Land Use Policy

Economic and Environmental Impact Assessment of Sustainable Future Irrigation

Practices in the Indus Basin of Pakistan
--Manuscript Draft--

Manuscript Number:

Article Type: Full Length Article

Keywords: irrigation technologies; irrigation costs; groundwater anomaly; groundwater depletion;

CO2 emissions; solar power supply

Corresponding Author: Muhammad Muzammil

GERMANY

First Author: Muhammad Muzammil

Order of Authors: Muhammad Muzammil

Azlan Zahid, PhD

Lutz Breuer, PhD

Abstract: Agriculture plays a substantial role in the economy of Pakistan, which is strongly
associated with the irrigated plains of the Indus basin. Insecure water supply and poor
irrigation practices are major threats to water security and the environment of the
country. In this research, we investigate the economic and environmental feasibility of
alternative improved irrigation technologies over the status quo (2002−2018) for the
Punjab and Sindh provinces, representing about 90% of the total irrigated area in the
country. We estimate the site−specific irrigation costs, groundwater anomalies, and
CO2 emissions. Improved irrigation technologies consider different energy sources
including solar power in combination with changes in the irrigation method. The status
quo irrigation costs are estimated to 1,301 million US$ yr−1, its groundwater depletion
to 6.3 mm yr−1 and CO2 emissions to 4.12 million t yr−1, of which 96% originate from
energy consumption and 4% via bicarbonate extraction from groundwater. Irrigation
costs of improved irrigation scenarios increase with all energy sources compared to the
status quo, which is mainly based on gravity-fed surface irrigation. This is because of
additional variable and fixed costs for system’s operation in the scenarios. Of these,
subsidized electricity induces the lowest costs for farmers with 63% extra costs
followed by solar energy with 77%. However, groundwater depletion can even be
reversed with up to 35% rise in groundwater levels via improved irrigation
technologies. Solar−powered irrigation can break down CO2 emissions by 81% whilst
other energy sources boost emissions by up to 410%. Results suggest that there is an
extremely opposing development between economic and ecological preferences in the
scenarios, requiring stakeholders to negotiate viable trade-offs.

Suggested Reviewers: Hamideh Nouri, PhD

Senior Research Fellow, University of Göttingen: Georg-August-Universitat Gottingen
hamideh.nouri@uni-goettingen.de

Noman Ali Buttar, PhD

Post Doc, Jiangsu University of Science and Technology: Jiangsu University
5103160333@stmail.ujs.edu.cn

Muhammad Saifullah, PhD

Assistant Professor, Muhammad Nawaz Shareef University of Agriculture
muhammad.saifullah@mnsuam.edu.pk

Muhammad Uzair Qamar, PhD

Assistant Professor, University of Agriculture Faisalabad
muhammad.uzair@uaf.edu.pk

Ikram ullah, PhD

Post Doc, Jiangsu University of Science and Technology: Jiangsu University
ikramullah@ujs.edu.cn

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Cover Letter

Institute for Landscape Ecology and Resources

 ILR ▪ Heinrich-Buff-Ring 26 ▪ 35392 Gießen ▪ Germany Management

Chair for Landscape, Water

and Biogeochemical Cycles
Editor in Chief
Land Use Policy Muhammad Muzammil
Heinrich-Buff-Ring 26
35392 Gießen
Germany

Phone 0641 / 99 - 37398

Fax 0641 / 99 - 37389
E Email: muhammad.muzammil@umwelt.uni-giessen.de

2021-06-28

Dear Sir,

We would like to submit an original research article entitled “Economic and Environmental Impact As-
sessment of Sustainable Future Irrigation Practices in the Indus Basin of Pakistan’’ for publication in
your prestigious journal ‘Land Use Policy’. I confirm that this work is original and has not been pub-
lished, nor is it currently under consideration for publication elsewhere.

This research manuscript evaluates and compares the cost-effectiveness and environmental conse-
quences of improved irrigation technologies over the status quo irrigation practices with the aims of
economic and environmental sustainability in one of the world’s largest irrigated area. We develop a
spatio-temporal modeling framework that estimates the irrigation cost, groundwater anomaly, and CO2
emissions. Further, we perform an inclusive analysis to find the optimum energy source for sustainable
irrigation through improved irrigation technologies. This is a novel, significant and needed study for the
Indus basin of Pakistan. We believe that this manuscript is appropriate for publication in Land Use
Policy.

Thank you for consideration of this manuscript.

Sincerely,

Muhammad Muzammil
Corresponding Author
Highlights

Highlights:
 Current irrigation practices in Pakistan risk water security and adversely affect the environment

 Improved irrigation methods tend to reduce groundwater depletion

 Economic benefits of irrigation technologies are insufficient over the status quo

 Solar-powered irrigation technologies could massively save CO 2 emissions

Title Page (with Author Details)

1 Economic and Environmental Impact Assessment of Sustainable Future

2 Irrigation Practices in the Indus Basin of Pakistan

3 Muhammad Muzammila,*, Azlan Zahidb, Lutz Breuera,c

4 a
Institute for Landscape Ecology and Resources Management (ILR), Research Centre for BioSystems, Land Use and Nutrition

5 (IFZ), Justus Liebig University Giessen, 35392 Giessen, Germany

6 b
Department of Agricultural and Biological Engineering, Penn State University, University Park, PA 16802, USA

7 c
Centre for International Development and Environmental Research (ZEU), Justus Liebig University Giessen, 35390 Giessen,

8 Germany

9 muhammad.muzammil@umwelt.uni−giessen.de (*corresponding author)

10 axz264@psu.edu

11 lutz.breuer@umwelt.uni−giessen.de

12 Abstract

13 Agriculture plays a substantial role in the economy of Pakistan, which is strongly associated with the

14 irrigated plains of the Indus basin. Insecure water supply and poor irrigation practices are major threats

15 to water security and the environment of the country. In this research, we investigate the economic and

16 environmental feasibility of alternative improved irrigation technologies over the status quo (2002−2018)

17 for the Punjab and Sindh provinces, representing about 90% of the total irrigated area in the country.

18 We estimate the site−specific irrigation costs, groundwater anomalies, and CO2 emissions. Improved

19 irrigation technologies consider different energy sources including solar power in combination with

20 changes in the irrigation method. The status quo irrigation costs are estimated to 1,301 million US$ yr−1,

21 its groundwater depletion to 6.3 mm yr−1 and CO2 emissions to 4.12 million t yr−1, of which 96% originate

22 from energy consumption and 4% via bicarbonate extraction from groundwater. Irrigation costs of

23 improved irrigation scenarios increase with all energy sources compared to the status quo, which is

24 mainly based on gravity-fed surface irrigation. This is because of additional variable and fixed costs for

25 system’s operation in the scenarios. Of these, subsidized electricity induces the lowest costs for farmers

26 with 63% extra costs followed by solar energy with 77%. However, groundwater depletion can even be

27 reversed with up to 35% rise in groundwater levels via improved irrigation technologies. Solar−powered

28 irrigation can break down CO2 emissions by 81% whilst other energy sources boost emissions by up to

29 410%. Results suggest that there is an extremely opposing development between economic and

30 ecological preferences in the scenarios, requiring stakeholders to negotiate viable trade-offs.

1
31 Keywords: irrigation technologies, irrigation costs, groundwater anomaly, groundwater depletion, CO2

32 emissions, solar power supply

33 Highlights:
34  Current irrigation practices risk water security and adversely affect the environment

35  Improved irrigation methods tend to reduce groundwater depletion

36  Economic benefits of irrigation technologies are insufficient over the status quo

37  Solar-powered irrigation technologies could massively save CO 2 emissions

38 1. INTRODUCTION

39 Agriculture of Pakistan is based on irrigation where rainfall marginally meets 15% of crop water

40 requirements (Qureshi et al., 2010). Irrigated agriculture is associated with the Indus Basin, which

41 provides irrigation water on a supply basis. Surface water resources are unable to fulfill the actual

42 irrigation demands owing to the high evapotranspiration and salinity environment in the plain (Basharat

43 et al., 2014). Groundwater covers 40−60% of the irrigation needs to meet the deficit in surface water

44 supplies (Cheema et al., 2014). The intensive groundwater use results in a decline of the water table

45 and accumulation of soil salinity, which originates from saline pockets of the aquifers (Qureshi et al.,

46 2008). Meanwhile, the increasing trend of groundwater pumping has become energy exhaustive. It

47 reduces the farmer’s income because of high irrigation costs and leads to massive greenhouse gas

48 releases, mainly CO2 emissions, through energy consumption (Qureshi, 2014). Furthermore, the role of

49 groundwater depletion in greenhouse emissions is still unaccounted, which can be a significant emission

50 source associated with bicarbonate extraction (Mishra et al., 2018; Wood and Hyndman, 2017). Despite

51 the poor situation of water resources availability in the country, inefficient irrigation methods dominate

52 in the region, with losses of up to 50% of available water in the fields (Rizwan et al., 2018).

53 In many arid and semi−arid countries, where water resources are limited and depleting rapidly, there is

54 pressure to reduce water consumption for the water security of the growing population (Kahlown et al.,

55 2007). Previous studies indicate that improved irrigation technologies enable farmers to cope with water

56 scarcity and insecure water supply (Muzammil et al., 2020; Wang et al., 2017; Zhang et al., 2019).

57 However, the impact of irrigation technologies remains a critical topic for sustainable irrigation. The

58 consequences of improved irrigation technologies vary among regions because of differences in

59 cost−benefits, off−farm environmental impacts, and social preferences (Wichelns and Oster, 2006).

2
60 Economic impact assessment should be part of the evaluation process and support the

61 decision−making. For example, Zou et al. (2013) analyzed water−saving strategies based on the climate

62 change response for China and proposed that channel lining is a preferable strategy from an economic

63 perspective compared to pressurized irrigation practices. Mahinda et al. (2016) investigated the

64 economic impact of sorghum production via drip irrigation in semi−arid regions of Tanzania and

65 recommended that two irrigations per day are beneficial to get higher economic returns.

66 Narayanamoorthy et al. (2018) studied the economic impact of drip irrigation on vegetable crops and

67 their findings indicate that the pressurized irrigation system offers high net returns compared to

68 conventional irrigation methods.

69 In the context of environmental consequences, irrigation development can have severe effects on the

70 investment in irrigation projects at regional and basin levels (Dogaru et al., 2019; FAO, 1997;

71 Velasco−Muñoz et al., 2019). For example, Daccache et al. (2014) projected that a pressurized irrigation

72 system is capable to increase irrigation efficiency, but CO2 emissions increase due to additional energy

73 consumption compared to a gravity−fed surface irrigation system. Shekhar et al. (2020) showed that

74 technology changes could have the potential to mitigate groundwater depletion through pressure

75 reduction on water resources. However, the lower percolation from fields with improved water saving

76 irrigation techniques may reduce aquifer recharge (Johnson et al., 1999). In Pakistan, the potential of

77 high−efficiency irrigation technologies related to water saving have been recognized. Several studies

78 revealed that it is possible to overcome water scarcity in Pakistan through the adoption of high−efficiency

79 irrigation systems (Latif et al., 2016; Rizwan et al., 2018; Zafar et al., 2020). Meanwhile, previous studies

80 show that future power supply for irrigation technologies should consider changes in the energy source,

81 including solar power supply (Hassan and Kamran, 2018; Mongat et al., 2015). Nevertheless, the

82 economic and environmental impacts of these technologies are still unknown over the status−quo

83 irrigation settings. An inclusive analysis of the cost−effectiveness of improved irrigation technologies

84 coping with ecological impact can support economic development and environmental sustainability in

85 the region.

86 In this study, we compare the economic and environmental impacts of the status−quo irrigation settings

87 with alternative, improved irrigation technologies. We use a coupled economic−environmental−modeling

88 framework to estimate the irrigation costs, groundwater depletion, and CO2 emissions to understand the

89 return on investment and environmental effects. We consider improved, more sustainable irrigation

3
 90 technologies that differ from the status−quo irrigation practices in terms of water consumption, irrigation

91 costs, and energy use. As the water consumption via improved irrigation technologies is lower than that

92 of conventional irrigation, the effect of groundwater recharge through surplus irrigation is diminishing,

93 which we take also into account. Furthermore, improving the established irrigation system needs a high

94 initial investment and, in the case where the gravity−fed irrigation system is replaced, additional

95 operational energy costs and associated CO2 emissions come into play, which are also analyzed.

96 The objectives of the current study are: (1) to investigate the economic impact of improved irrigation

97 technologies over status−quo irrigation practices, (2) to compare groundwater depletion and CO2

98 emissions of the status−quo irrigation settings with improved irrigation practices, and (3) to develop

99 alternative scenarios for improved irrigation technologies and identify sustainable energy use options in

100 the irrigation agriculture of Pakistan.

101 2. MATERIALS AND METHODS

102 2.1 Study Area

103 The study focuses on the irrigated areas of Punjab and Sindh provinces in the Indus basin of Pakistan.

104 Together, these cover 17 million ha (Figure 1), representing 90% of the total irrigated area in the country.

105 The topography of the plain falls from north to south, ranging from 540 to 4 m above mean sea level. The

106 basin has an arid to semi−arid climate with complex hydrological processes due to spatial and temporal

107 variation in the rainfall, temperature, land use, and water consumption. The average annual rainfall

108 amounts to 379 mm (2002−2018), while maximum temperature ranges from 34−44°C in the summer

109 (Apr−Sep) and 20−28°C in the winter (Dec−Feb). The annual potential evapotranspiration varies from

110 1,200 to 2,050 mm from the north to the south. Crops are harvested in two cropping seasons called

111 Kharif (wet season; Apr−Sep) and Rabi (dry season; Oct−Mar). Sugarcane, cotton, and rice are

112 dominant crops in the Kharif while wheat is a major crop in the Rabi season.

113 There are five major tributaries to the Indus (Indus, Chenab, Ravi, Jhelum, and Sutlej), which supply

114 irrigation water via a network of canals and watercourses. The provincial governments distribute the

115 surface water among farmers according to the landholding size and collect the water charges two times

116 in a year in the Kharif and Rabi seasons. The water charges vary from province to province i.e., the

117 Punjab government collects at a flat rate despite which crop is grown while it varies in Sindh by crop to

118 crop. Farmers using additional groundwater recourses via private units (tubewells), operate with diesel

4
119 engines or mains power for groundwater pumping. The government provides subsidized electricity to

120 farmers. However, diesel operated tubewells are common among farmers with 87% of share because

121 they have a lower initial investment than electric operated tubewells. Crops are widely irrigated via

122 surface irrigation with an application efficiency of 45−60%. Improved irrigation systems (drip and

123 sprinkler) are installed only in a limited area (50,000 ha) through a subsidized program of the World

124 Bank and the government of Punjab in the frame of the Punjab Irrigated-Agriculture Productivity

125 Improvement Project (PIPIP).

126
127 Figure 1. Map of the study area.

128 2.2 Modeling Framework

129 In this study, we develop an economic−environmental−modeling framework to evaluate the economic

130 and environmental impacts of the status−quo irrigation practices and a variety of scenarios with

131 improved irrigation technologies. The model is written in python by using the SciPy package. The

132 modeling approach uses gridded data and makes use of information such as the irrigation requirements,

133 harvested area, crop water consumption, groundwater level, energy use required for pumping water,

134 water prices and energy costs. The methodological steps of the modeling framework are summarized

135 in Figure 2, and the calculation methods are described in section 2.3. The input data and references

136 used in this study are given in Table A1.

5
137
138 Figure 2. Methodological steps to estimate the economic and environmental impact of the status−quo

139 irrigation settings and the improved irrigation technologies.

140 2.3 Calculation Methods

141 2.3.1 Irrigation Requirements

142 Regional irrigation requirements (IRRarea) are calculated for the entire area by combining all crop’s

143 productive (IRRprod) and unproductive (IRRunprod) consumptions of irrigation water along with the leaching

144 requirements (LR) (Eq1). IRRprod contributes to crop growth, while IRRunprod covers the water losses in

145 line with the efficiency of the irrigation system (IRReffi). IRRunprod does not result in crop production and

146 percolates from the root zone to the groundwater or evaporates at the soil surface. These water losses

147 partially cover the LR (Multsch et al., 2013). The LR is an additional amount of water that is otherwise

148 needed to leach salts from the root zone by assuming the salinity tolerance limit of each crop and the

149 salt fraction in the irrigation water (Ayers and Westcot, 1985).

IRRprod
150 IRR area = + LR (Eq1)
IRReffi

151 with IRRarea, IRRprod, and LR given in (km3 yr−1) and IRReffi in percentage (%).

152 In this study, we use data on the site−specific productive irrigation water IRR prod and leaching

153 requirement LR (2002−2016) from a recently published study (Muzammil et al., 2020). The dataset holds

154 information with a spatial resolution of 0.063° for Pakistan. Muzammil et al. (2020) used

155 SPARE:WATER, an open−source model integrated into a geographical information system to estimate

156 the crop water balance at the grid level (Multsch et al., 2013). SPARE:WATER follows the FAO56

6
157 guidelines to determine crop water requirements (Allen et al., 1998) and calculates the potential LR in

158 line with the salinity tolerance limit of crops and the salt fractions in the irrigation water. For this study,

159 we extended the simulation period of 2002−2016 from Muzammil et al. (2020) and included the years

160 2017 and 2018. A detailed list of model input data and parameters required to run the model is given in

161 Muzammil et al. (2020). The climatic data is obtained from the Pakistan Metrological Department, while

162 information on crops is provided from the Pakistan Statistics Bureau. The efficiencies of irrigation

163 systems are taken from the FAO dataset as 60%, 75%, and 90% for surface, sprinkler, and drip irrigation,

164 respectively (Brouwer et al., 1985).

165 2.3.2 Surface Water and Groundwater Use

166 As surface water and groundwater are used in the Indus basin to meet the irrigation demand, we

167 estimate the surface water share (km 3 yr−1) from a dataset of annual canals supply. The data is

168 preprocessed to exclude the off−farm water losses assuming a conveyance efficiency of 70% (Arshad

169 et al., 2005; Cheema et al., 2014). The volume of groundwater abstraction (km 3 yr−1) is determined

170 by subtracting the available surface water in the fields from the regional irrigation requirements.

171 2.3.3 Irrigation Costs

172 The total annual irrigation costs of the entire area (TCarea; million US$ yr−1) are estimated by adding the

173 total fixed costs (TFCarea) and total variable costs (TVCarea) (Eq2):

174 TCarea = TFCarea + TVCarea (Eq2)

175 1) Total Fixed Irrigation Costs

176 Total fixed irrigation costs per area (TFCarea) are estimated by adding its components on a regional

177 basis (Eq3), i.e., surface water prices (SWParea), tubewell construction costs (TCCarea), and irrigation

178 system costs (ISCarea).

179 TFCarea = SWParea + TCCarea + ISCarea (Eq3)

180 where SWParea (million US$ yr−1) results from summing up the products of costs occurring for surface

181 water for crop irrigation (US$ ha−1) times their harvested area (ha yr−1). TCCarea (million US$ yr−1) is

182 estimated by dividing the initial costs of all tubewells (million US$) for a given area from their average

183 lifetimes (years). The initial costs of tubewells are projected by combining the construction costs of all
7
184 diesel and electric operated tubewells. The TCCarea vary and depend on groundwater level and power

185 required for pumping groundwater (Qureshi et al., 2010). ISCarea are calculated by summing up the

186 product of all crops’ irrigation system costs per hectare (US$ ha−1) times their harvested area (ha yr−1).

187 Note that the annual irrigation system costs are split in halves for the crops of the two growing seasons

188 Kharif and Rabi, respectively. The annual irrigation system costs are derived from dividing the initial

189 costs of the systems by their average lifetimes (years). The status−quo irrigation system is based on

190 gravity, therefore ISCarea for surface irrigation are negligible (Latif et al., 2016). The initial costs of the

191 improved irrigation system vary from crop to crop and by changing the power source.

192 2) Total Variable Irrigation Costs

193 We use Eq4 to calculate the regional value of the total variable irrigation costs (TVCarea) by adding its

194 components, i.e., the operational costs (OCarea) and the maintenance costs (MCarea):

195 TVCarea = OCarea + MCarea (Eq4)

196 We further divide OCarea into two parts, i.e., the groundwater pumping costs (GPCarea) and the

197 operational costs of the irrigation system (OCSarea). Accordingly, the maintenance costs (MCarea) are

198 composed of the maintenance costs of the tubewells (MCTarea), and the maintenance costs of the

199 irrigation system (MCSarea).

200 The GPCarea (million US$ yr−1) is based on the costs for the energy sources diesel and electricity. The

201 share of diesel and electric pumping in the study area is estimated by using the fraction of diesel and

202 electric operated tubewells in a grid cell. GPCarea are projected by adding the groundwater pumping

203 costs of diesel (GPCarea(d)) and electric (GPCarea(e)) operated tubewells. Both, GPCarea(d) and GPCarea(e),

204 are calculated by summing up the product of the tubewell abstracted groundwater volumes (m3) times

205 the pumping costs (US$ m−3). Pumping costs are calculated by multiplying the energy consumed (kWh)

206 per m3 pumped groundwater and the energy price (US$ kWh−1). The energy consumption is determined

207 from Eq5 where V, TDH, and Ƞpp are abstracted groundwater volume (m 3), total dynamic head (m), and

208 pumping plant efficiency (%), respectively (Kay and Hatcho, 1992). In this study, the energy price for

209 the electric source is used directly as the given electricity price in the country (US$ kWh−1) while for

210 diesel consumption, fuel price (US$ L−1) is converted into an energy price (US$ kWh−1) by multiplying

211 fuel price with a conversion factor of 0.11 (Kay and Hatcho, 1992).

8
 V×TDH
212 Energy (kWh) = (Eq5)
367 × ηpp

213 The OCSarea (million US$ yr−1) consists of the energy and labor costs of the irrigation system. The energy

214 costs for the surface irrigation method are negligible as its operation is based on gravity (Khatri et al.,

215 2013). For the pressurized irrigation system, energy demand is estimated by multiplying the energy

216 required to run the irrigation system (kWh yr−1) and the energy price (US$ kWh−1), being either diesel or

217 electricity. The energy consumption is estimated from Eq5 where TDH indicates the total head required

218 to run the irrigation system, i.e., the operational head, friction losses, and suction lift. Labor costs are

219 calculated by summing up the product of labor charges (US$ ha−1) and the harvested area (ha yr−1).

220 MCTarea (million US$ yr−1) is calculated by summing up the annual maintenance costs of diesel and

221 electric operated tubewells in the region. The maintenance costs of diesel and electric operated

222 tubewells are estimated by multiplying the maintenance costs per tubewell and the number of electric

223 and diesel operated tubewells in the study area.

224 Finally, the MCSarea (million US$ yr−1) contains repair and cleaning costs of the watercourses, which is

225 calculated by multiplying the maintenance costs (US$ ha−1) and the total harvested area (ha yr−1). For

226 improved irrigation technologies, maintenance costs cover repair and security costs of the system. We

227 estimate it as 5% of the total operational costs (Buchanan, 2002).

228 2.3.4 Groundwater Storage

229 The annual aquifer recharge (mm) is estimated from the Water Table Fluctuation method by adding the

230 groundwater storage anomaly (mm) and the depth of pumped groundwater from the aquifer (mm)

231 (Bhanja et al., 2019; Wu et al., 2019). We use monthly terrestrial water storage data from the Gravity

232 Recovery and Climate Experiment (GRACE) to estimate the groundwater storage anomaly. GRACE

233 data has been validated for Pakistan in past studies (Iqbal et al., 2016; Tang et al., 2017). In this study,

234 we apply the GRACE Mascon solution, which does not need post−processing filtering and which is less

235 depending on scale factors (Save et al., 2016). Groundwater storage anomaly is derived by subtracting

236 the surface water storage (soil moisture, canopy water, snow water) from the terrestrial water storage.

237 The surface water storage is estimated up to 2 m of the soil column from the land surface model (NOAH)

238 dataset of the GLDAS product, which has been used in several regions where in situ measurements are

239 not available (Andersen et al., 2005; Leblanc et al., 2009; Rzepecka and Birylo, 2020; Tiwari et al.,

240 2009).

9
241 Further, we calculate the contributions of the fields’ percolation losses to total recharge. For the

242 status−quo irrigation settings, it is estimated from published data (Arshad, 2004). This data is simulated

243 via the GLEAMS hydrological model, which is used at the field scale to estimate the movement of water

244 content through percolation and contribution of recharge to the groundwater (Nicks, 1998). Accordingly,

245 water percolates from fields to the groundwater storage in the Indus basin of Pakistan at a rate of

246 0.314 mm day−1. It is assumed that this percolation is negligible for improved irrigation technologies

247 where irrigation surplus is marginal (Dewandel et al., 2008).

248 2.3.5 Carbon Dioxide Emissions

249 We estimate CO2 emissions from the status−quo irrigation practices and improved irrigation

250 technologies, where energy consumption and bicarbonate extraction from the groundwater are

251 considered as the major emissions sources.

252 1) CO2 Emissions from Energy Consumption

253 There are two energy consumption sources related to irrigation in the study area, i.e., groundwater

254 pumping and irrigation system operation. The annual mass of CO2 emissions depends on the amount

255 of energy consumed (kWh yr−1) and the types of these energy sources (Ramphull and Surroop, 2017),

256 represented by their respective emission factors. We apply a fixed emission factor for diesel engines of

257 0.32021 kg CO2 kWh−1 (Wang et al., 2012). For electricity, we calculate with a constant value of 0.47337

258 kg CO2 kWh−1 based on information on the major energy sources for power production in Pakistan

259 (Brander et al., 2011). Note that the status−quo irrigation system is based on gravity, therefore, no CO2

260 is emitted.

261 2) CO2 Emissions from Bicarbonates Extraction

262 In this study, we assume that the CO2 concentrations in recharging groundwater and pumped

263 groundwater are the same. If groundwater recharge is equal to the abstraction, there are no CO2

264 emissions (Wood and Hyndman, 2017). Hitherto, CO2 is emitted if groundwater is depleted and CO2 is

265 sequestered in the aquifer in cases of rising groundwater levels. We estimate CO2

266 emissions/sequestration (million t CO2 yr−1) by multiplying CO2 concentrations in the groundwater (mg

267 L−1) and groundwater depletion/increase (m3). Groundwater depletion/increase is estimated by

268 multiplying the groundwater storage anomaly (m) and surface area of the plain (m 2).

10
269 The CO2 concentrations in the groundwater depend on atmospheric CO2 dissolved in water, which

270 enters the groundwater body via percolation and thus depends on the groundwater recharge rate. During

271 solution, CO2 and H2O split into hydrogen (H+) and bicarbonate (HCO3−) ions (Eq6).

272 CO2 + H2 O → H + + HCO3− (Eq6)

273 It is assumed that half of the mass of total bicarbonates present in the groundwater originates from this

274 separation. While another half is formed when the CaCO 3 rich rock in the aquifer reacts with hydrogen

275 ions (H+) (Mishra et al., 2018) (Eq7):

276 H + + CaCO3 → HCO3− + Ca2+ (Eq7)

277 Depending on the resulting bicarbonate concentration in the groundwater, CO2 evolves into the

278 atmosphere according to Eq8 when groundwater is pumped.

279 Ca(HCO3 )2 → CO2 + H2 O + CaCO3 (Eq8)

280 The resulting CO2 concentration (mg L−1) in the groundwater is calculated by multiplying the molecular

281 mass ratio of HCO3− and CO2 with the bicarbonate concentration (mg L−1) (Eq9).

1 44
282 CO2 Concentration = HCO3− × (Eq9)
2 61

283 2.4 Scenario Development

284 We develop four future scenarios (SC−1 to SC−4) to derive a potential optimum plan for irrigation that

285 reduces the irrigation costs, groundwater depletion, and CO2 emissions in the Indus basin. Scenarios

286 are established by changing the status−quo irrigation methods (gravity−fed surface irrigation) to

287 improved irrigation technologies as this has been identified as a preferable solution to reduce total

288 amount of irrigation water (Muzammil et al. 2020). The year 2018 is considered as a baseline to which

289 scenarios are compared. We keep the harvested area from the baseline in the scenarios and convert

290 surface irrigation to drip irrigation for row crops and to sprinkler irrigation for field crops. The scenarios

291 are classified according to the energy sources required to operate the revised irrigation system. In SC−1,

292 the diesel engines are used to operate the irrigation system, SC−2 is run on electricity but assumes

293 subsidized prices as status quo conditions, SC−3 is also based on electricity, but considers the actual

294 energy price, and SC−4 is defined by using solar energy.

11
295 3. RESULTS

296 3.1 Water Consumption and Irrigation Costs

297 The shares of surface and groundwater in irrigation water are shown in Figure A1. The irrigation water

298 consumption (IRRarea) and the total irrigation costs (TCarea) for 2002−2018 are presented in Figure 3.

299 Results show that the southern part of Punjab has the highest irrigation requirements while the upper

300 portion of Punjab and the whole parts of Sindh have relatively lower irrigation water consumptions

301 (Figure 3a). We find strong inter−annual variation in the total irrigation requirements with the highest

302 IRRarea in 2002 (177 km3 yr−1) and the lowest in 2015 (130 km3 yr−1) (Figure 3b). A Mann−Kendall test

303 reveals that there is no trend in IRRarea from 2002−2018 (p = 0.23). Average IRRarea is estimated to 157

304 km3 yr−1, of which groundwater accounts for 52% (82 km3 yr−1) and surface water contributes to 48%

305 (75 km3 yr−1). Diesel pumping has the largest share in groundwater abstraction with 83%, followed by

306 electric pumping of 17%. Results of the irrigation costs also show a substantial variation in space and

307 year−to−year (Figure 3.1). The southern region of Punjab has the highest annual irrigation cost

308 compared to other parts of the study area (Figure 3c). The highest irrigation costs are calculated for

309 2014 (1,837 million US$) and the lowest one for 2003 (718 million US$) (Figure 3d). We find an overall

310 significant increasing trend for irrigation costs from 2002−2018 (R2 = 0.43, slope = 41.6, p=0.001). The

311 years 2015 and 2016 are striking with lower costs, which are due to the combined effect of a lower

312 IRRarea and reduced fuel prices compared to other years. The average irrigation costs for 2002−2018

313 are calculated to 1,301 million US$, of which fixed cost components account for 8% (85 million US$)

314 and variable costs account for by far the largest amount (1,216 million US$). Groundwater pumping

315 costs have the largest share in irrigation costs with 60%, followed by maintenance costs (32%), surface

316 water prices (3%), and tubewell construction costs (5%). Diesel pumping costs have a dominant part in

317 the total pumping costs of groundwater with 93%, while the electric pumping holds only 7%.

12
318
319 Figure 3. Irrigation water consumption (a, b) and irrigation cost (c, d) from 2002−2018.

320 3.2 Estimates of Groundwater Depletion

321 We project the groundwater storage from 2002−2018 by estimating the groundwater recharge and

322 abstraction in the study area. The results show that the northern part of the plain (Punjab province) faces

323 the largest depletion rate (−11 mm yr−1) while an increase in groundwater level (4 mm yr−1) is observed

324 in the southern part of the plain (Sindh province) (figure 4a). Overall, the groundwater storage anomaly

325 is significantly decreasing (R2 = 0.39, slope = −3.93, p=0.02) in the study area from 2002−2018 (figure

326 3.2b) at a rate of −6.3 mm yr−1 (−1.35 km3 yr−1). Annual differences are substantial, with the highest

327 depletion rate in the year 2018 (−78 mm yr−1; −16.7 km3 yr−1) and the largest surplus in groundwater

328 storage in 2003 (43 mm yr−1; 9.2 km3 yr−1). Overall, we do neither find significant trends for net

329 groundwater recharge (p = 0.06) nor for abstraction (p = 0.38). Further, the average net recharge rate

330 is estimated to 380 mm yr−1, of which 69% (263 mm yr−1) are contributed by from natural resources

331 (precipitation, and particular leaching from rivers, water bodies and canals) while the cropping fields add

332 another 31% (117 mm yr−1) as percolation losses from unproductive irrigation.

13
333
334 Figure 4. (a) Average groundwater depletion from 2002 to 2018, (b) Time series trend (2002−2018) of

335 groundwater abstraction, net recharge, natural recharge, and storage anomaly.

336 3.3 Estimates of CO2 Emissions

337 We estimate CO2 emissions from 2002−2018 according to the emission sources, i.e., energy

338 consumption and bicarbonate extraction from depleted groundwater volume (Figure 5). The southern

339 part of Punjab depicts the highest CO2 emissions from energy consumption (Figure 5a) while the upper

340 portion of Punjab shows the highest CO2 emissions due to groundwater depletion (Figure 5b). The

341 results further reveal that about 4.12 million t CO2 yr−1 are emitted in the plain, of which 96% (3.95 million

342 t yr−1) result from energy consumption while 4% (0.17 million t yr−1) are stemming from groundwater

343 depletion. The largest CO2 emissions are produced in the year 2018 (5.42 million t) and the lowest one

344 in 2015 (2.15 million t) (Figure 5c). Further, CO2 emissions from groundwater depletion are highly

345 variable over time with a maximum in 2018 (1.58 million t). For several years, we found even negative

346 values (i.e., an increase of the CO2 storage) due to a surplus of groundwater recharge over groundwater

347 abstraction. This results in rather substantial net storage of CO2 in 2003 (−0.93 million t). With regard to

348 the energy source, diesel pumping has a larger share (87%) than CO2 emissions from electric pumping.

14
349
350 Figure 5. Average annual CO2 emissions from (a) energy consumption and (b) groundwater depletion,

351 (c) temporal development of CO2 emissions from 2002−2018. (Note that the color-codes in the maps

352 (a) and (b) vary by a factor of 10)

353 3.4 Scenario Analysis

354 Scenarios are investigated to derive the optimum energy source for improved irrigation technologies

355 and compare the results with the status−quo irrigation method. We establish four scenarios SC1−4 to

356 identify the effect of improved irrigation technologies on irrigation costs, groundwater depletion, and CO2

357 emissions for more sustainable irrigation practices by using different energy sources in each scenario.

358 The changes in irrigation costs, groundwater depletion, and CO2 emissions for all scenarios are

359 presented in Figure 6 and Table A2. In SC−1, we change the gravity driven status−quo irrigation settings

360 with irrigation technologies and consider diesel as the primary energy source. The results indicate that

361 irrigation costs and CO2 emissions increase up to 170% and 410%, respectively, while the groundwater

362 depletion is reduced by up to 135%. SC−2 focuses on changing the status quo irrigation settings with

363 irrigation technologies that run on subsidized electricity from mains power. We find an increase in

364 irrigation costs and CO2 emissions of up to 63% and 165%, respectively. Meanwhile, the groundwater

365 depletion rate decreases by up to 135%. The scenario SC−3 has the same settings as SC−2 but we
15
366 use actual prices for electricity. In consequence, we observe an increase in irrigation costs of up to

367 130% of the baseline scenario. In SC−4, solar−powered irrigation technologies are used instead of the

368 surface irrigation method. The results show that irrigation costs increase by up to 77% while CO2

369 emissions and groundwater depletion are reduced by up to 81% and 135%, respectively.

370

371 Figure 6. Scenarios analysis for (a) irrigation costs (b) CO2 emissions and (c) groundwater balance,

372 via improved irrigation technologies.

373 4. DISCUSSIONS

374 4.1 Economic Impact of Irrigation Methods

375 In the status−quo conditions, the average irrigation water consumption in the study area is 157 km3 yr−1,

376 of which surface water contributes 48% and groundwater 52%. Despite the small difference in water

377 consumption from surface water and groundwater, there is a vast margin between prices with 3% for

378 surface water and 63% for groundwater of the total irrigation costs (1,301 million US$), respectively.

379 Alternatively, scenarios indicate that improved irrigation technologies can reduce water consumption by

380 32%, which could lead to a reduction in groundwater share of up to 61%, with at the same time 55%

381 decreasing groundwater pumping costs. However, improved irrigation technologies raise total irrigation

382 costs owing to the initial and running costs of the system. Scenarios specify that the operation of

383 improved irrigation technologies via subsidized electricity is an optimal scenario among others for

16
384 farmer’s perspective where irrigation costs increase by 63% compared to the status quo. Solar energy

385 is the second most feasible power source when no subsidized electricity is at hand, but still, irrigation

386 costs increase by 77% compared to that of the status quo. Highest costs are found for diesel operated

387 systems which boost irrigation costs by up to 170%. In short, the economic benefits of improved irrigation

388 technologies are insufficient over the status−quo practices to cover the additional expenditure of the

389 irrigation system. This is in line with various other studies that recognized that improved irrigation

390 technologies can increase farmer’s expenditures via capital investments and running costs (Huffaker,

391 2008; Pérez−Blanco et al., 2020; Soto−García et al., 2013). For example, Rodrigues et al. (2012) studied

392 the comparative advantages of drip and sprinkler irrigation in southern Brazil and concluded that

393 economic benefits from water−saving technologies are insufficient to recover the initial costs of the

394 system. Numerous studies revealed that the implication of improved irrigation technologies is a

395 challenge owing to an extra burden of investment compared to surface irrigation. In Pakistan, despite

396 the various awareness campaigns in the last three decades to introduce improved irrigation

397 technologies, farmers are still not willing to adopt the technologies because of the high initial costs of

398 the system. Thus, governments should provide subsidies to farmers for sustainable water consumption

399 (Cremades et al., 2015; Liu et al., 2008; Tiwari and Dinar, 2002), such as in the World Bank funded

400 Punjab Irrigated−Agriculture Productivity Improvement Project with a size of 50,000 ha. Such types of

401 projects have the capability to promote water−saving technologies among farmers. However, it is

402 doubtful that such a technical shift is sustainable from an economic viewpoint.

403 Part of this problem might be arising from the very low surface water prices in Pakistan, which do not

404 promote changing towards more efficient, but costly irrigation technologies. Qamar et al. (2018) studied

405 the implication strategies of improved irrigation technologies in the Indus basin of Pakistan and

406 concluded that the surface water prices should be higher to promote irrigation technologies among

407 farmers. We recommend that a comprehensive analysis should be conducted to study the adoption

408 strategies of improved irrigation technologies by changing the water prices. Such an analysis should not

409 only consider pure economic aspect, but also take into account societal barriers and personal

410 preferences as well as choices from farmers.

17
411 4.2 CO2 Emissions from Irrigation Practices

412 We estimated CO2 emissions from irrigation practices in the Indus basin of Pakistan by assuming

413 emissions from energy consumption and bicarbonate extraction. At the status−quo settings, diesel or

414 electric pumps are used to pump groundwater, which produces 96% of the total CO2 emissions (3.95

415 million t). Our estimates indicate that bicarbonate extraction is not a significant emissions source,

416 amounting to about 4% of the total CO2 emissions (0.17 million t), although groundwater makes up a

417 significant part of the irrigation water in the Indus basin. Mishra et al. (2018) estimated the annual CO2

418 emissions from groundwater bicarbonate extraction to around 0.72 million t, which is not a significant

419 emissions source either compared to energy consumption through groundwater pumping. Wood and

420 Hyndman (2017) calculated CO2 emissions from bicarbonates extraction in the USA and determined

421 that annual 1.7 million t of CO2 are released from this source. Despite a 10fold higher rate as compared

422 to the groundwater mediated CO2 emissions in the Indus basin, the total share of bicarbonate extraction

423 on US CO2 emissions is small with less than 0.5% (estimated from data published by Wood an

424 Hyndman, 2017).

425 Past studies proposed several strategies to reduce CO2 emissions from groundwater pumping. For

426 example, Shah and Kishore (2012) recommended on−site solar and wind energy for groundwater

427 pumping. However, the authors show serious concern that the availability of renewable energy will

428 encourage the farmers to pump additional groundwater because of the currently low pumping costs.

429 Dhillon et al. (2018) projected that an improvement in pumping plant efficiency could also reduce CO2

430 emissions. Zou et al. (2015) showed indirect effects through general water savings of improved irrigation

431 systems and subsequent lower CO2 emissions because of a reduced groundwater demand. However,

432 improved irrigation technologies might require further energy to run the system, which in turn can

433 increase overall CO2 emissions. Daccache et al. (2014) studied the environmental impact of irrigation

434 practices in the Mediterranean region of Spain. Similar to our results, they revealed that CO2 emissions

435 increased by 135% for improved irrigation technologies compared to the old−fashioned, gravity−based

436 surface irrigation method.

437 We estimate CO2 emissions for different scenarios of improved irrigation technologies by combining

438 emissions from groundwater pumping and irrigation system operation. Our results indicate that diesel

439 engines and mains power electricity are both detrimental energy sources for advancing irrigation

440 technologies compared to the status−quo settings, simply because of the huge increase in CO2
18
441 emissions by 410% and 165%, respectively. However, solar energy operating systems are most

442 effective, which can reduce CO2 emissions even of the status−quo technology by 81%. Many studies

443 revealed that solar energy is the best option for improved irrigation technologies for sustainable

444 development in a region or basin (Hartung and Pluschke, 2018; Roblin, 2016; Yang et al., 2014).

445 4.3 Groundwater Depletion

446 In the study area, the average groundwater depletion is 6.3 mm yr−1, which is comparatively low. For

447 example, Long et al. (2016) estimated the groundwater depletion to be 31 mm yr−1 in the Northwest

448 Indian Aquifer. Shen et al. (2015) investigated groundwater storage anomaly in the Hai River Basin

449 China and reported that groundwater is depleting at a rate of 17 mm yr−1. Despite the lower depletion

450 rate in the Indus basin, our estimates show that the groundwater depletion trend is increasing from

451 2002−2018. Tang et al. (2017) confirmed that groundwater storage is diminishing in the Indus basin. It

452 has been predicted that the depletion rate in the Indus basin will increase by 50% in 2050 compared to

453 the groundwater depletion trend in 2005 (OECD, 2017). We believe that an increasing trend of

454 groundwater depletion is a serious matter and quick measures are needed for sustainable groundwater

455 usage. In the sense of sustainability, the groundwater abstraction rate should be lower than the recharge

456 rate (Cools et al., 2002; Foster and Garduño, 2004; Shamsudduha et al., 2011). Our results show that

457 improved irrigation technologies are capable to reduce groundwater utilization compared to status−quo

458 irrigation. However, such improvements can also have negative side effects like the reduction

459 percolation losses from fields. These apparent negative losses lead, on the one hand, to a leaching of

460 salts from the soil (Muzammil et al., 2020) and, on the other hand, also to groundwater recharge. Overall,

461 our estimates verify that the reduction in groundwater abstraction is larger than field losses, resulting in

462 an overall recharge of the groundwater body.

463 Our overall findings reveal that the status quo irrigation practices are favorable where groundwater

464 depletion and CO2 emissions are not such a problem, i.e., the lower part of the Indus basin (Sindh).

465 While improved irrigation technologies could be valued in areas where groundwater consumption is

466 large (i.e., center Punjab), and where groundwater depletion rates, irrigation costs and CO2 emissions

467 are high. In line with the above findings, we recommend that the improved irrigation technologies should

468 be adopted particularly in regional hotspots where status quo irrigation practices are poor.

19
469 5. CONCLUSIONS

470 In this paper, we assess the economic and environmental impact of status−quo irrigation settings and

471 alternative improved irrigation technologies in the Indus basin of Pakistan. We evaluate four scenarios

472 by using different energy sources for improved irrigation systems and compare the overall outcomes

473 with the status−quo irrigation method. Results indicate that a reduction in groundwater depletion is

474 possible for all scenarios. CO2 emissions can be reduced, particularly when solar energy is considered

475 for power supply. For all other cases, the current status−quo is superior. We further show that irrigation

476 costs increase in all scenarios compared to the status−quo. However, subsidized electricity is the

477 preferable power source for improved irrigation technologies followed by solar energy, non−subsidized

478 electricity, and diesel engines. From a cost-point view, we recommend solar energy as the second−best

479 option for farmers if no subsidized electricity is available.

480 Apart from the benefits, the solar system might require a large area for panels installation, which could

481 cause a reduction in the availability of cultivated land (Weselek et al., 2019). Nevertheless,

482 state−of−the−art agro voltaic systems could offer a solution for the future, providing energy supply,

483 reducing drought stress and water consumption and thereby improving water use efficiency (Amaducci

484 et al., 2018; Weselek et al., 2019).

485 This study is conducted assuming the current boundary conditions of agricultural production in Punjab

486 and Sindh, i.e., irrigation needs, available water and energy resources, as well as energy prices. In

487 future studies, the impact of climate change, resulting glacier melt as well as demographic changes

488 should be taken into account when developing sustainable irrigation practices for Pakistan. We also

489 recommend that future estimates of irrigation costs should also include global CO2 market prices by

490 considering externalities of CO2 emissions (Poelhekke, 2019).

491 Further aspects that should be picked up in future sustainability analysis are related to stakeholders and

492 landowners. Our study does not consider any personal preferences and choices of farmers, which might

493 result in barriers when adopting new irrigation technologies. And finally, rebound effects should also be

494 considered when new technologies hit the market (Paul et al., 2019; Pfeiffer and Lin, 2014), particularly

495 if water costs are low and solar powered pumping becomes an economic alternative on the long-term.

20
496 Acknowledgements

497 M.M. would like to thanks the German Academic Exchange Services (DAAD) and the Higher Education

498 Commission of Pakistan (HEC) for providing the opportunity of PhD studies (DAAD−91716371). We are

499 thankful to the Pakistan Metrological Department (PMD), On Farm Water Management (OFWM), and

500 Irrigation and Power Department to provide the data. The authors also like to acknowledge Dr.

501 Muhammad Imran for clarification of specific questions related to economics.

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

21
525 Appendix A1

526 Table A1. Model input data.

Dataset Units Resolution Data Sources

Surface Irrigation

Water price US$ ha−1 Spatial dataset Pakistan Statistic Bureau

(https://www.pbs.gov.pk/publications)

Labor and maintenance cost US$ ha−1 − Directorate of Agriculture (Economics and Marketing)

(http://www.amis.pk/Surveys.aspx)

Improved Irrigation

System cost US$ ha−1 − Punjab Irrigated Productivity Project (PIPIP)

Pumping plant efficiency % − Punjab Irrigated Productivity Project (PIPIP)

Average lifetime Years − (Razzaq et al., 2018)

Labor cost US$ ha−1 − (Mian et al., 2019)

Maintenance cost US$ ha−1 − (Buchanan, 2002)

Tubewells

Density diesel and electric fraction Spatial Pakistan Statistic Bureau

dataset (https://www.pbs.gov.pk/publications)

Initial cost US$ − (Qureshi et al., 2010)

Average life Years − (Johnson, 1982)

Pumping plant efficiency % − (Pervaiz, 2010)

Groundwater lift m Point dataset Irrigation and Power Department

(https://irrigation.punjab.gov.pk/page/1071)

Maintenance cost US$ − (Qureshi et al., 2003)

CO2 Emission

Groundwater depletion mm yr−1 Spatial dataset GRACE (https://www.gfz-potsdam.de)

−1
HCO3 concentration Mg L Point dataset Pakistan Council in Research in Water Resources

(http://pcrwr.gov.pk/water-quality-reports)

Emission factor − − (Brander et al., 2011; Wang et al., 2012)

Miscellaneous

Location of irrigated areas (IWMI, 2018)

Conversion rate Pakistani − − Pakistan Statistic Bureau

rupee to US$ (https://www.pbs.gov.pk/publications)

Electricity price US$ kWh−1 − Pakistan Statistic Bureau

(https://www.pbs.gov.pk/publications)

Diesel price US$ L−1 − Pakistan Statistic Bureau

(https://www.pbs.gov.pk/publications)

527
22
528 Table A2. Scenarios analysis of improved irrigation technologies.

Irrigation Costs Groundwater anomaly CO2 emission

(million US$) (mm) (million tons)
Fixed Variable Total Energy Bicarbonates Total
Surface Tubewells Irrigation Operational Maintenance consumption extraction
water system GW System
pumping operation
Acronym SWParea TCCarea ISCarea GWParea OCSarea MCarea TCarea
Baseline 31 61 0 855 0 624 1,571 −780 3.7 1.6 5.3
SC−1 31 61 1,302 350 2,134 312 4,190 280 27.5 −0.5 27
SC−2 31 61 1,330 350 560 234 2,566 280 14.5 −0.5 14
SC−3 31 61 1,330 350 1,534 282 3,588 280 14.5 −0.5 14
SC−4 31 61 1,850 350 0 505 2,797 280 1.5 −0.5 1
529

530

531

532 Figure A1. (a) Surface water and (b) groundwater share in irrigation water consumption.

23
533 References

534 Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop evapotranspiration - Guidelines for
535 computing crop water requirements - FAO Irrigation and drainage paper 56 15.

536 Amaducci, S., Yin, X., Colauzzi, M., 2018. Agrivoltaic systems to optimise land use for electric energy
537 production. Applied Energy 220, 545–561. https://doi.org/10.1016/j.apenergy.2018.03.081

538 Andersen, O.B., Seneviratne, S.I., Hinderer, J., Viterbo, P., 2005. GRACE-derived terrestrial water
539 storage depletion associated with the 2003 European heat wave. Geophysical Research
540 Letters 32. https://doi.org/10.1029/2005GL023574

541 Arshad, M., 2004. Contribution of irrigation conveyance system components to recharge potential in
542 Rechna Doab under lined and unlined options. University of Agriculture, Faisalabad, Pakistan.

543 Arshad, M., Choudhry, M.R., Ahmed, R.N., 2005. Groundwater recharge contribution from various
544 components of irrigation water conveyance system of Rechna Doab of Punjab-Pakistan.
545 Pakistan Journal of Water Resources 9, 17.

546 Ayers, R.S., Westcot, D.W., 1985. Water quality for agriculture, FAO irrigation and drainage paper.
547 Food and Agriculture Organization of the United Nations, Rome.

548 Basharat, M., Umair Ali, S., Azhar, A.H., 2014. Spatial variation in irrigation demand and supply across
549 canal commands in Punjab: a real integrated water resources management challenge. Water
550 Policy 16, 397–421. https://doi.org/10.2166/wp.2013.060

551 Bhanja, S.N., Mukherjee, A., Rangarajan, R., Scanlon, B.R., Malakar, P., Verma, S., 2019. Long-term
552 groundwater recharge rates across India by in situ measurements. Hydrol. Earth Syst. Sci. 23,
553 711–722. https://doi.org/10.5194/hess-23-711-2019

554 Brander, M., Sood, A., Wylie, C., Haughton, A., Lovell, J., 2011. Technical paper| Electricity-specific
555 emission factors for grid electricity. Ecometrica, Emissionfactors. com.

556 Brouwer, C., Prins, K., Heibloem, M., 1985. Irrigation Water Management: Training Manual No. 4:
557 Irrigation scheduling.

558 Buchanan, J.R., 2002. PB1721-Irrigation Cost Analysis Handbook. The University of Tennessee
559 Agricultural Extension Service.

560 Cheema, M.J.M., Immerzeel, W.W., Bastiaanssen, W.G.M., 2014. Spatial quantification of
561 groundwater abstraction in the irrigated Indus basin. Groundwater 52, 25–36.

562 Cools, J., Batelaan, O., Smedt, F.D., 2002. How Much Groundwater can be used? Towards
563 Quantification of Sustainable Groundwater Management. 10.

564 Cremades, R., Wang, J., Morris, J., 2015. Policies, economic incentives and the adoption of modern
565 irrigation technology in China 12.

566 Daccache, A., Ciurana, J.S., Rodriguez Diaz, J.A., Knox, J.W., 2014. Water and energy footprint of
567 irrigated agriculture in the Mediterranean region. Environ. Res. Lett. 9, 124014.
568 https://doi.org/10.1088/1748-9326/9/12/124014

569 Dewandel, B., Gandolfi, J.-M., de Condappa, D., Ahmed, S., 2008. An efficient methodology for
570 estimating irrigation return flow coefficients of irrigated crops at watershed and seasonal scale.
571 Hydrol. Process. 22, 1700–1712. https://doi.org/10.1002/hyp.6738

572 Dhillon, M.S., Kaur, S., Sood, A., Aggarwal, R., 2018. Estimation of carbon emissions from
573 groundwater pumping in central Punjab. Carbon Management 9, 425–435.
574 https://doi.org/10.1080/17583004.2018.1518107

24
575 Dogaru, D., Mauser, W., Balteanu, D., Krimly, T., Lippert, C., Sima, M., Szolgay, J., Kohnova, S.,
576 Hanel, M., Nikolova, M., Szalai, S., Frank, A., 2019. Irrigation Water Use in the Danube Basin:
577 Facts, Governance and Approach to Sustainability. Journal of Environmental Geography 12.

578 FAO, 1997. Irrigation potential in Africa: A basin approach, FAO land and water bulletin. FAO, Rome,
579 Italy.

580 Foster, S., Garduño, H., 2004. China: Towards Sustainable Groundwater Resource Use for Irrigated
581 Agriculture on the North China Plain 16.

582 Hartung, H., Pluschke, L., 2018. The Benefits and Risks of Solar Powered Irrigation 87.

583 Hassan, W., Kamran, F., 2018. A hybrid PV/utility powered irrigation water pumping system for rural
584 agricultural areas. Cogent Engineering 5, 1466383.
585 https://doi.org/10.1080/23311916.2018.1466383

586 Huffaker, R., 2008. Conservation potential of agricultural water conservation subsidies. Water
587 Resources Research 44. https://doi.org/10.1029/2007WR006183

588 Iqbal, N., Hossain, F., Lee, H., Akhter, G., 2016. Satellite Gravimetric Estimation of Groundwater
589 Storage Variations Over Indus Basin in Pakistan. IEEE J. Sel. Top. Appl. Earth Observations
590 Remote Sensing 9, 3524–3534. https://doi.org/10.1109/JSTARS.2016.2574378

591 IWMI, 2018. Irrigated area mapping: Asia and Africa :: IWMI Data & Tools. International Water
592 Management Institute (IWMI). URL http://www.iwmi.cgiar.org/2018/06/irrigated-area-mapping-
593 asia-and-africa/ (accessed 11.29.19).

594 Johnson, G.S., Sullivan, W.H., Cosgrove, D.M., Schmidt, R.D., 1999. Recharge of the Snake Rwer
595 Plain Aquifer: Transitioning from Incidental to Managed1. JAWRA Journal of the American
596 Water Resources Association 35, 123–131. https://doi.org/10.1111/j.1752-
597 1688.1999.tb05457.x

598 Johnson, S.H., 1982. Large-Scale Irrigation and Drainage Schemes in Pakistan: A Study of Rigidities
599 in Public Decision Making. Food Research Institute Studies 18, 1–32.

600 Kahlown, M.A., Raoof, A., Zubair, M., Kemper, W.D., 2007. Water use efficiency and economic
601 feasibility of growing rice and wheat with sprinkler irrigation in the Indus Basin of Pakistan.
602 Agricultural Water Management 87, 292–298. https://doi.org/10.1016/j.agwat.2006.07.011

603 Kay, M., Hatcho, N., 1992. Irrigation Water Management Training Manual. Small-Scale Pumped
604 Irrigation: Energy and Cost.

605 Khatri, K.L., Memon, A.A., Shaikh, Y., Pathan, A.F.H., Shah, S.A., Pinjani, K.K., Soomro, R., Smith,
606 R., Almani, Z., 2013. Real-Time Modelling and Optimisation for Water and Energy Efficient
607 Surface Irrigation. JWARP 05, 681–688. https://doi.org/10.4236/jwarp.2013.57068

608 Latif, M., Haider, S.S., Rashid, M.U., 2016. Adoption of High Efficiency Irrigation Systems to
609 Overcome Scarcity of Irrigation Water in Pakistan 53, 243–252.

610 Leblanc, M.J., Tregoning, P., Ramillien, G., Tweed, S.O., Fakes, A., 2009. Basin-scale, integrated
611 observations of the early 21st century multiyear drought in southeast Australia. Water
612 Resources Research 45. https://doi.org/10.1029/2008WR007333

613 Liu, Y., Huang, J., Wang, J., Rozelle, S., 2008. Determinants of agricultural water saving technology
614 adoption: An empirical study of 10 provinces of China. Ecol. Econ 4, 462–472.

615 Long, D., Chen, X., Scanlon, B.R., Wada, Y., Hong, Y., Singh, V.P., Chen, Y., Wang, C., Han, Z.,
616 Yang, W., 2016. Have GRACE satellites overestimated groundwater depletion in the
617 Northwest India Aquifer? Sci Rep 6, 24398. https://doi.org/10.1038/srep24398

25
618 Mahinda, A.J., Gachene, C.K.K., Kilasara, M., 2016. Economic Impact of Drip Irrigation Regimes on
619 Sorghum Production in Semi-arid Areas of Tanzania, in: Lal, R., Kraybill, D., Hansen, D.O.,
620 Singh, B.R., Mosogoya, T., Eik, L.O. (Eds.), Climate Change and Multi-Dimensional
621 Sustainability in African Agriculture: Climate Change and Sustainability in Agriculture. Springer
622 International Publishing, Cham, pp. 227–240. https://doi.org/10.1007/978-3-319-41238-2_13

623 Mian, M.A., Lukeová, D., Krepl, V., 2019. Farmer’s Perception Regarding Effectiveness of Drip
624 Irrigation System in Attock, Pakistan. Presented at the Tropentag 2019, Kassel, Germany, p.
625 1.

626 Mishra, V., Asoka, A., Vatta, K., Lall, U., 2018. Groundwater Depletion and Associated CO 2
627 Emissions in India. Earth’s Future 6, 1672–1681. https://doi.org/10.1029/2018EF000939

628 Mongat, A.S., Arshad, M., Bakhsh, A., Shakoor, A., Anjum, L., Hameed, A., Shamim, F., 2015. Design,
629 Installation and Evaluation of Solar Drip Irrigation System at Mini Dam Command Area.
630 Pakistan Journal of Agricultural Sciences 8.

631 Multsch, S., Al-Rumaikhani, Y.A., Frede, H.-G., Breuer, L., 2013. A Site-sPecific Agricultural water
632 Requirement and footprint Estimator (SPARE:WATER 1.0). Geoscientific Model Development
633 6, 1043–1059. https://doi.org/10.5194/gmd-6-1043-2013

634 Muzammil, M., Zahid, A., Breuer, L., 2020. Water Resources Management Strategies for Irrigated
635 Agriculture in the Indus Basin of Pakistan. Water 12, 1429. https://doi.org/10.3390/w12051429

636 Narayanamoorthy, A., Bhattarai, M., Jothi, P., 2018. An assessment of the economic impact of drip
637 irrigation in vegetable production in India. Agri. Econ. Rese. Revi. 31, 105.
638 https://doi.org/10.5958/0974-0279.2018.00010.1

639 Nicks, A.D., 1998. GLEAMS Model Evaluation — Hydrology and Erosion Components, in: Boardman,
640 J., Favis-Mortlock, D. (Eds.), Modelling Soil Erosion by Water, NATO ASI Series. Springer,
641 Berlin, Heidelberg, pp. 55–63. https://doi.org/10.1007/978-3-642-58913-3_6

642 OECD, 2017. Water Risk Hotspots for Agriculture. OECD, Paris.

643 Paul, C., Techen, A.-K., Robinson, J.S., Helming, K., 2019. Rebound effects in agricultural land and
644 soil management: Review and analytical framework. Journal of Cleaner Production 227,
645 1054–1067. https://doi.org/10.1016/j.jclepro.2019.04.115

646 Pérez-Blanco, C.D., Hrast-Essenfelder, A., Perry, C., 2020. Irrigation Technology and Water
647 Conservation: A Review of the Theory and Evidence. Review of Environmental Economics
648 and Policy 14, 216–239. https://doi.org/10.1093/reep/reaa004

649 Pervaiz, S., 2010. Groundwater Management Using Vertical Electrical Sounding Survey and Tubewell
650 Auditing at Farmers’ Fields. University of Agriculture, Faisalabad, Pakistan.

651 Pfeiffer, L., Lin, C.-Y.C., 2014. Does efficient irrigation technology lead to reduced groundwater
652 extraction? Empirical evidence. Journal of Environmental Economics and Management 67,
653 189–208. https://doi.org/10.1016/j.jeem.2013.12.002

654 Poelhekke, S., 2019. How expensive should CO2 be? Fuel for the political debate on optimal climate
655 policy. Heliyon 5, e02936. https://doi.org/10.1016/j.heliyon.2019.e02936

656 Qamar, M.U., Azmat, M., Abbas, A., Usman, M., Shahid, M.A., Khan, Z.M., 2018. Water Pricing and
657 Implementation Strategies for the Sustainability of an Irrigation System: A Case Study within
658 the Command Area of the Rakh Branch Canal. Water 10, 509.
659 https://doi.org/10.3390/w10040509

660 Qureshi, A.S., 2014. Reducing Carbon Emissions through Improved Irrigation Management: A Case
661 Study from Pakistan. Irrig. and Drain. 63, 132–138. https://doi.org/10.1002/ird.1795

26
662 Qureshi, A.S., Gill, M.A., Sarwar, A., 2010. Sustainable groundwater management in Pakistan:
663 challenges and opportunities. Irrigation and Drainage 59, 107–116.
664 https://doi.org/10.1002/ird.455

665 Qureshi, A.S., McCornick, P.G., Qadir, M., Aslam, Z., 2008. Managing salinity and waterlogging in the
666 Indus Basin of Pakistan. Agricultural Water Management 95, 1–10.
667 https://doi.org/10.1016/j.agwat.2007.09.014

668 Qureshi, A.S., Shah, T., Akhtar, M., 2003. The groundwater economy of Pakistan, Working paper.
669 International Water Management Institute, Lahore.

670 Ramphull, M., Surroop, D., 2017. Greenhouse gas emission factor for the energy sector in Mauritius.
671 Journal of Environmental Chemical Engineering 5, 5994–6000.
672 https://doi.org/10.1016/j.jece.2017.11.027

673 Razzaq, A., Rehman, A., Qureshi, A.H., Javed, I., Saqib, R., Iqbal, N., 2018. An Economic Analysis of
674 High Efficiency Irrigation Systems in Punjab, Pakistan. Sarhad Journal of Agriculture 9.

675 Rizwan, M., Bakhsh, A., Li, X., Anjum, L., Jamal, K., Hamid, S., 2018. Evaluation of the Impact of
676 Water Management Technologies on Water Savings in the Lower Chenab Canal Command
677 Area, Indus River Basin. Water 10, 681. https://doi.org/10.3390/w10060681

678 Roblin, S., 2016. Solar-powered irrigation: A solution to water management in agriculture? Renewable
679 Energy Focus 17, 205–206. https://doi.org/10.1016/j.ref.2016.08.013

680 Rodrigues, G.C., Martins, J.D., Petry, M.T., Carlesso, R., Paredes, P., Rosa, R.D., 2012. Economic
681 impacts assessment of deficit irrigation and commodity prices. Application to maize in
682 southern Brazil. Presented at the International Conference on Agricultural Engineering,
683 Valencia, Spain.

684 Rzepecka, Z., Birylo, M., 2020. Groundwater Storage Changes Derived from GRACE and GLDAS on
685 Smaller River Basins—A Case Study in Poland. Geosciences 10, 124.
686 https://doi.org/10.3390/geosciences10040124

687 Save, H., Bettadpur, S., Tapley, B.D., 2016. High-resolution CSR GRACE RL05 mascons. Journal of
688 Geophysical Research: Solid Earth 121, 7547–7569. https://doi.org/10.1002/2016JB013007

689 Shah, T., Kishore, A., 2012. Solar-Powered Pump Irrigation and India’s Groundwater Economy 8.

690 Shamsudduha, M., Taylor, R.G., Ahmed, K.M., Zahid, A., 2011. The impact of intensive groundwater
691 abstraction on recharge to a shallow regional aquifer system: evidence from Bangladesh.
692 Hydrogeol J 19, 901–916. https://doi.org/10.1007/s10040-011-0723-4

693 Shekhar, S., Kumar, S., Densmore, A.L., van Dijk, W.M., Sinha, R., Kumar, M., Joshi, S.K., Rai, S.P.,
694 Kumar, D., 2020. Modelling water levels of northwestern India in response to improved
695 irrigation use efficiency. Scientific Reports 10, 13452. https://doi.org/10.1038/s41598-020-
696 70416-0

697 Shen, H., Leblanc, M., Tweed, S., Liu, W., 2015. Groundwater depletion in the Hai River Basin, China,
698 from in situ and GRACE observations. Hydrological Sciences Journal 60, 671–687.
699 https://doi.org/10.1080/02626667.2014.916406

700 Soto-García, M., Martínez-Alvarez, V., García-Bastida, P.A., Alcon, F., Martin-Gorriz, B., 2013. Effect
701 of water scarcity and modernisation on the performance of irrigation districts in south-eastern
702 Spain. Agricultural Water Management 124, 11–19.
703 https://doi.org/10.1016/j.agwat.2013.03.019

704 Tang, Y., Hooshyar, M., Zhu, T., Ringler, C., Sun, A.Y., Long, D., Wang, D., 2017. Reconstructing
705 annual groundwater storage changes in a large-scale irrigation region using GRACE data and

27
706 Budyko model. Journal of Hydrology 551, 397–406.
707 https://doi.org/10.1016/j.jhydrol.2017.06.021

708 Tiwari, D., Dinar, A., 2002. Role and use of economic incentives in irrigated agriculture. WORLD
709 BANK TECHNICAL PAPER 103–122.

710 Tiwari, V.M., Wahr, J., Swenson, S., 2009. Dwindling groundwater resources in northern India, from
711 satellite gravity observations. Geophysical Research Letters 36.
712 https://doi.org/10.1029/2009GL039401

713 Velasco-Muñoz, J.F., Aznar-Sánchez, J.A., Batlles-delaFuente, A., Fidelibus, M.D., 2019. Sustainable
714 Irrigation in Agriculture: An Analysis of Global Research. Water 11, 1758.
715 https://doi.org/10.3390/w11091758

716 Wang, J., Li, Y., Huang, J., Yan, T., Sun, T., 2017. Growing water scarcity, food security and
717 government responses in China. Global Food Security, Food Security Governance in Latin
718 America 14, 9–17. https://doi.org/10.1016/j.gfs.2017.01.003

719 Wang, J., Rothausen, S.G., Conway, D., Zhang, L., Xiong, W., Holman, I.P., Li, Y., 2012. China’s
720 water–energy nexus: greenhouse-gas emissions from groundwater use for agriculture.
721 Environmental Research Letters 7, 014035.

722 Weselek, A., Ehmann, A., Zikeli, S., Lewandowski, I., Schindele, S., Högy, P., 2019. Agrophotovoltaic
723 systems: applications, challenges, and opportunities. A review. Agron. Sustain. Dev. 39, 35.
724 https://doi.org/10.1007/s13593-019-0581-3

725 Wichelns, D., Oster, J.D., 2006. Sustainable irrigation is necessary and achievable, but direct costs
726 and environmental impacts can be substantial. Agricultural Water Management, Responsible
727 Management of Water in Agriculture 86, 114–127. https://doi.org/10.1016/j.agwat.2006.07.014

728 Wood, W.W., Hyndman, D.W., 2017. Groundwater Depletion: A Significant Unreported Source of
729 Atmospheric Carbon Dioxide. Earth’s Future 5, 1133–1135.
730 https://doi.org/10.1002/2017EF000586

731 Wu, Q., Si, B., He, H., Wu, P., 2019. Determining Regional-Scale Groundwater Recharge with GRACE
732 and GLDAS. Remote Sensing 11, 154. https://doi.org/10.3390/rs11020154

733 Yang, J., Olsson, A., Yan, J., Chen, B., 2014. A Hybrid Life-cycle Assessment of CO2 Emissions of a
734 PV Water Pumping System in China. Energy Procedia, International Conference on Applied
735 Energy, ICAE2014 61, 2871–2875. https://doi.org/10.1016/j.egypro.2014.12.326

736 Zafar, U., Arshad, M., Cheema, M.J.M., Ahmad, R., 2020. Sensor Based Drip Irrigation to Enhance
737 Crop Yield and Water Productivity in Semi-Arid Climatic Region of Pakistan 57, 1293–1301.

738 Zhang, B., Fu, Z., Wang, J., Zhang, L., 2019. Farmers’ adoption of water-saving irrigation technology
739 alleviates water scarcity in metropolis suburbs: A case study of Beijing, China. Agricultural
740 Water Management 212, 349–357. https://doi.org/10.1016/j.agwat.2018.09.021

741 Zou, X., Li, Y., Cremades, R., Gao, Q., Wan, Y., Qin, X., 2013. Cost-effectiveness analysis of water-
742 saving irrigation technologies based on climate change response: A case study of China.
743 Agricultural Water Management 129, 9–20. https://doi.org/10.1016/j.agwat.2013.07.004

744 Zou, X., Li, Y., Li, K., Cremades, R., Gao, Q., Wan, Y., Qin, X., 2015. Greenhouse gas emissions from
745 agricultural irrigation in China. Mitig Adapt Strateg Glob Change 20, 295–315.
746 https://doi.org/10.1007/s11027-013-9492-9

747

28
Manuscript (without Author Details) Click here to view linked References

1 Economic and Environmental Impact Assessment of Sustainable Future

2 Irrigation Practices in the Indus Basin of Pakistan

3 Abstract

4 Agriculture plays a substantial role in the economy of Pakistan, which is strongly associated with the

5 irrigated plains of the Indus basin. Insecure water supply and poor irrigation practices are major threats

6 to water security and the environment of the country. In this research, we investigate the economic and

7 environmental feasibility of alternative improved irrigation technologies over the status quo (2002−2018)

8 for the Punjab and Sindh provinces, representing about 90% of the total irrigated area in the country.

9 We estimate the site−specific irrigation costs, groundwater anomalies, and CO2 emissions. Improved

10 irrigation technologies consider different energy sources including solar power in combination with

11 changes in the irrigation method. The status quo irrigation costs are estimated to 1,301 million US$ yr−1,

12 its groundwater depletion to 6.3 mm yr−1 and CO2 emissions to 4.12 million t yr−1, of which 96% originate

13 from energy consumption and 4% via bicarbonate extraction from groundwater. Irrigation costs of

14 improved irrigation scenarios increase with all energy sources compared to the status quo, which is

15 mainly based on gravity-fed surface irrigation. This is because of additional variable and fixed costs for

16 system’s operation in the scenarios. Of these, subsidized electricity induces the lowest costs for farmers

17 with 63% extra costs followed by solar energy with 77%. However, groundwater depletion can even be

18 reversed with up to 35% rise in groundwater levels via improved irrigation technologies. Solar−powered

19 irrigation can break down CO2 emissions by 81% whilst other energy sources boost emissions by up to

20 410%. Results suggest that there is an extremely opposing development between economic and

21 ecological preferences in the scenarios, requiring stakeholders to negotiate viable trade-offs.

22 Keywords: irrigation technologies, irrigation costs, groundwater anomaly, groundwater depletion, CO2

23 emissions, solar power supply

24 Highlights:
25  Current irrigation practices risk water security and adversely affect the environment

26  Improved irrigation methods tend to reduce groundwater depletion

27  Economic benefits of irrigation technologies are insufficient over the status quo

28  Solar-powered irrigation technologies could massively save CO 2 emissions

1
29 1. INTRODUCTION

30 Agriculture of Pakistan is based on irrigation where rainfall marginally meets 15% of crop water

31 requirements (Qureshi et al., 2010). Irrigated agriculture is associated with the Indus Basin, which

32 provides irrigation water on a supply basis. Surface water resources are unable to fulfill the actual

33 irrigation demands owing to the high evapotranspiration and salinity environment in the plain (Basharat

34 et al., 2014). Groundwater covers 40−60% of the irrigation needs to meet the deficit in surface water

35 supplies (Cheema et al., 2014). The intensive groundwater use results in a decline of the water table

36 and accumulation of soil salinity, which originates from saline pockets of the aquifers (Qureshi et al.,

37 2008). Meanwhile, the increasing trend of groundwater pumping has become energy exhaustive. It

38 reduces the farmer’s income because of high irrigation costs and leads to massive greenhouse gas

39 releases, mainly CO2 emissions, through energy consumption (Qureshi, 2014). Furthermore, the role of

40 groundwater depletion in greenhouse emissions is still unaccounted, which can be a significant emission

41 source associated with bicarbonate extraction (Mishra et al., 2018; Wood and Hyndman, 2017). Despite

42 the poor situation of water resources availability in the country, inefficient irrigation methods dominate

43 in the region, with losses of up to 50% of available water in the fields (Rizwan et al., 2018).

44 In many arid and semi−arid countries, where water resources are limited and depleting rapidly, there is

45 pressure to reduce water consumption for the water security of the growing population (Kahlown et al.,

46 2007). Previous studies indicate that improved irrigation technologies enable farmers to cope with water

47 scarcity and insecure water supply (Muzammil et al., 2020; Wang et al., 2017; Zhang et al., 2019).

48 However, the impact of irrigation technologies remains a critical topic for sustainable irrigation. The

49 consequences of improved irrigation technologies vary among regions because of differences in

50 cost−benefits, off−farm environmental impacts, and social preferences (Wichelns and Oster, 2006).

51 Economic impact assessment should be part of the evaluation process and support the

52 decision−making. For example, Zou et al. (2013) analyzed water−saving strategies based on the climate

53 change response for China and proposed that channel lining is a preferable strategy from an economic

54 perspective compared to pressurized irrigation practices. Mahinda et al. (2016) investigated the

55 economic impact of sorghum production via drip irrigation in semi−arid regions of Tanzania and

56 recommended that two irrigations per day are beneficial to get higher economic returns.

57 Narayanamoorthy et al. (2018) studied the economic impact of drip irrigation on vegetable crops and

2
58 their findings indicate that the pressurized irrigation system offers high net returns compared to

59 conventional irrigation methods.

60 In the context of environmental consequences, irrigation development can have severe effects on the

61 investment in irrigation projects at regional and basin levels (Dogaru et al., 2019; FAO, 1997;

62 Velasco−Muñoz et al., 2019). For example, Daccache et al. (2014) projected that a pressurized irrigation

63 system is capable to increase irrigation efficiency, but CO2 emissions increase due to additional energy

64 consumption compared to a gravity−fed surface irrigation system. Shekhar et al. (2020) showed that

65 technology changes could have the potential to mitigate groundwater depletion through pressure

66 reduction on water resources. However, the lower percolation from fields with improved water saving

67 irrigation techniques may reduce aquifer recharge (Johnson et al., 1999). In Pakistan, the potential of

68 high−efficiency irrigation technologies related to water saving have been recognized. Several studies

69 revealed that it is possible to overcome water scarcity in Pakistan through the adoption of high−efficiency

70 irrigation systems (Latif et al., 2016; Rizwan et al., 2018; Zafar et al., 2020). Meanwhile, previous studies

71 show that future power supply for irrigation technologies should consider changes in the energy source,

72 including solar power supply (Hassan and Kamran, 2018; Mongat et al., 2015). Nevertheless, the

73 economic and environmental impacts of these technologies are still unknown over the status−quo

74 irrigation settings. An inclusive analysis of the cost−effectiveness of improved irrigation technologies

75 coping with ecological impact can support economic development and environmental sustainability in

76 the region.

77 In this study, we compare the economic and environmental impacts of the status−quo irrigation settings

78 with alternative, improved irrigation technologies. We use a coupled economic−environmental−modeling

79 framework to estimate the irrigation costs, groundwater depletion, and CO2 emissions to understand the

80 return on investment and environmental effects. We consider improved, more sustainable irrigation

81 technologies that differ from the status−quo irrigation practices in terms of water consumption, irrigation

82 costs, and energy use. As the water consumption via improved irrigation technologies is lower than that

83 of conventional irrigation, the effect of groundwater recharge through surplus irrigation is diminishing,

84 which we take also into account. Furthermore, improving the established irrigation system needs a high

85 initial investment and, in the case where the gravity−fed irrigation system is replaced, additional

86 operational energy costs and associated CO2 emissions come into play, which are also analyzed.

3
 87 The objectives of the current study are: (1) to investigate the economic impact of improved irrigation

88 technologies over status−quo irrigation practices, (2) to compare groundwater depletion and CO2

89 emissions of the status−quo irrigation settings with improved irrigation practices, and (3) to develop

90 alternative scenarios for improved irrigation technologies and identify sustainable energy use options in

91 the irrigation agriculture of Pakistan.

92 2. MATERIALS AND METHODS

93 2.1 Study Area

94 The study focuses on the irrigated areas of Punjab and Sindh provinces in the Indus basin of Pakistan.

95 Together, these cover 17 million ha (Figure 1), representing 90% of the total irrigated area in the country.

96 The topography of the plain falls from north to south, ranging from 540 to 4 m above mean sea level. The

97 basin has an arid to semi−arid climate with complex hydrological processes due to spatial and temporal

98 variation in the rainfall, temperature, land use, and water consumption. The average annual rainfall

99 amounts to 379 mm (2002−2018), while maximum temperature ranges from 34−44°C in the summer

100 (Apr−Sep) and 20−28°C in the winter (Dec−Feb). The annual potential evapotranspiration varies from

101 1,200 to 2,050 mm from the north to the south. Crops are harvested in two cropping seasons called

102 Kharif (wet season; Apr−Sep) and Rabi (dry season; Oct−Mar). Sugarcane, cotton, and rice are

103 dominant crops in the Kharif while wheat is a major crop in the Rabi season.

104 There are five major tributaries to the Indus (Indus, Chenab, Ravi, Jhelum, and Sutlej), which supply

105 irrigation water via a network of canals and watercourses. The provincial governments distribute the

106 surface water among farmers according to the landholding size and collect the water charges two times

107 in a year in the Kharif and Rabi seasons. The water charges vary from province to province i.e., the

108 Punjab government collects at a flat rate despite which crop is grown while it varies in Sindh by crop to

109 crop. Farmers using additional groundwater recourses via private units (tubewells), operate with diesel

110 engines or mains power for groundwater pumping. The government provides subsidized electricity to

111 farmers. However, diesel operated tubewells are common among farmers with 87% of share because

112 they have a lower initial investment than electric operated tubewells. Crops are widely irrigated via

113 surface irrigation with an application efficiency of 45−60%. Improved irrigation systems (drip and

114 sprinkler) are installed only in a limited area (50,000 ha) through a subsidized program of the World

4
115 Bank and the government of Punjab in the frame of the Punjab Irrigated-Agriculture Productivity

116 Improvement Project (PIPIP).

117
118 Figure 1. Map of the study area.

119 2.2 Modeling Framework

120 In this study, we develop an economic−environmental−modeling framework to evaluate the economic

121 and environmental impacts of the status−quo irrigation practices and a variety of scenarios with

122 improved irrigation technologies. The model is written in python by using the SciPy package. The

123 modeling approach uses gridded data and makes use of information such as the irrigation requirements,

124 harvested area, crop water consumption, groundwater level, energy use required for pumping water,

125 water prices and energy costs. The methodological steps of the modeling framework are summarized

126 in Figure 2, and the calculation methods are described in section 2.3. The input data and references

127 used in this study are given in Table A1.

128
5
129 Figure 2. Methodological steps to estimate the economic and environmental impact of the status−quo

130 irrigation settings and the improved irrigation technologies.

131 2.3 Calculation Methods

132 2.3.1 Irrigation Requirements

133 Regional irrigation requirements (IRRarea) are calculated for the entire area by combining all crop’s

134 productive (IRRprod) and unproductive (IRRunprod) consumptions of irrigation water along with the leaching

135 requirements (LR) (Eq1). IRRprod contributes to crop growth, while IRRunprod covers the water losses in

136 line with the efficiency of the irrigation system (IRReffi). IRRunprod does not result in crop production and

137 percolates from the root zone to the groundwater or evaporates at the soil surface. These water losses

138 partially cover the LR (Multsch et al., 2013). The LR is an additional amount of water that is otherwise

139 needed to leach salts from the root zone by assuming the salinity tolerance limit of each crop and the

140 salt fraction in the irrigation water (Ayers and Westcot, 1985).

IRRprod
141 IRR area = + LR (Eq1)
IRReffi

142 with IRRarea, IRRprod, and LR given in (km3 yr−1) and IRReffi in percentage (%).

143 In this study, we use data on the site−specific productive irrigation water IRR prod and leaching

144 requirement LR (2002−2016) from a recently published study (Muzammil et al., 2020). The dataset holds

145 information with a spatial resolution of 0.063° for Pakistan. Muzammil et al. (2020) used

146 SPARE:WATER, an open−source model integrated into a geographical information system to estimate

147 the crop water balance at the grid level (Multsch et al., 2013). SPARE:WATER follows the FAO56

148 guidelines to determine crop water requirements (Allen et al., 1998) and calculates the potential LR in

149 line with the salinity tolerance limit of crops and the salt fractions in the irrigation water. For this study,

150 we extended the simulation period of 2002−2016 from Muzammil et al. (2020) and included the years

151 2017 and 2018. A detailed list of model input data and parameters required to run the model is given in

152 Muzammil et al. (2020). The climatic data is obtained from the Pakistan Metrological Department, while

153 information on crops is provided from the Pakistan Statistics Bureau. The efficiencies of irrigation

154 systems are taken from the FAO dataset as 60%, 75%, and 90% for surface, sprinkler, and drip irrigation,

155 respectively (Brouwer et al., 1985).

6
156 2.3.2 Surface Water and Groundwater Use

157 As surface water and groundwater are used in the Indus basin to meet the irrigation demand, we

158 estimate the surface water share (km 3 yr−1) from a dataset of annual canals supply. The data is

159 preprocessed to exclude the off−farm water losses assuming a conveyance efficiency of 70% (Arshad

160 et al., 2005; Cheema et al., 2014). The volume of groundwater abstraction (km 3 yr−1) is determined

161 by subtracting the available surface water in the fields from the regional irrigation requirements.

162 2.3.3 Irrigation Costs

163 The total annual irrigation costs of the entire area (TCarea; million US$ yr−1) are estimated by adding the

164 total fixed costs (TFCarea) and total variable costs (TVCarea) (Eq2):

165 TCarea = TFCarea + TVCarea (Eq2)

166 1) Total Fixed Irrigation Costs

167 Total fixed irrigation costs per area (TFCarea) are estimated by adding its components on a regional

168 basis (Eq3), i.e., surface water prices (SWParea), tubewell construction costs (TCCarea), and irrigation

169 system costs (ISCarea).

170 TFCarea = SWParea + TCCarea + ISCarea (Eq3)

171 where SWParea (million US$ yr−1) results from summing up the products of costs occurring for surface

172 water for crop irrigation (US$ ha−1) times their harvested area (ha yr−1). TCCarea (million US$ yr−1) is

173 estimated by dividing the initial costs of all tubewells (million US$) for a given area from their average

174 lifetimes (years). The initial costs of tubewells are projected by combining the construction costs of all

175 diesel and electric operated tubewells. The TCCarea vary and depend on groundwater level and power

176 required for pumping groundwater (Qureshi et al., 2010). ISCarea are calculated by summing up the

177 product of all crops’ irrigation system costs per hectare (US$ ha−1) times their harvested area (ha yr−1).

178 Note that the annual irrigation system costs are split in halves for the crops of the two growing seasons

179 Kharif and Rabi, respectively. The annual irrigation system costs are derived from dividing the initial

180 costs of the systems by their average lifetimes (years). The status−quo irrigation system is based on

181 gravity, therefore ISCarea for surface irrigation are negligible (Latif et al., 2016). The initial costs of the

182 improved irrigation system vary from crop to crop and by changing the power source.

7
183 2) Total Variable Irrigation Costs

184 We use Eq4 to calculate the regional value of the total variable irrigation costs (TVCarea) by adding its

185 components, i.e., the operational costs (OCarea) and the maintenance costs (MCarea):

186 TVCarea = OCarea + MCarea (Eq4)

187 We further divide OCarea into two parts, i.e., the groundwater pumping costs (GPCarea) and the

188 operational costs of the irrigation system (OCSarea). Accordingly, the maintenance costs (MCarea) are

189 composed of the maintenance costs of the tubewells (MCTarea), and the maintenance costs of the

190 irrigation system (MCSarea).

191 The GPCarea (million US$ yr−1) is based on the costs for the energy sources diesel and electricity. The

192 share of diesel and electric pumping in the study area is estimated by using the fraction of diesel and

193 electric operated tubewells in a grid cell. GPCarea are projected by adding the groundwater pumping

194 costs of diesel (GPCarea(d)) and electric (GPCarea(e)) operated tubewells. Both, GPCarea(d) and GPCarea(e),

195 are calculated by summing up the product of the tubewell abstracted groundwater volumes (m3) times

196 the pumping costs (US$ m−3). Pumping costs are calculated by multiplying the energy consumed (kWh)

197 per m3 pumped groundwater and the energy price (US$ kWh−1). The energy consumption is determined

198 from Eq5 where V, TDH, and Ƞpp are abstracted groundwater volume (m 3), total dynamic head (m), and

199 pumping plant efficiency (%), respectively (Kay and Hatcho, 1992). In this study, the energy price for

200 the electric source is used directly as the given electricity price in the country (US$ kWh−1) while for

201 diesel consumption, fuel price (US$ L−1) is converted into an energy price (US$ kWh−1) by multiplying

202 fuel price with a conversion factor of 0.11 (Kay and Hatcho, 1992).

V×TDH
203 Energy (kWh) = (Eq5)
367 × ηpp

204 The OCSarea (million US$ yr−1) consists of the energy and labor costs of the irrigation system. The energy

205 costs for the surface irrigation method are negligible as its operation is based on gravity (Khatri et al.,

206 2013). For the pressurized irrigation system, energy demand is estimated by multiplying the energy

207 required to run the irrigation system (kWh yr−1) and the energy price (US$ kWh−1), being either diesel or

208 electricity. The energy consumption is estimated from Eq5 where TDH indicates the total head required

209 to run the irrigation system, i.e., the operational head, friction losses, and suction lift. Labor costs are

210 calculated by summing up the product of labor charges (US$ ha−1) and the harvested area (ha yr−1).

8
211 MCTarea (million US$ yr−1) is calculated by summing up the annual maintenance costs of diesel and

212 electric operated tubewells in the region. The maintenance costs of diesel and electric operated

213 tubewells are estimated by multiplying the maintenance costs per tubewell and the number of electric

214 and diesel operated tubewells in the study area.

215 Finally, the MCSarea (million US$ yr−1) contains repair and cleaning costs of the watercourses, which is

216 calculated by multiplying the maintenance costs (US$ ha−1) and the total harvested area (ha yr−1). For

217 improved irrigation technologies, maintenance costs cover repair and security costs of the system. We

218 estimate it as 5% of the total operational costs (Buchanan, 2002).

219 2.3.4 Groundwater Storage

220 The annual aquifer recharge (mm) is estimated from the Water Table Fluctuation method by adding the

221 groundwater storage anomaly (mm) and the depth of pumped groundwater from the aquifer (mm)

222 (Bhanja et al., 2019; Wu et al., 2019). We use monthly terrestrial water storage data from the Gravity

223 Recovery and Climate Experiment (GRACE) to estimate the groundwater storage anomaly. GRACE

224 data has been validated for Pakistan in past studies (Iqbal et al., 2016; Tang et al., 2017). In this study,

225 we apply the GRACE Mascon solution, which does not need post−processing filtering and which is less

226 depending on scale factors (Save et al., 2016). Groundwater storage anomaly is derived by subtracting

227 the surface water storage (soil moisture, canopy water, snow water) from the terrestrial water storage.

228 The surface water storage is estimated up to 2 m of the soil column from the land surface model (NOAH)

229 dataset of the GLDAS product, which has been used in several regions where in situ measurements are

230 not available (Andersen et al., 2005; Leblanc et al., 2009; Rzepecka and Birylo, 2020; Tiwari et al.,

231 2009).

232 Further, we calculate the contributions of the fields’ percolation losses to total recharge. For the

233 status−quo irrigation settings, it is estimated from published data (Arshad, 2004). This data is simulated

234 via the GLEAMS hydrological model, which is used at the field scale to estimate the movement of water

235 content through percolation and contribution of recharge to the groundwater (Nicks, 1998). Accordingly,

236 water percolates from fields to the groundwater storage in the Indus basin of Pakistan at a rate of

237 0.314 mm day−1. It is assumed that this percolation is negligible for improved irrigation technologies

238 where irrigation surplus is marginal (Dewandel et al., 2008).

9
239 2.3.5 Carbon Dioxide Emissions

240 We estimate CO2 emissions from the status−quo irrigation practices and improved irrigation

241 technologies, where energy consumption and bicarbonate extraction from the groundwater are

242 considered as the major emissions sources.

243 1) CO2 Emissions from Energy Consumption

244 There are two energy consumption sources related to irrigation in the study area, i.e., groundwater

245 pumping and irrigation system operation. The annual mass of CO2 emissions depends on the amount

246 of energy consumed (kWh yr−1) and the types of these energy sources (Ramphull and Surroop, 2017),

247 represented by their respective emission factors. We apply a fixed emission factor for diesel engines of

248 0.32021 kg CO2 kWh−1 (Wang et al., 2012). For electricity, we calculate with a constant value of 0.47337

249 kg CO2 kWh−1 based on information on the major energy sources for power production in Pakistan

250 (Brander et al., 2011). Note that the status−quo irrigation system is based on gravity, therefore, no CO2

251 is emitted.

252 2) CO2 Emissions from Bicarbonates Extraction

253 In this study, we assume that the CO2 concentrations in recharging groundwater and pumped

254 groundwater are the same. If groundwater recharge is equal to the abstraction, there are no CO2

255 emissions (Wood and Hyndman, 2017). Hitherto, CO2 is emitted if groundwater is depleted and CO2 is

256 sequestered in the aquifer in cases of rising groundwater levels. We estimate CO2

257 emissions/sequestration (million t CO2 yr−1) by multiplying CO2 concentrations in the groundwater (mg

258 L−1) and groundwater depletion/increase (m3). Groundwater depletion/increase is estimated by

259 multiplying the groundwater storage anomaly (m) and surface area of the plain (m 2).

260 The CO2 concentrations in the groundwater depend on atmospheric CO2 dissolved in water, which

261 enters the groundwater body via percolation and thus depends on the groundwater recharge rate. During

262 solution, CO2 and H2O split into hydrogen (H+) and bicarbonate (HCO3−) ions (Eq6).

263 CO2 + H2 O → H + + HCO3− (Eq6)

264 It is assumed that half of the mass of total bicarbonates present in the groundwater originates from this

265 separation. While another half is formed when the CaCO 3 rich rock in the aquifer reacts with hydrogen

266 ions (H+) (Mishra et al., 2018) (Eq7):

10
267 H + + CaCO3 → HCO3− + Ca2+ (Eq7)

268 Depending on the resulting bicarbonate concentration in the groundwater, CO2 evolves into the

269 atmosphere according to Eq8 when groundwater is pumped.

270 Ca(HCO3 )2 → CO2 + H2 O + CaCO3 (Eq8)

271 The resulting CO2 concentration (mg L−1) in the groundwater is calculated by multiplying the molecular

272 mass ratio of HCO3− and CO2 with the bicarbonate concentration (mg L−1) (Eq9).

1 44
273 CO2 Concentration = HCO3− × (Eq9)
2 61

274 2.4 Scenario Development

275 We develop four future scenarios (SC−1 to SC−4) to derive a potential optimum plan for irrigation that

276 reduces the irrigation costs, groundwater depletion, and CO2 emissions in the Indus basin. Scenarios

277 are established by changing the status−quo irrigation methods (gravity−fed surface irrigation) to

278 improved irrigation technologies as this has been identified as a preferable solution to reduce total

279 amount of irrigation water (Muzammil et al. 2020). The year 2018 is considered as a baseline to which

280 scenarios are compared. We keep the harvested area from the baseline in the scenarios and convert

281 surface irrigation to drip irrigation for row crops and to sprinkler irrigation for field crops. The scenarios

282 are classified according to the energy sources required to operate the revised irrigation system. In SC−1,

283 the diesel engines are used to operate the irrigation system, SC−2 is run on electricity but assumes

284 subsidized prices as status quo conditions, SC−3 is also based on electricity, but considers the actual

285 energy price, and SC−4 is defined by using solar energy.

286 3. RESULTS

287 3.1 Water Consumption and Irrigation Costs

288 The shares of surface and groundwater in irrigation water are shown in Figure A1. The irrigation water

289 consumption (IRRarea) and the total irrigation costs (TCarea) for 2002−2018 are presented in Figure 3.

290 Results show that the southern part of Punjab has the highest irrigation requirements while the upper

291 portion of Punjab and the whole parts of Sindh have relatively lower irrigation water consumptions

292 (Figure 3a). We find strong inter−annual variation in the total irrigation requirements with the highest

293 IRRarea in 2002 (177 km3 yr−1) and the lowest in 2015 (130 km3 yr−1) (Figure 3b). A Mann−Kendall test
11
294 reveals that there is no trend in IRRarea from 2002−2018 (p = 0.23). Average IRRarea is estimated to 157

295 km3 yr−1, of which groundwater accounts for 52% (82 km3 yr−1) and surface water contributes to 48%

296 (75 km3 yr−1). Diesel pumping has the largest share in groundwater abstraction with 83%, followed by

297 electric pumping of 17%. Results of the irrigation costs also show a substantial variation in space and

298 year−to−year (Figure 3.1). The southern region of Punjab has the highest annual irrigation cost

299 compared to other parts of the study area (Figure 3c). The highest irrigation costs are calculated for

300 2014 (1,837 million US$) and the lowest one for 2003 (718 million US$) (Figure 3d). We find an overall

301 significant increasing trend for irrigation costs from 2002−2018 (R2 = 0.43, slope = 41.6, p=0.001). The

302 years 2015 and 2016 are striking with lower costs, which are due to the combined effect of a lower

303 IRRarea and reduced fuel prices compared to other years. The average irrigation costs for 2002−2018

304 are calculated to 1,301 million US$, of which fixed cost components account for 8% (85 million US$)

305 and variable costs account for by far the largest amount (1,216 million US$). Groundwater pumping

306 costs have the largest share in irrigation costs with 60%, followed by maintenance costs (32%), surface

307 water prices (3%), and tubewell construction costs (5%). Diesel pumping costs have a dominant part in

308 the total pumping costs of groundwater with 93%, while the electric pumping holds only 7%.

12
309
310 Figure 3. Irrigation water consumption (a, b) and irrigation cost (c, d) from 2002−2018.

311 3.2 Estimates of Groundwater Depletion

312 We project the groundwater storage from 2002−2018 by estimating the groundwater recharge and

313 abstraction in the study area. The results show that the northern part of the plain (Punjab province) faces

314 the largest depletion rate (−11 mm yr−1) while an increase in groundwater level (4 mm yr−1) is observed

315 in the southern part of the plain (Sindh province) (figure 4a). Overall, the groundwater storage anomaly

316 is significantly decreasing (R2 = 0.39, slope = −3.93, p=0.02) in the study area from 2002−2018 (figure

317 3.2b) at a rate of −6.3 mm yr−1 (−1.35 km3 yr−1). Annual differences are substantial, with the highest

318 depletion rate in the year 2018 (−78 mm yr−1; −16.7 km3 yr−1) and the largest surplus in groundwater

319 storage in 2003 (43 mm yr−1; 9.2 km3 yr−1). Overall, we do neither find significant trends for net

320 groundwater recharge (p = 0.06) nor for abstraction (p = 0.38). Further, the average net recharge rate

321 is estimated to 380 mm yr−1, of which 69% (263 mm yr−1) are contributed by from natural resources

322 (precipitation, and particular leaching from rivers, water bodies and canals) while the cropping fields add

323 another 31% (117 mm yr−1) as percolation losses from unproductive irrigation.

13
324
325 Figure 4. (a) Average groundwater depletion from 2002 to 2018, (b) Time series trend (2002−2018) of

326 groundwater abstraction, net recharge, natural recharge, and storage anomaly.

327 3.3 Estimates of CO2 Emissions

328 We estimate CO2 emissions from 2002−2018 according to the emission sources, i.e., energy

329 consumption and bicarbonate extraction from depleted groundwater volume (Figure 5). The southern

330 part of Punjab depicts the highest CO2 emissions from energy consumption (Figure 5a) while the upper

331 portion of Punjab shows the highest CO2 emissions due to groundwater depletion (Figure 5b). The

332 results further reveal that about 4.12 million t CO2 yr−1 are emitted in the plain, of which 96% (3.95 million

333 t yr−1) result from energy consumption while 4% (0.17 million t yr−1) are stemming from groundwater

334 depletion. The largest CO2 emissions are produced in the year 2018 (5.42 million t) and the lowest one

335 in 2015 (2.15 million t) (Figure 5c). Further, CO2 emissions from groundwater depletion are highly

336 variable over time with a maximum in 2018 (1.58 million t). For several years, we found even negative

337 values (i.e., an increase of the CO2 storage) due to a surplus of groundwater recharge over groundwater

338 abstraction. This results in rather substantial net storage of CO2 in 2003 (−0.93 million t). With regard to

339 the energy source, diesel pumping has a larger share (87%) than CO2 emissions from electric pumping.

14
340
341 Figure 5. Average annual CO2 emissions from (a) energy consumption and (b) groundwater depletion,

342 (c) temporal development of CO2 emissions from 2002−2018. (Note that the color-codes in the maps

343 (a) and (b) vary by a factor of 10)

344 3.4 Scenario Analysis

345 Scenarios are investigated to derive the optimum energy source for improved irrigation technologies

346 and compare the results with the status−quo irrigation method. We establish four scenarios SC1−4 to

347 identify the effect of improved irrigation technologies on irrigation costs, groundwater depletion, and CO2

348 emissions for more sustainable irrigation practices by using different energy sources in each scenario.

349 The changes in irrigation costs, groundwater depletion, and CO2 emissions for all scenarios are

350 presented in Figure 6 and Table A2. In SC−1, we change the gravity driven status−quo irrigation settings

351 with irrigation technologies and consider diesel as the primary energy source. The results indicate that

352 irrigation costs and CO2 emissions increase up to 170% and 410%, respectively, while the groundwater

353 depletion is reduced by up to 135%. SC−2 focuses on changing the status quo irrigation settings with

354 irrigation technologies that run on subsidized electricity from mains power. We find an increase in

355 irrigation costs and CO2 emissions of up to 63% and 165%, respectively. Meanwhile, the groundwater

356 depletion rate decreases by up to 135%. The scenario SC−3 has the same settings as SC−2 but we
15
357 use actual prices for electricity. In consequence, we observe an increase in irrigation costs of up to

358 130% of the baseline scenario. In SC−4, solar−powered irrigation technologies are used instead of the

359 surface irrigation method. The results show that irrigation costs increase by up to 77% while CO2

360 emissions and groundwater depletion are reduced by up to 81% and 135%, respectively.

361

362 Figure 6. Scenarios analysis for (a) irrigation costs (b) CO2 emissions and (c) groundwater balance,

363 via improved irrigation technologies.

364 4. DISCUSSIONS

365 4.1 Economic Impact of Irrigation Methods

366 In the status−quo conditions, the average irrigation water consumption in the study area is 157 km3 yr−1,

367 of which surface water contributes 48% and groundwater 52%. Despite the small difference in water

368 consumption from surface water and groundwater, there is a vast margin between prices with 3% for

369 surface water and 63% for groundwater of the total irrigation costs (1,301 million US$), respectively.

370 Alternatively, scenarios indicate that improved irrigation technologies can reduce water consumption by

371 32%, which could lead to a reduction in groundwater share of up to 61%, with at the same time 55%

372 decreasing groundwater pumping costs. However, improved irrigation technologies raise total irrigation

373 costs owing to the initial and running costs of the system. Scenarios specify that the operation of

374 improved irrigation technologies via subsidized electricity is an optimal scenario among others for

16
375 farmer’s perspective where irrigation costs increase by 63% compared to the status quo. Solar energy

376 is the second most feasible power source when no subsidized electricity is at hand, but still, irrigation

377 costs increase by 77% compared to that of the status quo. Highest costs are found for diesel operated

378 systems which boost irrigation costs by up to 170%. In short, the economic benefits of improved irrigation

379 technologies are insufficient over the status−quo practices to cover the additional expenditure of the

380 irrigation system. This is in line with various other studies that recognized that improved irrigation

381 technologies can increase farmer’s expenditures via capital investments and running costs (Huffaker,

382 2008; Pérez−Blanco et al., 2020; Soto−García et al., 2013). For example, Rodrigues et al. (2012) studied

383 the comparative advantages of drip and sprinkler irrigation in southern Brazil and concluded that

384 economic benefits from water−saving technologies are insufficient to recover the initial costs of the

385 system. Numerous studies revealed that the implication of improved irrigation technologies is a

386 challenge owing to an extra burden of investment compared to surface irrigation. In Pakistan, despite

387 the various awareness campaigns in the last three decades to introduce improved irrigation

388 technologies, farmers are still not willing to adopt the technologies because of the high initial costs of

389 the system. Thus, governments should provide subsidies to farmers for sustainable water consumption

390 (Cremades et al., 2015; Liu et al., 2008; Tiwari and Dinar, 2002), such as in the World Bank funded

391 Punjab Irrigated−Agriculture Productivity Improvement Project with a size of 50,000 ha. Such types of

392 projects have the capability to promote water−saving technologies among farmers. However, it is

393 doubtful that such a technical shift is sustainable from an economic viewpoint.

394 Part of this problem might be arising from the very low surface water prices in Pakistan, which do not

395 promote changing towards more efficient, but costly irrigation technologies. Qamar et al. (2018) studied

396 the implication strategies of improved irrigation technologies in the Indus basin of Pakistan and

397 concluded that the surface water prices should be higher to promote irrigation technologies among

398 farmers. We recommend that a comprehensive analysis should be conducted to study the adoption

399 strategies of improved irrigation technologies by changing the water prices. Such an analysis should not

400 only consider pure economic aspect, but also take into account societal barriers and personal

401 preferences as well as choices from farmers.

17
402 4.2 CO2 Emissions from Irrigation Practices

403 We estimated CO2 emissions from irrigation practices in the Indus basin of Pakistan by assuming

404 emissions from energy consumption and bicarbonate extraction. At the status−quo settings, diesel or

405 electric pumps are used to pump groundwater, which produces 96% of the total CO2 emissions (3.95

406 million t). Our estimates indicate that bicarbonate extraction is not a significant emissions source,

407 amounting to about 4% of the total CO2 emissions (0.17 million t), although groundwater makes up a

408 significant part of the irrigation water in the Indus basin. Mishra et al. (2018) estimated the annual CO2

409 emissions from groundwater bicarbonate extraction to around 0.72 million t, which is not a significant

410 emissions source either compared to energy consumption through groundwater pumping. Wood and

411 Hyndman (2017) calculated CO2 emissions from bicarbonates extraction in the USA and determined

412 that annual 1.7 million t of CO2 are released from this source. Despite a 10fold higher rate as compared

413 to the groundwater mediated CO2 emissions in the Indus basin, the total share of bicarbonate extraction

414 on US CO2 emissions is small with less than 0.5% (estimated from data published by Wood an

415 Hyndman, 2017).

416 Past studies proposed several strategies to reduce CO2 emissions from groundwater pumping. For

417 example, Shah and Kishore (2012) recommended on−site solar and wind energy for groundwater

418 pumping. However, the authors show serious concern that the availability of renewable energy will

419 encourage the farmers to pump additional groundwater because of the currently low pumping costs.

420 Dhillon et al. (2018) projected that an improvement in pumping plant efficiency could also reduce CO2

421 emissions. Zou et al. (2015) showed indirect effects through general water savings of improved irrigation

422 systems and subsequent lower CO2 emissions because of a reduced groundwater demand. However,

423 improved irrigation technologies might require further energy to run the system, which in turn can

424 increase overall CO2 emissions. Daccache et al. (2014) studied the environmental impact of irrigation

425 practices in the Mediterranean region of Spain. Similar to our results, they revealed that CO2 emissions

426 increased by 135% for improved irrigation technologies compared to the old−fashioned, gravity−based

427 surface irrigation method.

428 We estimate CO2 emissions for different scenarios of improved irrigation technologies by combining

429 emissions from groundwater pumping and irrigation system operation. Our results indicate that diesel

430 engines and mains power electricity are both detrimental energy sources for advancing irrigation

431 technologies compared to the status−quo settings, simply because of the huge increase in CO2
18
432 emissions by 410% and 165%, respectively. However, solar energy operating systems are most

433 effective, which can reduce CO2 emissions even of the status−quo technology by 81%. Many studies

434 revealed that solar energy is the best option for improved irrigation technologies for sustainable

435 development in a region or basin (Hartung and Pluschke, 2018; Roblin, 2016; Yang et al., 2014).

436 4.3 Groundwater Depletion

437 In the study area, the average groundwater depletion is 6.3 mm yr−1, which is comparatively low. For

438 example, Long et al. (2016) estimated the groundwater depletion to be 31 mm yr−1 in the Northwest

439 Indian Aquifer. Shen et al. (2015) investigated groundwater storage anomaly in the Hai River Basin

440 China and reported that groundwater is depleting at a rate of 17 mm yr−1. Despite the lower depletion

441 rate in the Indus basin, our estimates show that the groundwater depletion trend is increasing from

442 2002−2018. Tang et al. (2017) confirmed that groundwater storage is diminishing in the Indus basin. It

443 has been predicted that the depletion rate in the Indus basin will increase by 50% in 2050 compared to

444 the groundwater depletion trend in 2005 (OECD, 2017). We believe that an increasing trend of

445 groundwater depletion is a serious matter and quick measures are needed for sustainable groundwater

446 usage. In the sense of sustainability, the groundwater abstraction rate should be lower than the recharge

447 rate (Cools et al., 2002; Foster and Garduño, 2004; Shamsudduha et al., 2011). Our results show that

448 improved irrigation technologies are capable to reduce groundwater utilization compared to status−quo

449 irrigation. However, such improvements can also have negative side effects like the reduction

450 percolation losses from fields. These apparent negative losses lead, on the one hand, to a leaching of

451 salts from the soil (Muzammil et al., 2020) and, on the other hand, also to groundwater recharge. Overall,

452 our estimates verify that the reduction in groundwater abstraction is larger than field losses, resulting in

453 an overall recharge of the groundwater body.

454 Our overall findings reveal that the status quo irrigation practices are favorable where groundwater

455 depletion and CO2 emissions are not such a problem, i.e., the lower part of the Indus basin (Sindh).

456 While improved irrigation technologies could be valued in areas where groundwater consumption is

457 large (i.e., center Punjab), and where groundwater depletion rates, irrigation costs and CO2 emissions

458 are high. In line with the above findings, we recommend that the improved irrigation technologies should

459 be adopted particularly in regional hotspots where status quo irrigation practices are poor.

19
460 5. CONCLUSIONS

461 In this paper, we assess the economic and environmental impact of status−quo irrigation settings and

462 alternative improved irrigation technologies in the Indus basin of Pakistan. We evaluate four scenarios

463 by using different energy sources for improved irrigation systems and compare the overall outcomes

464 with the status−quo irrigation method. Results indicate that a reduction in groundwater depletion is

465 possible for all scenarios. CO2 emissions can be reduced, particularly when solar energy is considered

466 for power supply. For all other cases, the current status−quo is superior. We further show that irrigation

467 costs increase in all scenarios compared to the status−quo. However, subsidized electricity is the

468 preferable power source for improved irrigation technologies followed by solar energy, non−subsidized

469 electricity, and diesel engines. From a cost-point view, we recommend solar energy as the second−best

470 option for farmers if no subsidized electricity is available.

471 Apart from the benefits, the solar system might require a large area for panels installation, which could

472 cause a reduction in the availability of cultivated land (Weselek et al., 2019). Nevertheless,

473 state−of−the−art agro voltaic systems could offer a solution for the future, providing energy supply,

474 reducing drought stress and water consumption and thereby improving water use efficiency (Amaducci

475 et al., 2018; Weselek et al., 2019).

476 This study is conducted assuming the current boundary conditions of agricultural production in Punjab

477 and Sindh, i.e., irrigation needs, available water and energy resources, as well as energy prices. In

478 future studies, the impact of climate change, resulting glacier melt as well as demographic changes

479 should be taken into account when developing sustainable irrigation practices for Pakistan. We also

480 recommend that future estimates of irrigation costs should also include global CO2 market prices by

481 considering externalities of CO2 emissions (Poelhekke, 2019).

482 Further aspects that should be picked up in future sustainability analysis are related to stakeholders and

483 landowners. Our study does not consider any personal preferences and choices of farmers, which might

484 result in barriers when adopting new irrigation technologies. And finally, rebound effects should also be

485 considered when new technologies hit the market (Paul et al., 2019; Pfeiffer and Lin, 2014), particularly

486 if water costs are low and solar powered pumping becomes an economic alternative on the long-term.

487

488

20
489 Appendix A1

490 Table A1. Model input data.

Dataset Units Resolution Data Sources

Surface Irrigation

Water price US$ ha−1 Spatial dataset Pakistan Statistic Bureau

(https://www.pbs.gov.pk/publications)

Labor and maintenance cost US$ ha−1 − Directorate of Agriculture (Economics and Marketing)

(http://www.amis.pk/Surveys.aspx)

Improved Irrigation

System cost US$ ha−1 − Punjab Irrigated Productivity Project (PIPIP)

Pumping plant efficiency % − Punjab Irrigated Productivity Project (PIPIP)

Average lifetime Years − (Razzaq et al., 2018)

Labor cost US$ ha−1 − (Mian et al., 2019)

Maintenance cost US$ ha−1 − (Buchanan, 2002)

Tubewells

Density diesel and electric fraction Spatial Pakistan Statistic Bureau

dataset (https://www.pbs.gov.pk/publications)

Initial cost US$ − (Qureshi et al., 2010)

Average life Years − (Johnson, 1982)

Pumping plant efficiency % − (Pervaiz, 2010)

Groundwater lift m Point dataset Irrigation and Power Department

(https://irrigation.punjab.gov.pk/page/1071)

Maintenance cost US$ − (Qureshi et al., 2003)

CO2 Emission

Groundwater depletion mm yr−1 Spatial dataset GRACE (https://www.gfz-potsdam.de)

−1
HCO3 concentration Mg L Point dataset Pakistan Council in Research in Water Resources

(http://pcrwr.gov.pk/water-quality-reports)

Emission factor − − (Brander et al., 2011; Wang et al., 2012)

Miscellaneous

Location of irrigated areas (IWMI, 2018)

Conversion rate Pakistani − − Pakistan Statistic Bureau

rupee to US$ (https://www.pbs.gov.pk/publications)

Electricity price US$ kWh−1 − Pakistan Statistic Bureau

(https://www.pbs.gov.pk/publications)

Diesel price US$ L−1 − Pakistan Statistic Bureau

(https://www.pbs.gov.pk/publications)

491
21
492 Table A2. Scenarios analysis of improved irrigation technologies.

Irrigation Costs Groundwater anomaly CO2 emission

(million US$) (mm) (million tons)
Fixed Variable Total Energy Bicarbonates Total
Surface Tubewells Irrigation Operational Maintenance consumption extraction
water system GW System
pumping operation
Acronym SWParea TCCarea ISCarea GWParea OCSarea MCarea TCarea
Baseline 31 61 0 855 0 624 1,571 −780 3.7 1.6 5.3
SC−1 31 61 1,302 350 2,134 312 4,190 280 27.5 −0.5 27
SC−2 31 61 1,330 350 560 234 2,566 280 14.5 −0.5 14
SC−3 31 61 1,330 350 1,534 282 3,588 280 14.5 −0.5 14
SC−4 31 61 1,850 350 0 505 2,797 280 1.5 −0.5 1
493

494

495

496 Figure A1. (a) Surface water and (b) groundwater share in irrigation water consumption.

22
497 References

498 Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop evapotranspiration - Guidelines for
499 computing crop water requirements - FAO Irrigation and drainage paper 56 15.

500 Amaducci, S., Yin, X., Colauzzi, M., 2018. Agrivoltaic systems to optimise land use for electric energy
501 production. Applied Energy 220, 545–561. https://doi.org/10.1016/j.apenergy.2018.03.081

502 Andersen, O.B., Seneviratne, S.I., Hinderer, J., Viterbo, P., 2005. GRACE-derived terrestrial water
503 storage depletion associated with the 2003 European heat wave. Geophysical Research
504 Letters 32. https://doi.org/10.1029/2005GL023574

505 Arshad, M., 2004. Contribution of irrigation conveyance system components to recharge potential in
506 Rechna Doab under lined and unlined options. University of Agriculture, Faisalabad, Pakistan.

507 Arshad, M., Choudhry, M.R., Ahmed, R.N., 2005. Groundwater recharge contribution from various
508 components of irrigation water conveyance system of Rechna Doab of Punjab-Pakistan.
509 Pakistan Journal of Water Resources 9, 17.

510 Ayers, R.S., Westcot, D.W., 1985. Water quality for agriculture, FAO irrigation and drainage paper.
511 Food and Agriculture Organization of the United Nations, Rome.

512 Basharat, M., Umair Ali, S., Azhar, A.H., 2014. Spatial variation in irrigation demand and supply across
513 canal commands in Punjab: a real integrated water resources management challenge. Water
514 Policy 16, 397–421. https://doi.org/10.2166/wp.2013.060

515 Bhanja, S.N., Mukherjee, A., Rangarajan, R., Scanlon, B.R., Malakar, P., Verma, S., 2019. Long-term
516 groundwater recharge rates across India by in situ measurements. Hydrol. Earth Syst. Sci. 23,
517 711–722. https://doi.org/10.5194/hess-23-711-2019

518 Brander, M., Sood, A., Wylie, C., Haughton, A., Lovell, J., 2011. Technical paper| Electricity-specific
519 emission factors for grid electricity. Ecometrica, Emissionfactors. com.

520 Brouwer, C., Prins, K., Heibloem, M., 1985. Irrigation Water Management: Training Manual No. 4:
521 Irrigation scheduling.

522 Buchanan, J.R., 2002. PB1721-Irrigation Cost Analysis Handbook. The University of Tennessee
523 Agricultural Extension Service.

524 Cheema, M.J.M., Immerzeel, W.W., Bastiaanssen, W.G.M., 2014. Spatial quantification of
525 groundwater abstraction in the irrigated Indus basin. Groundwater 52, 25–36.

526 Cools, J., Batelaan, O., Smedt, F.D., 2002. How Much Groundwater can be used? Towards
527 Quantification of Sustainable Groundwater Management. 10.

528 Cremades, R., Wang, J., Morris, J., 2015. Policies, economic incentives and the adoption of modern
529 irrigation technology in China 12.

530 Daccache, A., Ciurana, J.S., Rodriguez Diaz, J.A., Knox, J.W., 2014. Water and energy footprint of
531 irrigated agriculture in the Mediterranean region. Environ. Res. Lett. 9, 124014.
532 https://doi.org/10.1088/1748-9326/9/12/124014

533 Dewandel, B., Gandolfi, J.-M., de Condappa, D., Ahmed, S., 2008. An efficient methodology for
534 estimating irrigation return flow coefficients of irrigated crops at watershed and seasonal scale.
535 Hydrol. Process. 22, 1700–1712. https://doi.org/10.1002/hyp.6738

536 Dhillon, M.S., Kaur, S., Sood, A., Aggarwal, R., 2018. Estimation of carbon emissions from
537 groundwater pumping in central Punjab. Carbon Management 9, 425–435.
538 https://doi.org/10.1080/17583004.2018.1518107

23
539 Dogaru, D., Mauser, W., Balteanu, D., Krimly, T., Lippert, C., Sima, M., Szolgay, J., Kohnova, S.,
540 Hanel, M., Nikolova, M., Szalai, S., Frank, A., 2019. Irrigation Water Use in the Danube Basin:
541 Facts, Governance and Approach to Sustainability. Journal of Environmental Geography 12.

542 FAO, 1997. Irrigation potential in Africa: A basin approach, FAO land and water bulletin. FAO, Rome,
543 Italy.

544 Foster, S., Garduño, H., 2004. China: Towards Sustainable Groundwater Resource Use for Irrigated
545 Agriculture on the North China Plain 16.

546 Hartung, H., Pluschke, L., 2018. The Benefits and Risks of Solar Powered Irrigation 87.

547 Hassan, W., Kamran, F., 2018. A hybrid PV/utility powered irrigation water pumping system for rural
548 agricultural areas. Cogent Engineering 5, 1466383.
549 https://doi.org/10.1080/23311916.2018.1466383

550 Huffaker, R., 2008. Conservation potential of agricultural water conservation subsidies. Water
551 Resources Research 44. https://doi.org/10.1029/2007WR006183

552 Iqbal, N., Hossain, F., Lee, H., Akhter, G., 2016. Satellite Gravimetric Estimation of Groundwater
553 Storage Variations Over Indus Basin in Pakistan. IEEE J. Sel. Top. Appl. Earth Observations
554 Remote Sensing 9, 3524–3534. https://doi.org/10.1109/JSTARS.2016.2574378

555 IWMI, 2018. Irrigated area mapping: Asia and Africa :: IWMI Data & Tools. International Water
556 Management Institute (IWMI). URL http://www.iwmi.cgiar.org/2018/06/irrigated-area-mapping-
557 asia-and-africa/ (accessed 11.29.19).

558 Johnson, G.S., Sullivan, W.H., Cosgrove, D.M., Schmidt, R.D., 1999. Recharge of the Snake Rwer
559 Plain Aquifer: Transitioning from Incidental to Managed1. JAWRA Journal of the American
560 Water Resources Association 35, 123–131. https://doi.org/10.1111/j.1752-
561 1688.1999.tb05457.x

562 Johnson, S.H., 1982. Large-Scale Irrigation and Drainage Schemes in Pakistan: A Study of Rigidities
563 in Public Decision Making. Food Research Institute Studies 18, 1–32.

564 Kahlown, M.A., Raoof, A., Zubair, M., Kemper, W.D., 2007. Water use efficiency and economic
565 feasibility of growing rice and wheat with sprinkler irrigation in the Indus Basin of Pakistan.
566 Agricultural Water Management 87, 292–298. https://doi.org/10.1016/j.agwat.2006.07.011

567 Kay, M., Hatcho, N., 1992. Irrigation Water Management Training Manual. Small-Scale Pumped
568 Irrigation: Energy and Cost.

569 Khatri, K.L., Memon, A.A., Shaikh, Y., Pathan, A.F.H., Shah, S.A., Pinjani, K.K., Soomro, R., Smith,
570 R., Almani, Z., 2013. Real-Time Modelling and Optimisation for Water and Energy Efficient
571 Surface Irrigation. JWARP 05, 681–688. https://doi.org/10.4236/jwarp.2013.57068

572 Latif, M., Haider, S.S., Rashid, M.U., 2016. Adoption of High Efficiency Irrigation Systems to
573 Overcome Scarcity of Irrigation Water in Pakistan 53, 243–252.

574 Leblanc, M.J., Tregoning, P., Ramillien, G., Tweed, S.O., Fakes, A., 2009. Basin-scale, integrated
575 observations of the early 21st century multiyear drought in southeast Australia. Water
576 Resources Research 45. https://doi.org/10.1029/2008WR007333

577 Liu, Y., Huang, J., Wang, J., Rozelle, S., 2008. Determinants of agricultural water saving technology
578 adoption: An empirical study of 10 provinces of China. Ecol. Econ 4, 462–472.

579 Long, D., Chen, X., Scanlon, B.R., Wada, Y., Hong, Y., Singh, V.P., Chen, Y., Wang, C., Han, Z.,
580 Yang, W., 2016. Have GRACE satellites overestimated groundwater depletion in the
581 Northwest India Aquifer? Sci Rep 6, 24398. https://doi.org/10.1038/srep24398

24
582 Mahinda, A.J., Gachene, C.K.K., Kilasara, M., 2016. Economic Impact of Drip Irrigation Regimes on
583 Sorghum Production in Semi-arid Areas of Tanzania, in: Lal, R., Kraybill, D., Hansen, D.O.,
584 Singh, B.R., Mosogoya, T., Eik, L.O. (Eds.), Climate Change and Multi-Dimensional
585 Sustainability in African Agriculture: Climate Change and Sustainability in Agriculture. Springer
586 International Publishing, Cham, pp. 227–240. https://doi.org/10.1007/978-3-319-41238-2_13

587 Mian, M.A., Lukeová, D., Krepl, V., 2019. Farmer’s Perception Regarding Effectiveness of Drip
588 Irrigation System in Attock, Pakistan. Presented at the Tropentag 2019, Kassel, Germany, p.
589 1.

590 Mishra, V., Asoka, A., Vatta, K., Lall, U., 2018. Groundwater Depletion and Associated CO 2
591 Emissions in India. Earth’s Future 6, 1672–1681. https://doi.org/10.1029/2018EF000939

592 Mongat, A.S., Arshad, M., Bakhsh, A., Shakoor, A., Anjum, L., Hameed, A., Shamim, F., 2015. Design,
593 Installation and Evaluation of Solar Drip Irrigation System at Mini Dam Command Area.
594 Pakistan Journal of Agricultural Sciences 8.

595 Multsch, S., Al-Rumaikhani, Y.A., Frede, H.-G., Breuer, L., 2013. A Site-sPecific Agricultural water
596 Requirement and footprint Estimator (SPARE:WATER 1.0). Geoscientific Model Development
597 6, 1043–1059. https://doi.org/10.5194/gmd-6-1043-2013

598 Muzammil, M., Zahid, A., Breuer, L., 2020. Water Resources Management Strategies for Irrigated
599 Agriculture in the Indus Basin of Pakistan. Water 12, 1429. https://doi.org/10.3390/w12051429

600 Narayanamoorthy, A., Bhattarai, M., Jothi, P., 2018. An assessment of the economic impact of drip
601 irrigation in vegetable production in India. Agri. Econ. Rese. Revi. 31, 105.
602 https://doi.org/10.5958/0974-0279.2018.00010.1

603 Nicks, A.D., 1998. GLEAMS Model Evaluation — Hydrology and Erosion Components, in: Boardman,
604 J., Favis-Mortlock, D. (Eds.), Modelling Soil Erosion by Water, NATO ASI Series. Springer,
605 Berlin, Heidelberg, pp. 55–63. https://doi.org/10.1007/978-3-642-58913-3_6

606 OECD, 2017. Water Risk Hotspots for Agriculture. OECD, Paris.

607 Paul, C., Techen, A.-K., Robinson, J.S., Helming, K., 2019. Rebound effects in agricultural land and
608 soil management: Review and analytical framework. Journal of Cleaner Production 227,
609 1054–1067. https://doi.org/10.1016/j.jclepro.2019.04.115

610 Pérez-Blanco, C.D., Hrast-Essenfelder, A., Perry, C., 2020. Irrigation Technology and Water
611 Conservation: A Review of the Theory and Evidence. Review of Environmental Economics
612 and Policy 14, 216–239. https://doi.org/10.1093/reep/reaa004

613 Pervaiz, S., 2010. Groundwater Management Using Vertical Electrical Sounding Survey and Tubewell
614 Auditing at Farmers’ Fields. University of Agriculture, Faisalabad, Pakistan.

615 Pfeiffer, L., Lin, C.-Y.C., 2014. Does efficient irrigation technology lead to reduced groundwater
616 extraction? Empirical evidence. Journal of Environmental Economics and Management 67,
617 189–208. https://doi.org/10.1016/j.jeem.2013.12.002

618 Poelhekke, S., 2019. How expensive should CO2 be? Fuel for the political debate on optimal climate
619 policy. Heliyon 5, e02936. https://doi.org/10.1016/j.heliyon.2019.e02936

620 Qamar, M.U., Azmat, M., Abbas, A., Usman, M., Shahid, M.A., Khan, Z.M., 2018. Water Pricing and
621 Implementation Strategies for the Sustainability of an Irrigation System: A Case Study within
622 the Command Area of the Rakh Branch Canal. Water 10, 509.
623 https://doi.org/10.3390/w10040509

624 Qureshi, A.S., 2014. Reducing Carbon Emissions through Improved Irrigation Management: A Case
625 Study from Pakistan. Irrig. and Drain. 63, 132–138. https://doi.org/10.1002/ird.1795

25
626 Qureshi, A.S., Gill, M.A., Sarwar, A., 2010. Sustainable groundwater management in Pakistan:
627 challenges and opportunities. Irrigation and Drainage 59, 107–116.
628 https://doi.org/10.1002/ird.455

629 Qureshi, A.S., McCornick, P.G., Qadir, M., Aslam, Z., 2008. Managing salinity and waterlogging in the
630 Indus Basin of Pakistan. Agricultural Water Management 95, 1–10.
631 https://doi.org/10.1016/j.agwat.2007.09.014

632 Qureshi, A.S., Shah, T., Akhtar, M., 2003. The groundwater economy of Pakistan, Working paper.
633 International Water Management Institute, Lahore.

634 Ramphull, M., Surroop, D., 2017. Greenhouse gas emission factor for the energy sector in Mauritius.
635 Journal of Environmental Chemical Engineering 5, 5994–6000.
636 https://doi.org/10.1016/j.jece.2017.11.027

637 Razzaq, A., Rehman, A., Qureshi, A.H., Javed, I., Saqib, R., Iqbal, N., 2018. An Economic Analysis of
638 High Efficiency Irrigation Systems in Punjab, Pakistan. Sarhad Journal of Agriculture 9.

639 Rizwan, M., Bakhsh, A., Li, X., Anjum, L., Jamal, K., Hamid, S., 2018. Evaluation of the Impact of
640 Water Management Technologies on Water Savings in the Lower Chenab Canal Command
641 Area, Indus River Basin. Water 10, 681. https://doi.org/10.3390/w10060681

642 Roblin, S., 2016. Solar-powered irrigation: A solution to water management in agriculture? Renewable
643 Energy Focus 17, 205–206. https://doi.org/10.1016/j.ref.2016.08.013

644 Rodrigues, G.C., Martins, J.D., Petry, M.T., Carlesso, R., Paredes, P., Rosa, R.D., 2012. Economic
645 impacts assessment of deficit irrigation and commodity prices. Application to maize in
646 southern Brazil. Presented at the International Conference on Agricultural Engineering,
647 Valencia, Spain.

648 Rzepecka, Z., Birylo, M., 2020. Groundwater Storage Changes Derived from GRACE and GLDAS on
649 Smaller River Basins—A Case Study in Poland. Geosciences 10, 124.
650 https://doi.org/10.3390/geosciences10040124

651 Save, H., Bettadpur, S., Tapley, B.D., 2016. High-resolution CSR GRACE RL05 mascons. Journal of
652 Geophysical Research: Solid Earth 121, 7547–7569. https://doi.org/10.1002/2016JB013007

653 Shah, T., Kishore, A., 2012. Solar-Powered Pump Irrigation and India’s Groundwater Economy 8.

654 Shamsudduha, M., Taylor, R.G., Ahmed, K.M., Zahid, A., 2011. The impact of intensive groundwater
655 abstraction on recharge to a shallow regional aquifer system: evidence from Bangladesh.
656 Hydrogeol J 19, 901–916. https://doi.org/10.1007/s10040-011-0723-4

657 Shekhar, S., Kumar, S., Densmore, A.L., van Dijk, W.M., Sinha, R., Kumar, M., Joshi, S.K., Rai, S.P.,
658 Kumar, D., 2020. Modelling water levels of northwestern India in response to improved
659 irrigation use efficiency. Scientific Reports 10, 13452. https://doi.org/10.1038/s41598-020-
660 70416-0

661 Shen, H., Leblanc, M., Tweed, S., Liu, W., 2015. Groundwater depletion in the Hai River Basin, China,
662 from in situ and GRACE observations. Hydrological Sciences Journal 60, 671–687.
663 https://doi.org/10.1080/02626667.2014.916406

664 Soto-García, M., Martínez-Alvarez, V., García-Bastida, P.A., Alcon, F., Martin-Gorriz, B., 2013. Effect
665 of water scarcity and modernisation on the performance of irrigation districts in south-eastern
666 Spain. Agricultural Water Management 124, 11–19.
667 https://doi.org/10.1016/j.agwat.2013.03.019

668 Tang, Y., Hooshyar, M., Zhu, T., Ringler, C., Sun, A.Y., Long, D., Wang, D., 2017. Reconstructing
669 annual groundwater storage changes in a large-scale irrigation region using GRACE data and

26
670 Budyko model. Journal of Hydrology 551, 397–406.
671 https://doi.org/10.1016/j.jhydrol.2017.06.021

672 Tiwari, D., Dinar, A., 2002. Role and use of economic incentives in irrigated agriculture. WORLD
673 BANK TECHNICAL PAPER 103–122.

674 Tiwari, V.M., Wahr, J., Swenson, S., 2009. Dwindling groundwater resources in northern India, from
675 satellite gravity observations. Geophysical Research Letters 36.
676 https://doi.org/10.1029/2009GL039401

677 Velasco-Muñoz, J.F., Aznar-Sánchez, J.A., Batlles-delaFuente, A., Fidelibus, M.D., 2019. Sustainable
678 Irrigation in Agriculture: An Analysis of Global Research. Water 11, 1758.
679 https://doi.org/10.3390/w11091758

680 Wang, J., Li, Y., Huang, J., Yan, T., Sun, T., 2017. Growing water scarcity, food security and
681 government responses in China. Global Food Security, Food Security Governance in Latin
682 America 14, 9–17. https://doi.org/10.1016/j.gfs.2017.01.003

683 Wang, J., Rothausen, S.G., Conway, D., Zhang, L., Xiong, W., Holman, I.P., Li, Y., 2012. China’s
684 water–energy nexus: greenhouse-gas emissions from groundwater use for agriculture.
685 Environmental Research Letters 7, 014035.

686 Weselek, A., Ehmann, A., Zikeli, S., Lewandowski, I., Schindele, S., Högy, P., 2019. Agrophotovoltaic
687 systems: applications, challenges, and opportunities. A review. Agron. Sustain. Dev. 39, 35.
688 https://doi.org/10.1007/s13593-019-0581-3

689 Wichelns, D., Oster, J.D., 2006. Sustainable irrigation is necessary and achievable, but direct costs
690 and environmental impacts can be substantial. Agricultural Water Management, Responsible
691 Management of Water in Agriculture 86, 114–127. https://doi.org/10.1016/j.agwat.2006.07.014

692 Wood, W.W., Hyndman, D.W., 2017. Groundwater Depletion: A Significant Unreported Source of
693 Atmospheric Carbon Dioxide. Earth’s Future 5, 1133–1135.
694 https://doi.org/10.1002/2017EF000586

695 Wu, Q., Si, B., He, H., Wu, P., 2019. Determining Regional-Scale Groundwater Recharge with GRACE
696 and GLDAS. Remote Sensing 11, 154. https://doi.org/10.3390/rs11020154

697 Yang, J., Olsson, A., Yan, J., Chen, B., 2014. A Hybrid Life-cycle Assessment of CO2 Emissions of a
698 PV Water Pumping System in China. Energy Procedia, International Conference on Applied
699 Energy, ICAE2014 61, 2871–2875. https://doi.org/10.1016/j.egypro.2014.12.326

700 Zafar, U., Arshad, M., Cheema, M.J.M., Ahmad, R., 2020. Sensor Based Drip Irrigation to Enhance
701 Crop Yield and Water Productivity in Semi-Arid Climatic Region of Pakistan 57, 1293–1301.

702 Zhang, B., Fu, Z., Wang, J., Zhang, L., 2019. Farmers’ adoption of water-saving irrigation technology
703 alleviates water scarcity in metropolis suburbs: A case study of Beijing, China. Agricultural
704 Water Management 212, 349–357. https://doi.org/10.1016/j.agwat.2018.09.021

705 Zou, X., Li, Y., Cremades, R., Gao, Q., Wan, Y., Qin, X., 2013. Cost-effectiveness analysis of water-
706 saving irrigation technologies based on climate change response: A case study of China.
707 Agricultural Water Management 129, 9–20. https://doi.org/10.1016/j.agwat.2013.07.004

708 Zou, X., Li, Y., Li, K., Cremades, R., Gao, Q., Wan, Y., Qin, X., 2015. Greenhouse gas emissions from
709 agricultural irrigation in China. Mitig Adapt Strateg Glob Change 20, 295–315.
710 https://doi.org/10.1007/s11027-013-9492-9

711

27
Conflict of Interest

Declaration of interests

The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.
Author Statement

AUTHORS CREDIT STATEMENT

Muhammad Muzammil –Conceptualization, Methodology, Model development, data analysis, writing -

original draft

Azlan Zahid- reviewing and editing

Lutz Breuer- Conceptualization, Supervision, data analysis, reviewing and editing