRI), P O Box 30709, Nairobi, 00100 Kenya
PUBLISHED 4
Commonwealth Scientific and Industrial Research Organization (CSIRO), St. Lucia, QLD 4067, Australia
16 February 2016 5
Stockholm Resilience Centre, Stockholm University, Stockholm 106 91, Sweden
E-mail: jonas.jaegermeyr@pik-potsdam.de
Original content from this
work may be used under Keywords: sustainable intensification, yield gap, water harvesting, conservation agriculture, irrigation efficiency, food security, climate
the terms of the Creative
Commons Attribution 3.0 change adaptation
licence. Supplementary material for this article is available online
Any further distribution of
this work must maintain
attribution to the
author(s) and the title of
the work, journal citation
Abstract
and DOI. As planetary boundaries are rapidly being approached, humanity has little room for additional
expansion and conventional intensification of agriculture, while a growing world population further
spreads the food gap. Ample evidence exists that improved on-farm water management can close
water-related yield gaps to a considerable degree, but its global significance remains unclear. In this
modeling study we investigate systematically to what extent integrated crop water management might
contribute to closing the global food gap, constrained by the assumption that pressure on water
resources and land does not increase. Using a process-based bio-/agrosphere model, we simulate the
yield-increasing potential of elevated irrigation water productivity (including irrigation expansion
with thus saved water) and optimized use of in situ precipitation water (alleviated soil evaporation,
enhanced infiltration, water harvesting for supplemental irrigation) under current and projected
future climate (from 20 climate models, with and without beneficial CO2 effects). Results show that
irrigation efficiency improvements can save substantial amounts of water in many river basins
(globally 48% of non-productive water consumption in an ‘ambitious’ scenario), and if rerouted to
irrigate neighboring rainfed systems, can boost kcal production significantly (26% global increase).
Low-tech solutions for small-scale farmers on water-limited croplands show the potential to increase
rainfed yields to a similar extent. In combination, the ambitious yet achievable integrated water
management strategies explored in this study could increase global production by 41% and close the
water-related yield gap by 62%. Unabated climate change will have adverse effects on crop yields in
many regions, but improvements in water management as analyzed here can buffer such effects to a
significant degree.

1. Introduction appropriation of resources and conventional intensifi-

cation of agriculture (Steffen et al 2015). Inevitably,
Demand for food increases as populations grow and competition for energy, land and water rises with
gain wealth, thus the world might need a 60%-100% growing food demand, which fuels the challenge of
extra kcal production by 2050 to end hunger closing the global food gap (crop calorie requirements
(IAASTD 2009, Tilman et al 2011, Alexandratos and above domestic production and imports, now and in
Bruinsma 2012, Valin et al 2014). However, it is the future) (e.g. Godfray et al 2010, Foley et al 2011,
becoming increasingly apparent that planetary guard- Searchinger et al 2013). Climate change might exacer-
rails narrow down humanityʼs prospects for additional bate this situation by increasing water stress and

© 2016 IOP Publishing Ltd

Environ. Res. Lett. 11 (2016) 025002

(i.e. increasing the yield output per unit of water con-

sumed) in both rainfed and irrigated systems paired
with an increase in consumptive water use are a sine
qua non for raising food production to the tre-
mendous amount required (Molden 2007,
IAASTD 2009).
However, the attainable extent and potential of
integrated crop water management at the global level
under both current and future climates remains insuf-
ficiently quantified (e.g. Pretty et al 2011, Rost
et al 2009, IAASTD 2009, Brauman et al 2013). In this
global modeling study we investigate the potential to
Figure 1. Rainfall and irrigation water partitioning as growing increase yields through large-scale implementations of
season averages across global cropland, calculated with integrated crop water management (defined here as a
LPJmL (1980–2009). Blue numbers refer to irrigated systems
(precipitation plus irrigation withdrawal), green numbers mix of various farm water management
refer to rainfed systems. Evaporation also includes intercep- interventions).
tion losses. Note that water outflux can exceed growing season Particularly in semi-arid rainfed agriculture, sub-
rainfall due to soil moisture availability at planting. Produc-
tive consumption (i.e. transpiration) can be much lower ject to the largest water constraints to low yields, rain-
regionally; generalized estimates for rainfed systems in sub- fall variability (dry spells, periodic water scarcity) often
Saharan Africa are: transpiration 15%-30%, evaporation:
30%-50%, runoff: 10%-25%, seepage: 10%-30% (Rockström poses a much greater problem than the total amount
and Falkenmark 2015). Spatial patterns in transpiration of precipitation. In addition, in semiarid tropical sys-
coefficient simulated with LPJmL are displayed in figure S1. tems root zone drought and low yields (1-2 t ha-1) are
often caused by poor farm water management with
excessive on-farm water losses (Oweis and
Hachum 2006, Rockström et al 2007, Wani et al 2009).
hydroclimatic variability particularly in developing Accordingly, the transpiration coefficient (TC, crop
countries (Porter et al 2014, Rosenzweig et al 2014). transpiration per unit rain and withdrawn irrigation
Agriculture is the single largest user of freshwater water, figure 1) is often <30%, as non-productive soil
and the most important reason why the world is trans- evaporation can consume up to 50% on low-yielding
gressing planetary boundaries (Rockström and Karl- fields (Daamen et al 1995, Rockström 2003, Wani
berg 2010). The challenge of producing enough food et al 2009), and 10%–30% can be lost to surface runoff
becomes especially delicate, as it must be met mainly (Welderufael et al 2008, Araya and Stroosnijder 2010).
on currently cultivated land since expansion and con- These factors indicate key hydroclimatic opportu-
ventional intensification of agriculture comes at major nities. In fact, there is a portfolio of measures available
environmental costs (local to global scale factors: ero- to increase plant water availability through e.g. max-
sion, biodiversity loss, salinization, water pollution imizing soil infiltration, minimizing soil evaporation,
and eutrophication, water scarcity, greenhouse gas collecting surface runoff for supplemental irrigation,
emissions) (Matson et al 1997, Foley et al 2005, Rey- and improving irrigation systems (to expand irrigated
nolds et al 2015). Furthermore, significant yield gaps areas using saved water). Supplemental irrigation
exist across various farming systems, indicating a sub- during dry spells can trigger important positive pro-
stantial scope for yield gains through mitigation of duction shifts (Fox and Rockström 2003, Biazin
nutrient and water deficiencies (Mueller et al 2012, et al 2012, Burney et al 2013), and water harvesting
Licker et al 2010, Global Yield Gap Atlas 2015). (WH) and soil moisture conservation (SMC) techni-
Increasing production on existing agricultural ques can double smallholder yields in drought-prone
land by managing available resources more efficiently, regions while at the same time improving resilience to
placing less pressure on the environment and sustain- climate risks (Rockström et al 2003, Oweis and
ing future capacities, i.e. sustainable intensification, is Hachum 2006, Dile et al 2013). These long-known
thus seen as an important part of a solution and high practices are being implemented sporadically around
on the global policy agenda (Tilman 1999, The Royal the world, leaving open vast potential to scale up (Bar-
Society 2009, Garnett et al 2013, World Bank 2013, ron et al 2015, Searchinger et al 2013, Mati et al 2007).
Dobermann and Nelson 2015). The renewed Sustain- Irrigated farming systems on the other hand, are the
able Development Goals now stipulate sustainable single largest global user of water abstractions (80%-
agriculture as an agreed goal among all nations (Uni- 90% of consumption), but they use water often ineffi-
ted Nations 2016), but there is little quantitative evi- ciently (Gleick et al 2009, Molden 2007). Irrigation
dence of how to achieve it. While most global improvements have the potential to save and redis-
strategies focus on improving soil fertility, Rockström tribute water to underperforming systems (Rockström
and Falkenmark (2015) urge an international high- et al 2007, Kijne et al 2009, Brauman et al 2013, Jäger-
level consideration of integrated crop water manage- meyr et al 2015, Fishman et al 2015). In particular the
ment. In fact, such water productivity improvements combination of such measures, i.e. integrated farm

2
Environ. Res. Lett. 11 (2016) 025002

water management proved successful to boost yields ambition levels for harvesting runoff during the
across various farming systems (Oweis and growing season: on 10, 25, 50, and 85% of rainfed
Hachum 2003, Molden 2007, Mazvimavi et al 2008). cropland in each grid cell with a maximum storage
However, the potential significance of integrated capacity of 200 mm, respectively. Case studies support
crop water management at the global level remains similar up-scaling potentials across watersheds using
unclear, because upscaling is a challenge given the het- gravity-fed and pump-based SI (Kahinda et al 2008,
erogeneity of farming systems and downstream water Barron et al 2015). We define a rather high storage
trade-offs (e.g. Falkenmark et al 2001, Ngigi 2003, capacity to allow evaluating SI potentials, despite
Pretty et al 2011, Dile et al 2013). Only few studies have challenges to its large-scale implementation (in section
used the capacity of modeling approaches in evaluat- 3.6. we show that 100 mm suffice in 95% of all cases).
ing complex interactions of up-scaling water manage- Water is assumed to be reapplied on the same land
ment interventions (e.g. Tsubo and Walker 2007, where it was collected, if (i) root available relative soil
Kahinda et al 2007, Wisser et al 2010, Barron moisture <40% of field capacity, (ii) daily precipita-
et al 2015). Lebel et al (2015) quantify WH potential tion is below 5 mm, and (iii) soil water supply falls
for maize in the whole of Africa using an empirical short of soil water demand. Sensitivity analyses for the
approach. Rost et al (2009) simulate effects of WH and cistern size and irrigation threshold are displayed in
SMC on global crop NPP with the dynamic agro- figure S2.
hydological model used herein. A knowledge gap
remains, to provide a global assessment of integrated 2.1.2. In situ water harvesting (WHin)
water management in rainfed and irrigated agriculture Micro-catchment systems, e.g. pitting, terracing,
and using a large ensemble of climate change micro-basins, but also conservation tillage, and
scenarios. mulching can hinder water from running off the field
This study investigates systematically the global and thus help increasing infiltration capacity. Particu-
potential of integrated crop water management larly the combination of micro-catchments and
through implementing the most approved interven- mulching is observed to reduce runoff and soil
tions into the dynamic global bio-agrosphere model, evaporation considerably (Botha et al 2007, Biazin
LPJmL. We present a process-based simulation of et al 2012). In LPJmL, infiltration rate In depends on
crop yields with high spatial, temporal and agronomic soil properties, current soil moisture and the infiltra-
detail, explicitly accounting for downstream effects tion parameter p. By default, without management
and catchment hydrology. The study shows by how interventions, p=2, but here we also simulate
much (i) global crop production could be intensified increased infiltration rates assuming four different
sustainably (in terms of not using additional water or intensity levels ( p = 3, 4, 5, 6).
land inputs), (ii) the water gap (see figure 2) could be In is calculated for the upper soil layer as:
closed, and (iii) these opportunities might buffer wa
potential climate change impacts, assuming various In = prir ´ p 1- , (1)
Wsat - Wpwp
ambition levels for large-scale adoption of integrated
crop water management. where prir is daily rain and applied irrigation water, wa
is the available soil water content, and Wsat and Wpwp
are soil water content at saturation and wilting point,
2. Methods respectively (all in mm). A sensitivity analysis for p is
displayed in figure S2. Hereinafter, WH refers to the
The representation of water harvesting, soil moisture combination of WHin and WHex measures at the four
conservation and irrigation transitions in LPJmL is respective ambition levels.
outlined first (summarized in table 1), followed by
basic characteristics of the model and the simulation 2.1.3. Soil moisture conservation (SMC)
setup. Non-productive soil moisture depletion can be alle-
viated through organic or plastic film mulching, and
2.1. Simulation of water management interventions different conservation tillage systems. These techni-
2.1.1. Ex situ water harvesting (WHex) ques can improve grain yield remarkably through
This measure describes the concentration, collection, conserving soil moisture for additional plant tran-
and storage of surface runoff in ponds or cisterns for spiration, suppressing weeds, and improving cold
supplementary irrigation (SI) during dry spells. Reser- tolerance (Liu et al 2014). Organic crop residues
voirs are often sized to provide about 100–200 mm SI covering 50% of the soil surface can reduce soil
(Biazin et al 2012, Barron and Okwach 2005, Oweis evaporation by ~25%, plastic mulching can reduce
and Hachum 2006). Its implementation is site-speci- soil evaporation by ~50%–90% (Bos et al 2007, Bu
fic, depending on various biophysical, economic and et al 2013). In our simulations, we reduce soil
social factors (Barron et al 2015, Studer and Lin- evaporation on rainfed and irrigated cropland during
ger 2013). We therefore simulate in LPJmL four the growing season by 10, 25, 50, and 85%, respectively

3
Environ. Res. Lett. 11 (2016) 025002

Rainfed water gap Irrigated water gap

(a) (b)

0 5 10 20 30 40 50 60 >
Water gap [%]

Figure 2. ’Water gap’ (i.e. the gap between current and potential yields in the absence of water constraints) at basin level simulated by
LPJmL for rainfed (a) and irrigated agriculture (b), averaged for the time period 1980 to 2009. Global area-weighted averages are 29%
for rainfed and 6% for irrigated systems.

(applied as a simple factor to the evaporation calcul- for further investigation and for climate change
ation). SMC is also applied on irrigated fields and simulations:
therefore helps saving irrigation requirements. As it is
not affecting downstream water availability, SMC can (1) ‘Low’: ‘50% surface’ irrigation scenario + 25%
be considered a ‘crop per drop’ improvement also at SMC + 25% WH + irrigation expansion with
basin scale. saved water.
(2) ‘Ambitious’: ‘best practice’ irrigation scenario +
2.1.4. Irrigation improvements and expansion with
50% SMC + 50% WH + irrigation expansion with
saved water (IRRexp)
saved water.
Irrigation is represented through mechanistic simula-
tion of surface, sprinkler and drip systems, depending (3) ‘Max’: ‘all drip’ irrigation scenario + 85% SMC +
on country and crop type. System efficiencies are 85% WH + irrigation expansion with saved water.
directly linked to vegetation dynamics, weather and
soil conditions, and water availability (Jägermeyr It is worth to highlight that both the ‘max’, and ‘all
et al 2015). To simulate irrigation improvements, we drip’ scenarios are designed to evaluate planetary bio-
define three theoretical transition scenarios: physical limits, not to represent feasible transition
targets.
(1) ‘50% surface’, half of non-paddy surface irrigation
is replaced by more efficient sprinkler systems. 2.2. LPJmL model
The model LPJmL globally represents biogeochemical
(2) ‘Best practice’, drip systems are established
land surface processes of vegetation and soils (Bon-
where applicable (based on crop suitability, (see
deau et al 2007, Fader et al 2010, Jägermeyr et al 2015),
Jägermeyr et al 2015), the remainder is under
simulating daily water and carbon fluxes in direct
sprinkler irrigation, but paddy rice remains with
coupling with the establishment, growth, and produc-
surface systems.
tivity of major natural and agricultural plant types at
(3) ‘All drip’, drip irrigation on all irrigated land. 0.5° resolution.
Agricultural land is represented by 12 specified
Improving irrigation performances can release crop functional types (CFTs), a class ‘others’ that
water which in turn can be exploited for expanding the includes a suite of crops collectively parameterized as
target area. To calculate expansion potentials, we only annual crops, and pastures (Bondeau et al 2007). All
consider saved water that otherwise would have been CFTs are either irrigated or rainfed and its spatial dis-
consumed non-beneficially, as irrigation return-flows tribution and their irrigated fraction is prescribed as in
are often crucial for downstream water availability. Jägermeyr et al (2015).
Expansion of irrigated land is assumed to be further Assimilated carbon (in the process of photosynth-
constrained by current rainfed cropland within a river esis) is allocated to harvestable storage organs (e.g. cer-
basin. Table 3 presents global numbers of expansion. eal grain) and three other pools (roots, leafs, stems).
Sowing dates are dynamically calculated based on cli-
2.1.5. Integrated water management mate and crop type (Waha et al 2012). Crops are har-
In addition to individual simulations of water manage- vested when they reach maturity, defined either
ment interventions (sections 2.1.1–2.1.4), we run through a CFT-specific maximum value of daily accu-
cross-combinations of WH, SMC, and IRRexp (see mulated phenological heat units or expiration of the
table 2), from which we select three pointer scenarios growing season. Storage organs are subsequently

4
Environ. Res. Lett. 11 (2016) 025002

Table 1. Selection of water (and soil) management interventions simulated in this study.

Name Goal Measure LPJmL implementation Rainfed / irrigated

Soil moisture Alleviation of Mulching (organic Soil evaporation during growing Rainfed and irrigated cropland
conserva- non- residues, plastic season reduced by 10%–85%
tion (SMC) productive films), conservation
soil moisture tillage
depletion

ex situ water Supplementary Collecting surface Surface runoff during growing season collected on Rainfed cropland
harvesting irrigation runoff in 10%-85% of cropland (storage capacity 200 mm),
(WHex) for designated catch- suppl. irrigation if soil moisture <40% of field
dry spell ment area, storage in capacity
mitigation cisterns etc., supple-
mentary irrigation

in situ water Maximizing soil Pitting, contouring, Increased infiltration rate Rainfed cropland
harvesting infiltration terracing, micro- (see equation (1))
(WHin) capacity and redu- basins, plowing, crop
cing non-produc- residues, conserva-
tive surface runoff tion tillage

Irrigation Reducing non- Improving perfor- Replacing surface irrigation with Irrigated cropland, expanding
improve- productive water mance of irrigation sprinkler or drip systems into rainfed with saved water
ment (IRR) consumption and systems
using thus saved
water for
expansion

removed from the field. Root growth and distribution monthly climatology for temperature, cloudiness
within soil layers is CFT-specific, while the soil profile (Harris et al 2014) and with the Global Precipitation
is discretized into 5 hydrologically active layers Climatology Centres (GPCC) precipitation data
(Schaphoff et al 2013). (Schneider et al 2014). The number of monthly rain
Plant growth is currently not directly nutrient-lim- days was derived from CRU and GPCC data as
ited in LPJmL, yet constrained by temperature, radia- described in Heinke et al (2013). To cover uncertain-
tion, water and atmospheric CO2 concentration. We ties in climate change simulations (2009–2099), we
calibrate crop yields with national FAO statistics based considered four representative concentration path-
on three model parameters (as in Fader et al 2010) to ways (RCPs: 2.6, 4.5, 6.0, 8.5), each being represented
account for CFT-specific management intensities. by 20 global climate models (GCMs) obtained from
LPJmL partitions precipitation and applied irriga- the CMIP5 multi-model ensemble dataset (table S2)
tion water into interception, transpiration, soil eva- (Taylor et al 2012). Monthly GCM output was
poration, soil moisture, and runoff. Surplus water that bilinearly interpolated and bias-corrected to the refer-
cannot infiltrate generates surface runoff. Subsurface ence period 1970–2000 using a method adapted from
soil water above saturation runs off in lateral direction, Watanabe et al (2012). To analyze the CO2 effect on
while remaining soil water above field capacity perco- crop growth, each simulation was performed with
lates to the layer beneath, depending on its soil water constant (at year 2000) and transient CO2 concentra-
content and hydraulic conductivity. Surface and sub- tion. Model runs follow a 1000-year spinup (recycling
surface runoff are accumulated along the river net- the first 30 years of input climatology) and sowing
work and subsequently available for downstream dates are fixed during the simulation period after 1960
reuse. to allow the comparison of water management poten-
A recently implemented mechanistic irrigation tials between different runs and otherwise they would
module provides the framework for irrigation transi- represent a form of adaptation not intended here.
tions (Jägermeyr et al 2015). In addition, we account Spatially explicit global information on cropland
for household, industry and livestock water use and extent is obtained from the MIRCA2000 land-use
include a representation of dams and reservoirs to dataset (Portmann et al 2010). The extent of areas
improve the simulation of available surface water (Bie- equipped for irrigation from 1900–2005 is imported
mans et al 2011). from Siebert et al (2015) and the distribution of irriga-
tion systems from Jägermeyr et al (2015). Land use
2.3. Simulation protocol patterns are fixed after the year 2005. Irrigation with-
For the time period 1901–2009, we ran LPJmL forced drawal is constrained by local, renewable water sto-
with the Climate Research Units (CRU) TS 3.1 rage, i.e. there is no implicit assumption about

5
Environ. Res. Lett. 11 (2016) 025002

Table 2. Overview of model simulations. RCP = representative concentration pathways, GCMs=global climate models.

Reanalysis climate 1901–2009 Simulations

ACT Current management calibrated with FAO data, reference run for all other simulations. 1
POT Potentially achievable yields under unconstrained water availability (nutrient deficiencies remain) 1

SMC Soil evaporation reduced by 10, 25, 50, and 85%. 4

WHex Surface runoff collected by 10, 25, 50, and 85% 4
WHin Infiltration rate increased in four sequential steps (equation 1) 4
WH ex situ and in situ WH combined 4

IRR Irrigation improvements: ‘50% surface’, ‘Best practice’, ‘All drip’ 3

IRRexp Irrigation expansion using saved consumptive water from IRR 3
IRRexp + SMCexp Irrigation expansion using saved consumptive water from IRR and SMC implementation 3

Combined Cross-combinations of SMC, WH and IRR 12

Combinedexp Cross-combinations of SMC, WH and IRRexp + SMCexp (include the ‘low’, ‘ambitious’, ‘max’ 12
scenario)

Climate change 2009–2099: 4 RCP scenarios, 20 GCMs each, constant and transient CO2 each

CC Climate change impact 160

CC + manage Climate change plus water management scenarios: ‘low’, ‘ambitious’, ‘max’ 480
Σ 688

contributions from fossil groundwater or diverted riv- and include large parts of the Middle East, central Asia,
ers. Only potentially achievable yields (figures 2 and 4) North China plains, Australia, southern Africa, and
are simulated under unrestricted water availability. the western United States. Based on LPJmL, current
global rainfed farming operates with a water gap of
29% relative to its unconstrained water potential (yet
3. Results neglecting nutrient deficiencies). In the ‘low’, ‘ambi-
tious’ and ‘max’ scenario, this gap could be closed up
3.1. Effects of integrated water management on crop to 17%, 11% and 5%, respectively (figure 4).
production, transpiration coefficient, and the The irrigation water gap is necessarily smaller than
water gap the rainfed, as irrigation largely closes the gap. Under
Simulated crop water management increases global current conditions, global irrigated farming is simu-
kcal production by 41% under the ‘ambitious’ sce- lated to be only 6% short of its unconstrained water
nario (all measures combined, including irrigation potential (spatial patterns in figure 2(b)). While better
expansion), while using existing agricultural land, yet water management can further narrow this gap (local
cutting irrigation abstractions. Production increases significance), important benefits at the global level are
by more than 55% in many river basins between the associated with water savings (figure 4).
Middle East, central Asia, China, Australia, southern
Africa and North and South America (figure 3(a)). 3.2. Water savings potentials of irrigation systems
Under the ‘low’ and ‘max’ scenario global kcal Figure 5(a) confirms that improved irrigation and
production increases by 18 and 60%, respectively SMC implementations can only marginally increase
(figure 3(b)). Individual effects of irrigation transitions irrigated production at the global level (by <2%).
(IRR), soil moisture conservation (SMC) and water More importantly however, these measures (‘low’,
harvesting (WH) are specified below (section 3.2. and ‘ambitious’, ‘max’ scenario without expansion) show
the following). the potential to cut consumptive losses (i.e. soil
On current farmland, we calculate the average evaporation, interception, and evaporative convey-
transpiration coefficient (TC) at 46% for rainfed and ance losses) by respectively 24%, 48%, and 85%
33% for irrigated systems (figure 1). Simulated water (figure 5(b)). This results in significant reductions of
management significantly shifts water towards tran- global irrigation withdrawal from currently 2507 km3
spiration through alleviating soil evaporation, surface to 2071, 1248 and 808 km3 (figure 4), because
runoff and irrigation losses. Therefore, the TC (com- alleviated soil evaporation and higher conveyance and
bined for rainfed and irrigated systems) increases from application efficiencies strongly reduce irrigation
42% to 49%, 54%, and 61%, respectively, in the ‘low’, requirements.
‘ambitious’, and ‘max’ scenario.
Production gains are particularly steep in regions 3.3. Irrigation expansion with saved irrigation losses
currently experiencing large water gaps. Figure 2(a) These water savings would theoretically allow for an
highlights basins where the water gap exceeds 50%, additional 90, 345, and 597 Mha expansion into

6
Environ. Res. Lett. 11 (2016) 025002

rainfed cropland, respectively, for the ‘low’,‘ambi- strong fertilization effect (in LPJmL not directly
tious’, and ‘max’ scenario. These numbers are sub- constrained by nutrient limitation) actually increases
stantial in perspective of current irrigated land of global production by 4.3% (RCP 2.6) to 13% (RCP
about 300 Mha and the expected slow expansion pace. 8.5) despite concurrent climate impacts. A ‘moderate’
But future irrigation expectations are curbed due to CO2 fertilization (mean of constant and transient CO2
land constraints under current system efficiencies simulations) suggests marginal global production
(Alexandratos and Bruinsma 2012); farmers who changes (−2.6% to 1.6%). Regionally however, India,
pursue efforts to save water often use it to expand their Pakistan, west Australia, African Sahel, and east Brazil
irrigated share of cropland (Fishman et al 2015). face negative changes from −5% to < −20% under
Global total kcal production (rainfed + irrigated) RCP 2.6 (‘moderate’ CO2), while strong increases
could thereby increase by 7%, 26%, and 43% with occur in large parts of Russia, east and southern Africa,
considerably higher numbers in specific basins parti- and parts of central and south America (5% to >20%).
cularly between the middle East, large parts of Asia, In an RCP 8.5 world (‘moderate’ CO2) the Mediterra-
and Central to North America (figure 5(b); aggregated nean region, major parts of the United States and
to the basin level, as upstream irrigation improve- Mexico, and southern Asia appear additionally on the
ments can have water trade offs downstream). Note map with distinct negative changes, and many basins
that irrigation expansion (with higher efficiencies), but show kcal declines from −10 to < −30% (figure 7(c)).
also SMC and WH, lead to higher productive plant In the ‘low’ scenario, most adverse climate change
transpiration, which increases global irrigation water impacts are simulated to be buffered in a RCP 2.6
consumption from currently 1268 km3 to 1350, 1515, world (figure 7, see figure S4 and S5 for constant and
and 1607 km3 , respectively (table 3, ‘low’, ‘ambitious’, transient CO2). The ‘ambitious’ scenario can ease
‘max’ scenario), while non-productive losses still negative impacts in an RCP 8.5 world in many basins,
decrease (not all saved water used up for expansion as but some regions, notably east Brazil and west Africa,
some basins lack sufficient available rainfed cropland, remain with negative impacts. Despite large uncer-
figure 5(a)). Overall, the total global withdrawal tainties associated with the CO2 effect, global crop
amount is simulated to decrease by 128, 448, and 689 production is simulated to increase by >40% under
km3 for the three respective scenarios, despite the ‘ambitious’ water management for all but the most
growth of irrigated areas (table 3). severe climate change scenario (35% with RCP8.5,
table 4).
3.4. Effects of soil moisture conservation and water
harvesting on rainfed systems 3.6. Evaluation of results and modeling issues
SMC shows considerable potential to amplify rainfed Supplemental irrigation, mulching, and conservation
kcal production (3%–14% globally, figure 6(a)). tillage demonstrably increased yields in case studies by
Regions with high sensitivity are concentrated in semi- 56%, 44%, and 30%, respectively (Araya and Stroos-
arid to arid regions such as the Sahel, southern Africa, nijder 2010, Welderufael et al 2008, Fox and Rock-
central Asia, and Australia, where production ström 2003). A major case study (286 projects in 57
increases reach >20% (figure 6(b), no downstream countries) documents average yield increases by 79%
effect and thus displayed at the grid cell level). As for through a number of conservation agriculture inter-
water harvesting, WHex exhibits much higher produc- ventions, including water harvesting and conservation
tion potentials compared to WHin, but combining tillage (Pretty et al 2006). A wider collection of case
both measures appears especially beneficial at low studies shows similar ranges (table S1). Lebel et al
intensity levels, and could increase global kcal produc- (2015) simulate for maize in sub-Saharan Africa (SSA)
tion by 7%–24% (figure 6(c)). Figure 6(d) shows an average yield increase with WH intervention of 9%-
spatial patterns of WH (50% level) with high sensitivity 39%. We arrive at 0%–33% across SSA basins. In our
also in semi-arid regions, but in addition in sub- simulation (50% level), on average 57 mm supple-
humid regions with high rainfall variability and runoff mental irrigation are applied during the growing
excess, across tropical and temperate regions. season (in 95% of all cases less than 100 mm, figure
S3), Fox and Rockström (2003) document 60-90 mm
3.5. Climate change impact in a Burkina Faso case study. More generally we can
Climate change is simulated to have adverse effects on reproduce the documented scale of observed yield
global crop production, while high uncertainty is gains using LPJmL, and our analysis extends case study
associated with the direct effect of carbon dioxide on insights to a broader set of climates, locations, and
plant growth. In simulations with constant CO2 crops and thus refines management potentials at the
concentration (performed to isolate the climate global level (Rost et al 2009).
change effect), global kcal production is projected to We point out that it is critical to evaluate the local
change by −3% (RCP 2.6) to −18.2% (RCP 8.5, table 4, feasibility of WH catchment and storage systems. This
median of 20 GCMs), mostly due to increased water depends on different factors (e.g. terrain type, soil
deficiencies. With transient CO2 concentration, a structure, hydroclimatic setting, social and financial

7
Environ. Res. Lett. 11 (2016) 025002

Figure 3. Potential for increasing global kcal production through integrated crop water management (panel a, ‘ambitious’ scenario).
Global sums of kcal production for various simulated water management scenarios (combinedexp, table 2) is shown in (b), with bars
representing the irrigation scenario and stacks indicating the intensity of soil moisture conservation (SMC) and water harvesting
(WH). Indicated ‘low’, ‘ambitious’, and ‘max’ pointer scenarios derive from the combination of each irrigation scenario with the
respective SMC and WH intensity (defined in section 2.1.5), all for the time period 1980 to 2009.

Figure 4. Possible closure of the water gap through crop water management in rainfed and irrigated systems. The gap is calculated as
the difference between achieved production (colored circle) compared to potential production (white circle) for different
management scenarios. Global irrigation withdrawal is indicated by bottom italic numbers, while the inner white ring illustrates the
proportion to ‘actual’ withdrawal. Irrigation expansion is not included. See figure 2 for spatial patterns of the water gap.

8
Environ. Res. Lett. 11 (2016) 025002

Figure 5. Effects of irrigation improvements (IRR) and soil moisture conservation (SMC) on crop production, consumptive irrigation
losses and the transpiration coefficient of irrigated cropland. Stacks in panel (a) show global sums for currently irrigated cropland (top
row) and for total cropland with expanded irrigation into rainfed areas (bottom row, SMC does not apply to rainfed systems in this
figure). The map (b) shows spatial patterns of changes in total kcal production with the ‘ambitious’ IRRexp + SMCexp scenario (’best
practice’ irrigation and 50% SMC and expansion with saved water), all for the time period 1980 to 2009.

Table 3. Global area (Mha) of rainfed and irrigated agricultural land Therefore, we regard the ‘low’ and ‘ambitious’ sce-
(including pastures) aligned with irrigation water consumption nario as potentially achievable, while the ‘max’ sce-
(IWC, km3 ) and irrigation withdrawal (IWD), for current land use
and the three scenarios of combined water management ‘low’, nario, locally proven though, indeed appears unlikely
‘ambitious’, ‘max’, including irrigation expansion (Combinedexp in to become implemented globally. Nevertheless, it pro-
table 2).
vides important insights into planetary biophysical
Rainfed Irrigated Total IWC IWD capacities. Furthermore, it is difficult to quantify the
[Mha] [ km3 ] extent to which farmers already adopted WH and
Actual 3984 297 4282 1268 2507
SMC measures. Although initial up-scaling projects
Low prove successful regionally (e.g. Zhu and Yuan-
‘50% surface’ 3895 387 4282 1350 2379 hong 2006), they still represent only marginal areas at
irrigation the global level.
+ 25% SMC Our estimates of irrigation withdrawal and con-
+ 25% WH
sumption agree well with previous estimates
Ambitious
‘Best practice’ 3639 642 4282 1515 2059 (FAO 2014, Döll et al 2014, Wada and Bierkens 2014),
irrigation albeit featuring much more process detail. Irrigation
+ 50% SMC expansion adds a noticeable share to production increa-
+ 50% WH ses simulated in this study (figures 3(b) and 5(a)).
Max
The expansion of irrigated crops replaces rainfed crops,
‘All drip’ 3388 894 4282 1607 1818
irrigation which results in a propagation of irrigated cropland into
+ 85% SMC pastures in some cells, as the share of irrigated pastures
+ 85% WH is generally low. However, in SSA only 5% of the crop-
land is under irrigation today, which first explains our
flat irrigation improvement potentials in SSA (figure 5
(b)), and second outlines scope for irrigation expansion
using currently untapped water resources (FAO 2005,
capital (Studer and Linger 2013, Falkenmark Burney et al 2013, Xie et al 2014).
et al 2001, Mati et al 2007)), addressing those in detail Finally, it is important to note that upstream IRR
is beyond the scope of this broad-scale study of bio- and WH interventions can lead to reduced return-
physical potentials. Although case studies show that flows and runoff, which can negatively affect water
often only 10%-20% of the land is unsuitable for WH availability downstream. Despite noticeable impacts
and SMC adoption (Barron et al 2015, Kahinda locally, gains at the basin level over-compensate losses
et al 2008, van Rensburg et al 2012), we might exagge- in all basins (figure S4). This appears beneficial from a
rate WH suitability. However, this is faced by our very food production perspective, but there is a clear need
conservative assumption on the catchment area that is for policies and institutional orders to regulate water
limited to existing cropland only, and which is often reallocations. In this context it is crucial to quantify
much larger in reality. But it appears infeasible to contributions from groundwater and water diver-
delineate additional suitable catchment areas with a sions, given the complex recharge and transboundary
sufficient degree of detail globally. issues involved.

9
Environ. Res. Lett. 11 (2016) 025002

SMC: 50% level

(a) (b)

25
20
Kcal production
change [%]

15
SMC

Management
intensity
85%

10
50%
25%
10%

5
0 SMC
WH: 50% level

(c) (d)
25
Kcal production

20
change [%]

15
WH

10
5
0

WH WH WH
ex situ in situ
0 5 10 15 20 25 30 >

Kcal production change [%]

Figure 6. Effects of soil moisture conservation (SMC) and water harvesting (WH) interventions on rainfed cropping systems: barplots
show global sums of kcal production change. Maps show spatial patterns of SMC (b) and WH (d), respectively at the 50% intensity
level for the time period 1980 to 2009.

4. Discussion crop water management targets. In fact, such targets

are outright missing from the recently passed sustain-
This study is the first to systematically quantify able development goals and from the supporting lit-
potential contributions of different strategies of erature (United Nations 2015, United Nations 2016).
farm water management to increase global crop Our study adds confidence that not targeting dedi-
production without increasing pressure on land and cated water goals means we are set to miss substantial
water boundaries. Based on spatially and temporally opportunities to advance a sustainable food system
detailed process-based modeling, we advance the and its climate resilience.
quantification of the global achievable scope of water On the way towards a sustainable food future,
management in rainfed and irrigated agriculture. water management is accompanied with essential co-
Simulated yield potentials are well in line with farm- benefits (that are not modeled here). Among the most
level experiences, but we exploit the dynamic model- important are reducing soil erosion through water
ing capacity of LPJmL for complex up-scaling of water harvesting and mulching, currently affecting67% ̃ of
interactions to arrive at robust global estimates. 41% SSA cropland (Liniger et al 2011). But large-scale
production growth, at global scale, released through implementations of plastic mulching can also lead to
‘ambitious’ water management outlines tremendous environmental pollution (Liu et al 2014). Better irriga-
opportunities. While grand challenges lie ahead to its tion technology helps reducing nutrients and pesti-
large-scale implementation, the ‘ambitious’ potentials cides application (better location and timing)
simulated here appear feasible from a biophysical and (Christian-Smith et al 2012, Calderón et al 2014),
also an agronomic perspective. More than 800 million while conservation agriculture in general will help
people today remain chronically undernourished mitigate greenhouse gas emissions (e.g. Mahdi
(United Nations 2015)—a kcal production gain of et al 2015, Karimi et al 2012, Liniger et al 2011). Water
40% realized by 2050 might be sufficient to halve the management that leads to stabilized water supply
widening global food gap, assuming that we need throughout the growing season is prerequisite for
60%-100% additional crop calories to eradicate hun- smallholders to invest in higher inputs (fertilizer,
ger (a gap of 80% roughly relates to 7.6 * 1015 kcal per breeds) (Biazin et al 2012, Burney et al 2013). Low-cost
year, compared to the production of 9.5 * 1015 kcal in interventions (organic mulching, conservation tillage,
2006 (Alexandratos and Bruinsma 2012, Searchinger simple drip kits) can directly translate in synergies in
et al 2013)). livelihoods; as most poor live in water-constrained
Although sustainable intensification appears high agriculture, the associated scope for poverty allevia-
on the policy agenda, there is a lack of institutionalized tion and improved local food security is tremendous

10
Environ. Res. Lett. 11 (2016) 025002

Table 4. Potential climate change impact (CC) on global crop production as against three scenarios of water management under four RCP
scenarios and different levels of CO2 fertilization; for the time period 2070–2099 versus 1980–2009, as averages across 20 GCMs.

RCP 2.6 RCP 4.5 RCP 6 RCP 8.5

const.1 moderate trans. const. moderate trans. const. moderate trans. const. moderate trans.

CC −3.0 0.7 4.3 −7.6 0.9 9.3 −9.4 1.6 12.7 −18.2 −2.6 13.0
CC + manage
‘Low’ 12.6 16.2 19.8 8.1 16.4 24.6 5.8 16.7 27.5 −3.8 11.3 26.4
‘Ambitious’ 38.4 42.2 46.1 33.1 41.8 50.5 30.8 42.2 53.7 18.9 34.9 50.9
‘Max’ 53.1 57.3 61.5 47.1 56.7 66.2 44.6 57.1 69.7 31.4 49.1 66.8

1
Note. const.: CO2 concentration fixed at year 2000, trans.: transient CO2 concentration, moderate: moderate CO2 effect, mean yields of
constant and transient CO2 .

Climate change Climate change and water management

RCP 2.6

(a) (b)
RCP 8.5

(c) (d)

< -30 -20 -10 -5 0 5 10 20 30 >

Change in kcal production [%]

Figure 7. Spatial patterns of potential climate change impact on global crop production under RCP 2.6 (a) and opposed to ‘low’ water
management (b); under RCP 8.5 (c) and opposed to ‘ambitious’ water management (d), all for the time period 2070 to 2099 vs
1980–2009 as averages across 20 GCMs and with ‘moderate’ CO2 effect (compare table 4). Corresponding maps for constant and
transient CO2 are presented in figure S4 and S5.

(Postel et al 2001, Dillon 2011, Pretty et al 2011, this study we show that such savings could be redir-
Kahinda and Taigbenu 2011, Burney and ected to support vast currently rainfed farms with
Naylor 2012). additional irrigation water. Although initial invest-
At the global scale, this study suggest that both ment needs are steep, long-term economic analyses
smallholder on-farm techniques and large-scale confirmed the substantial net profits achievable (Bia-
improvements of irrigation systems and WH imple- zin et al 2012, Fox et al 2005).
mentation are needed, while respecting environ- However, water management is not a panacea and
mental flow requirements of riverine ecosystems and needs to be combined with other components to sus-
other environmental boundaries. Our results show tainable farm management to exploit the strong
that large-scale adoptions of these measures lead into synergy between water, soil and nutrient management
water reallocations that would benefit from institu- (Oweis and Hachum 2006). Especially in SSA, many
tional support and water legislations as mentioned cropping systems are highly nutrient-deficient and
above (Molden 2007). Future investments must focus water management cannot fully take off, unless deple-
on enhancing system productivity on current arable ted soils become replenished (Sánchez 2010, Fox and
land, integrating management in rainfed and irrigated Rockström 2003). But it is clear that the challenge of
agriculture in an integrated landscape approach achieving sustainable food security is not only a sup-
(Faurès et al 2007, Rockström et al 2007). Jägermeyr ply-side problem. Urgent action is also needed on
et al (2015) show that technical irrigation saving holding down the growth in food consumption, redu-
potentials are substantial at the global level, while in cing waste, and achieve replacement level fertility

11
Environ. Res. Lett. 11 (2016) 025002

(Garnett et al 2013, Searchinger et al 2013, DeFries production in Northern Ethiopia Agricultural Water
et al 2015). Management 97 841–7
Barron J, Kemp-Benedict E, Morris J, de Bruin A, Wang G and
Fencl A 2015 Mapping the potential success of agricultural
5. Conclusion water management interventions for smallholders: Where are
the best opportunities? Water Resources and Rural
Development 6 24–9
This study quantifies the significance of integrated Barron J and Okwach G 2005 Run-off water harvesting for dry spell
crop water management at the global scale to intensify mitigation in maize (Zea mays L.): results from on-farm
rainfed and irrigated farming. Simulated measures are research in semi-arid Kenya Agricultural Water Management
74 1–21
constrained by the assumption that pressure on water Biazin B, Sterk G, Temesgen M, Abdulkedir A and Stroosnijder L
resources and land does not increase, which delineates 2012 Rainwater harvesting and management in rainfed
an effective strategy to minimize agricultural impacts agricultural systems in sub-saharan africa — a review Physics
on the biosphere. Based on detailed, process-based and Chemistry of the Earth, Parts A/B/C 47-48 139–51
Biemans H, Haddeland I, Kabat P, Ludwig F, Hutjes R W a,
simulation of underlying local biophysical conditions Heinke J, von Bloh W and Gerten D 2011 Impact of reservoirs
and with high spatio-temporal resolution, we system- on river discharge and irrigation water supply during the 20th
atically investigate scenarios of irrigation improve- century Water Resour. Res. 47 W03509
ments and expansion, water harvesting, and soil Bondeau A et al 2007 Modelling the role of agriculture for the 20th
century global terrestrial carbon balance Glob. Change Biol.
moisture conservation. Under a ‘low’ intensity sce- 13 679–706
nario we arrive at a global kcal gain of 18%. With an Bos M, Kselik R, Allen R and Molden D 2009 Water Requirements for
‘ambitious’, yet achievable scenario we reveal global Irrigation and the Environment (New York: Springer)
production potentials of 41%. Such water manage- Botha J J, Anderson J J, Groenewald D C, Nhlabatsi N N, Zere T B,
Mdibe N and Baiphethi M N 2007 On-farm application of in-
ment interventions would also about halve the current field rainwater harvesting techniques on small plots in the
global water gap in agriculture. Moreover, thus central region of south africa, volume 1 of 2: main report
improved water management offers the opportunity Technical Report (Gezina: Water Research Commission,
to buffer potential negative climate change impacts in Private Bag)
Brauman K A, Siebert S and Foley J A 2013 Improvements in crop
many world regions. The ‘low’ intensity scenario water productivity increase water sustainability and food
might over-compensate climate change impacts under security—a global analysis Environ. Res. Lett. 8 024030
relatively low RCP 2.6 emissions (globally + 40% kcal Bu L D, Liu J L, Zhu L, Luo S S, Chen X P, Li S Q, Lee Hill R and
production), while the ‘ambitious’ scenario could ease Zhao Y 2013 The effects of mulching on maize growth, yield
and water use in a semi-arid region Agricultural Water
most negative impacts in a RCP 8.5 world (globally +
Management 123 71–8
33% kcal production). Such kcal gain might be Burney J A and Naylor R L 2012 Smallholder irrigation as a poverty
sufficient to halve the global food gap by 2050. In alleviation tool in sub-saharan africa World Development 40
conclusion, this study highlights that not focussing on 110–23
Burney J A, Naylor R L and Postel S L 2013 The case for distributed
systematic implementation of integrated crop water
irrigation as a development priority in sub-saharan africa
management means to miss substantial opportunities Proc. Natl. Acad. Sci. USA 110 12513–7
in intensifying global farming systems within plane- Calderón F, Oppenheimer J, Stern N and Al E 2014 Better growth,
tary boundaries and to negotiate climate-associated better climate—the new climate economy report—the
risks in smallholder agriculture. synthesis report Technical Report (Washington, DC: the
global commission on the economy and climate)
Christian-Smith J, Cooley H and Gleick P H 2012 Potential water
savings associated with agricultural water efficiency
Acknowledgments improvements: A case study of California, USA Water Policy
14 194–213
This study was partly funded by the FACCE MACSUR Daamen C C, Simmonds L P and Sivakumar M V 1995 The impact
project (031A103B) and within the framework of the of sparse millet crops on evaporation from soil in semi-arid
niger Agricultural Water Management 27 225–42
Leibniz Competition (SAW-2013-PIK-5). We thank
DeFries R, Fanzo J, Remans R, Palm C, Wood S and Anderman T L
Matti Kummu for helpful comments. Moreover, we 2015 Metrics for land-scarce agriculture Science 349 238–40
acknowledge the World Climate Research Pro- Dile Y T, Karlberg L, Temesgen M and Rockström J 2013 The role of
grammeʼs Working Group on Coupled Modelling, water harvesting to achieve sustainable agricultural
intensification and resilience against water related shocks in
which is responsible for CMIP, and we thank
sub-saharan africa Agriculture, Ecosystems & Environment 181
the climate modeling groups (listed in table S2) 69–79
for producing and making available their model Dillon A 2011 The effect of irrigation on poverty reduction, asset
output. accumulation, and informal insurance: evidence from
northern mali World Development 39 2165–75
Dobermann A and Nelson R 2015 GSDR 2015 Brief: transformative
References changes of agriculture and food systems Technical Report
(United Nations)
Alexandratos N and Bruinsma J 2012 World agriculture towards Döll P, Müller Schmied H, Schuh C, Portmann F T and Eicker A
2030/2050: the 2012 revision Technical Report 12, FAO 2014 Global-scale assessment of groundwater depletion and
(Rome: FAO) related groundwater abstractions: combining hydrological
Araya A and Stroosnijder L 2010 Effects of tied ridges and mulch on modeling with information from well observations and
barley (hordeum vulgare) rainwater use efficiency and GRACE satellites Water Resour. Res. 50 5698–720

12
Environ. Res. Lett. 11 (2016) 025002

Fader M, Rost S, Müller C, Bondeau A and Gerten D 2010 Virtual Kijne J, Barron J, Hoff H and Rockström J 2009 Opportunities to
water content of temperate cereals and maize: present and increase water productivity in agriculture with special
potential future patterns J. Hydrol. 384 218–31 reference to africa and south asia Technical Report
Falkenmark M, Fox P, Persson G and Rockström J 2001 Water (Stockholm Environmental Institute)
harvesting for upgrading of rainfed agriculture Technical Lebel S, Fleskens L, Forster P M, Jackson L S and Lorenz S 2015
Report (Stockholm, Sweden: Stockholm International Water Evaluation of in situ rainwater harvesting as an adaptation
Institute) strategy to climate change for maize production in rainfed
FAO 2005 Irrigation in Africa in figures AQUASTAT Survey, africa Water Resources Management 29 4803–16
Technical report, Food and Agriculture Organization of the Licker R, Johnston M, Foley J a, Barford C, Kucharik C J,
United Nations, Rome, Italy Monfreda C and Ramankutty N 2010 Mind the gap: how do
FAO 2014 AQUASTAT database—Food and Agriculture climate and agricultural management explain the ‘yield gap’
Organization of the United Nations (FAO) (http://fao.org/ of croplands around the world? Glob Ecol. Biogeogr. 19
nr/water/aquastat/data/query/index.html?lang=en) 769–82
Faurès J-M, Svendsen M and Turral H 2007 Reinventing irrigation Liniger H, Studer R M, Hauert C and Gurtner M 2011 Sustainable
(Water for Food, Water for Life: a Comprehensive Assessment of land management in practice Technical Report (FAO)
Water Management in Agriculture) ed D Molden (London: Liu E K, He W Q and Yan C R 2014 ‘White revolution’ to ‘white
Earthscan, and Colombo: International Water Management pollution’ - agricultural plastic film mulch in China Environ.
Institute) ch 9, pp 353–94 Res. Lett. 9 091001
Fishman R, Devineni N and Raman S 2015 Can improved Mahdi S S, Dhekale B S, Choudhury S R and Bangroo S A 2015 On
agricultural water use efficiency save Indiaʼs groundwater? the climate risks in crop production and management in
Environ. Res. Lett. 10 084022 India: A review Australian J. Crop Sci 9 585–95
Foley J A et al 2005 Global consequences of land use Science 309 Mati B, De Bock T, Malesu M, Khaka E, Oduor A, Nyabenge M and
570–4 Oduor V 2007 Mapping the potential of rainwater harvesting
Foley J A et al 2011 Solutions for a cultivated planet Nature 478 technologies in Africa: a GIS overview on development
337–42 domains for the continent and ten selected countries.
Fox P and Rockström J 2003 Supplemental irrigation for dry-spell Technical manual No. 6 Technical Report (Kenya: World
mitigation of rainfed agriculture in the Sahel Agricultural Agroforestry Centre (ICRAF)), Netherlands Ministry of
Water Management 61 29–50 Foreign 10 Affairs, Nairobi)
Fox P, Rockström J and Barron J 2005 Risk analysis and economic Matson P A, Parton W J, Power G a and Swift M J 1997 Agricultural
viability of water harvesting for supplemental irrigation in intensification and ecosystem properties Science 277 504–9
semi-arid burkina faso and kenya Agricultural Systems 83 Mazvimavi K, Twomlow S, Belder P and Hove L 2008 An assessment
231–50 of the sustainable uptake of conservation farming in
Garnett T et al 2013 Sustainable intensification in agriculture: Zimbabwe. Global theme on agroecosystems report no. 39.
Premises and policies Science 341 33–4 Technical Report (Zimbabwe: International Crops Research
Gleick P H, Cooley H, Cohen M J, Morikawa M, Morrison J and Institute for the Semi-Arid Tropics, Bulawayo)
Palaniappan M 2009 The World’s Water 2008-2009: The Molden D (ed) 2007 Water for Food, Water for Life: A Comprehensive
Biennal Report on Freshwater Resources (Washington, DC: Assessment of Water Management in Agriculture (London:
Island Press) Earthscan, and Colombo: International Water Management
Global Yield Gap Atlas 2015 www.yieldgap.org/ Institute)
Godfray H C J, Beddington J R, Crute I R, Haddad L, Lawrence D, Mueller N D, Gerber J S, Johnston M, Ray D K, Ramankutty N and
Muir J F, Pretty J, Robinson S, Thomas S M and Toulmin C Foley J A 2012 Closing yield gaps through nutrient and water
2010 Food security: the challenge of feeding 9 billion people management Nature 490 254–7
Science 327 812–8 Ngigi S N 2003 What is the limit of up-scaling rainwater harvesting
Harris I, Jones P D, Osborn T J and Lister D H 2014 Updated high- in a river basin? Physics and Chemistry of the Earth, Parts A/B/
resolution grids of monthly climatic observations—the CRU C 28 943–56
TS3.10 dataset Int. J. Climatol. 34 623–42 Oweis T and Hachum A 2006 Water harvesting and supplemental
Heinke J, Ostberg S, Schaphoff S, Frieler K, Müller C, Gerten D, irrigation for improved water productivity of dry farming
Meinshausen M and Lucht W 2013 A new climate dataset for systems in West Asia and North Africa Agricultural Water
systematic assessments of climate change impacts as a Management 80 57–73
function of global warming Geoscientific Model Development 6 Oweis T Y and Hachum A Y 2003 Improving water productivity in
1689–703 the dry areas of west asia and north africa (Water Productivity
IAASTD 2009 International Assessment of Agricultural Knowledge, in Agriculture: Limits and Opportunities for Improvement) ed
Science and Technology for Development Global Report J Kijne et al (CAB International) ch 11, pp 179–98
(Washington, DC: Island) Porter J, Xie L, Challinor A, Cochrane K, Howden S, Iqbal M,
Jägermeyr J, Gerten D, Heinke J, Schaphoff S, Kummu M and Lobell D and Travasso M 2014 Food security and food
Lucht W 2015 Water savings potentials of irrigation systems: production systems Climate Change 2014: Impacts,
global simulation of processes and linkages Hydrol. Earth Adaptation, and Vulnerability: A. Global and Sectoral Aspects.
System Sci. 19 3073–91 Contribution of Working Group II to the Fifth Assessment
Kahinda J M, Lillie E, Taigbenu a E, Taute M and Boroto R 2008 Report of the Intergovernmental Panel on Climate Change ed
Developing suitability maps for rainwater harvesting in South C Field et al (Cambridge: Cambridge University Press)
Africa Physics and Chemistry of the Earth, Parts A/B/C 33 pp 485–533
788–99 Portmann F T, Siebert S and Döll P 2010 MIRCA2000—Global
Kahinda J, Rockström J, Taigbenu A and Dimes J 2007 Rainwater monthly irrigated and rainfed crop areas around the year
harvesting to enhance water productivity of rainfed 2000: A new high-resolution data set for agricultural and
agriculture in the semi-arid zimbabwe Physics and Chemistry hydrological modeling Glob. Biogeochem. Cycles 24 GB1011
of the Earth, Parts A/B/C 32 1068–73 Postel S, Polak P, Gonzales F and Keller J 2001 Drip irrigation for
Kahinda J and Taigbenu A 2011 Rainwater harvesting in south small farmers Water International 26 3–13
africa: Challenges and opportunities Physics and Chemistry of Pretty J N, Noble D a, Bossio D, Dixon J, Hine R E,
the Earth, Parts A/B/C 36 968–76 De Vries F W T P and Morison J I L 2006 Resource-
Karimi P, Qureshi A S, Bahramloo R and Molden D 2012 Reducing conserving agriculture increases yields in developing
carbon emissions through improved irrigation and countries Environ. Sci. Technol. 40 1114–9
groundwater management: a case study from iran Pretty J, Toulmin C and Williams S 2011 Sustainable intensification
Agricultural Water Management 108 52–60 in african agriculture Int. J. Agri. Sustainability 9 5–24

13
Environ. Res. Lett. 11 (2016) 025002

Reynolds T W, Waddington S R, Anderson C L, Chew A, True Z and Taylor K, Stouffer R J and Meehl G A 2012 An overview of CMIP5
Cullen A 2015 Environmental impacts and constraints and the experiment design Bulletin of the American
associated with the production of major food crops in sub- Meteorological Society 3 485–98
saharan africa and south asia Food Security 7 795–822 The Royal Society 2009 Reaping the benefits: science and the
Rockström J 2003 Water for food and nature in drought-prone sustainable intensification of global agriculture (London: The
tropics: vapour shift in rain-fed agriculture Phil. Trans. R. Soc. Royal Society)
Lond. B 358 1997–2009 Tilman D 1999 Global environmental impacts of agricultural
Rockström J, Barron J and Fox P 2003 Water productivity in rain- expansion: the need for sustainable and efficient practices
fed agriculture: challenges and opportunities for smallholder Proc. Natl. Acad. Sci. USA 96 5995–6000
farmers in drought-prone tropical agroecosystems (Water Tilman D, Balzer C, Hill J and Befort B L 2011 Global food demand
Productivity in Agriculture: Limits and Opportunities for and the sustainable intensification of agriculture Proc. Natl.
Improvement) ed J Kijne et al (Wallingford: CAB Acad. Sci. USA 108 20260–4
International) ch 9 145–62 Tsubo M and Walker S 2007 An assessment of productivity of maize
Rockström J and Falkenmark M 2015 Agriculture: increase water grown under water harvesting system in a semi-arid region
harvesting in Africa Nature 519 283–5 with special reference to ENSO J. Arid Environ. 71 299–311
Rockström J and Karlberg L 2010 The quadruple squeeze: United Nations 2015 Global Sustainable Development Report—
defining the safe operating space for freshwater use to achieve 2015 Edition, United Nations
a triply green revolution in the anthropocene Ambio 39 United Nations 2016 Sustainable Development Goals (www.un.org/
257–65 sustainabledevelopment/) (accessed 2 February 2016)
Rockström J, Lannerstad M and Falkenmark M 2007 Assessing the Valin H et al 2014 The future of food demand: understanding
water challenge of a new green revolution in developing differences in global economic models Agricultural Economics
countries Proc. Natl. Acad. Sci. USA 104 6253–60 45 51–67
Rosenzweig C et al 2014 Assessing agricultural risks of climate van Rensburg L D, Bothma C B, Fraenkel C H, Le Roux P A L and
change in the 21st century in a global gridded crop model Hensley M 2012 In-field rainwater harvesting: mechanical
intercomparison Proc. Natl. Acad. Sci. USA 111 3268–73 tillage implements and scope for upscaling Irrigation and
Rost S, Gerten D, Hoff H, Lucht W, Falkenmark M and Rockström J Drainage 61 138–47
2009 Global potential to increase crop production through Wada Y and Bierkens M F P 2014 Sustainability of global water use:
water management in rainfed agriculture Environ. Res. Lett. 4 past reconstruction and future projections Environ. Res. Lett.
044002 9 104003
Sánchez P a 2010 Tripling crop yields in tropical africa Nat. Geosci. 3 Waha K, van Bussel L G J, Müller C and Bondeau A 2012 Climate-
299–300 driven simulation of global crop sowing dates Glob. Ecol.
Schaphoff S, Heyder U, Ostberg S, Gerten D, Heinke J and Lucht W Biogeogr. 21 247–59
2013 Contribution of permafrost soils to the global carbon Wani S, Sreedevi T, Rockström J and Ramakrishna Y 2009 Rainfed
budget Environ. Res. Lett. 8 014026 agriculture — past trends and future prospects Rainfed
Schneider U, Becker A, Finger P, Meyer-Christoffer A, Ziese M and Agriculture: Unlocking the Potential ed S Wani et al
Rudolf B 2014 GPCCʼs new land surface precipitation (Wallingford: CAB International) ch 1, pp 1–35
climatology based on quality-controlled in situ data and its Watanabe S, Kanae S, Seto S, Yeh P J-F, Hirabayashi Y and Oki T
role in quantifying the global water cycle Theor. Appl. 2012 Intercomparison of bias-correction methods for
Climatol. 115 15–40 monthly temperature and precipitation simulated by
Searchinger T, Hanson C, Ranganathan J, Lipinski B, Waite R, multiple climate models J Geophys. Res. 117 D23114
Winterbottom R, Dinshaw A and Heimlich R 2013 Welderufael W a, Le Roux P a L and Hensley M 2008 Quantifying
Creating a sustainable food future—a menu of solutions rainfall-runoff relationships on the dera calcic fluvic regosol
to sustainably feed more than 9 billion people by 2050 ecotope in ethiopia Agricultural Water Management 95 1223–32
Technical Report (Washington, DC: World Resources Wisser D, Frolking S, Douglas E M, Fekete B M, Schumann A H and
Institute) Vörösmarty C J 2010 The significance of local water resources
Siebert S, Kummu M, Porkka M, Döll P, Ramankutty N and captured in small reservoirs for crop production — a global-
Scanlon B R 2015 A global data set of the extent of irrigated scale analysis J. Hydrol. 384 264–75
land from 1900 to 2005 Hydrology and Earth System Sciences World Bank 2013 The Sahel: New Push to Transform Agriculture with
19 1521–45 More Support for Pastoralism and Irrigation (http://
Steffen W et al 2015 Planetary boundaries: Guiding worldbank.org/en/news/press-release/2013/10/27/the-
human development on a changing planet Science 347 sahel-new-push-to-transform-agriculture-with-more-
1259855 support-for-pastoralism-and-irrigation)
Studer R and Linger H 2013 Water harvesting: Guidelines to good Xie H, You L, Wielgosz B and Ringler C 2014 Estimating the
practice Technical Report (Rome: Centre for Development potential for expanding smallholder irrigation in sub-saharan
and Environment (CDE), Bern; Rainwater Harvesting africa Agricultural Water Management 131 183–93
Implementation Network (RAIN), Amsterdam; MetaMeta, Zhu Q and Yuanhong L 2006 Rainwater harvesting: the key to
Wageningen; The 50 International Fund for Agricultural sustainable rural development in Gansu, China Waterlines 24
Development (IFAD)) 4–7

14