Understanding Unsustainable Irrigation Practices in a Regionally Contested Large River Basin in Peninsular India Through the Lens of the Water–Energy–Food–Environment (WEFE) Nexus

Abstract

Water management is a long-standing source of dispute between the riparian states of Karnataka and Tamil Nadu. Recently, these disputes have intensified due to impacts from climate change and Bangalore’s rapid growth to megacity status. Despite well-defined national water governance instruments, competition between state actors and limited access to reliable hydrometric data have led to a fragmented regulatory regime, allowing unchecked exploitation of surface and groundwater resources. Meanwhile, subsidised energy for groundwater pumping incentivises the unsustainable irrigation of high-value, water-intensive crops, resulting in overextraction and harm to aquatic ecosystems. Here, we employ a water–energy–food–environment (WEFE) nexus approach to examine the socio-political, economic, and environmental factors driving unsustainable irrigation practices in the Cauvery River Basin (CRB) of Southern India. Our methodology integrates spatially explicit analysis using digitised irrigation census data, theoretical energy modelling, and crop water demand simulations to assess groundwater use patterns and energy consumption for irrigation and their links with governance and economic growth. We analyse spatio-temporal irrigation patterns across the whole basin (about 85,000 km²) and reveal the correlation between energy access and groundwater extraction. Our study highlights four key findings. First, groundwater pumping during the Rabi (short-rain) season consumes 24 times more energy than during the Kharif (long-rain) season, despite irrigating 40% less land. Second, the increasing depth of borewells, driven by falling water table levels, is a major factor in rising energy consumption. Third, energy input is highest in regions dominated by paddy cultivation. Fourth, water pumping in the Cauvery region accounts for about 16% of India’s agricultural energy use, despite covering only 4% of the country’s net irrigated area. Our study reinforces the existing literature advocating for holistic, catchment-wide planning, aligned with all UN Sustainable Development Goals.

1. Introduction

The Cauvery River Basin (CRB), a vital food-producing region in semi-arid India and a growing industrial and urban hub, exemplifies the unsustainable Water–Energy–Food (WEF) nexus. Unchecked groundwater rights for landowners [1] and intensive energy use strain limited-capacity aquifers [2], leading to declining water tables, ecosystem damage [3,4], and escalating socio-economic challenges, including increased water abstraction costs, conflicts, and heightened climate vulnerability [5,6]. Although groundwater pumping fuelled agricultural gains during the Green Revolution, it also accelerated resource depletion and carbon emissions, creating a cycle that now threatens farming livelihoods and national food security [2,7]. Addressing this crisis requires a deep understanding of the complex interconnections within the system.

As we are facing a climate emergency, effective resource management demands anticipatory governance to navigate complex trade-offs and uncertainties [8,9,10]. The WEF nexus framework has emerged as a powerful analytical tool for addressing these challenges [11,12,13], integrating contextual and quantitative information to reveal interconnections and feedback loops within complex systems [14,15]. This approach helps identify key leverage points for intervention. Our study applies the WEF nexus to understand the drivers and water–energy trade-offs in the CRB’s irrigation sector.

Recently, two studies applied the nexus approach to understand the dependency of cropping patterns on energy-intensive irrigation. While these studies [16,17] have generated valuable field-based estimates of water and energy use in agriculture through surveys, their methodologies do not offer insights on the spatial relationships between the resource use patterns. Our study offers a scalable and spatially explicit framework by leveraging national-level administrative and geospatial datasets. By using mixed methods i.e., driver assessment, estimation of irrigation energy use and associated theoretical water productivity and simulating crop water needs, we provide a more complete picture of irrigation inefficiencies and the water–energy–food–environment trade-offs across an entire basin. The highly localised focus of the recent nexus studies and absence of integrated metrics such as comparisons between theoretical and actual pumping requirements, make it difficult to connect resource-intensive irrigation practices to broader questions of policy, governance, and farmers’ decision-making logic.

Since the last century, government interventions and investments have supported agricultural growth in the region through surface water harvesting, groundwater irrigation, and electricity subsidies for agricultural pumping [2,10]. Although these measures aimed to boost food productivity, they also fuelled conflicts over competing water resource claims [18,19,20], profoundly shaping the region’s agro-economy. Studies analysing the CRB through geopolitical and governance lenses link water and energy overexploitation to resource conflicts [19,21,22], uncoordinated irrigation planning [23,24], and energy subsidies [2,25], all of which undermine long-term agricultural productivity. Meanwhile, hydrological and climate studies highlight the ecological consequences of excessive groundwater pumping for irrigation [6,26,27,28,29,30].

While groundwater irrigation demand is a well-recognised challenge in CRB water resource management, studies often overlook how regional socio-economic factors shape irrigation practices and intensify energy use [31,32]. Understanding these local variations is crucial for sustainable solutions. To address this gap, we use a spatially explicit approach to visualise variations in groundwater structures and energy consumption, identifying resource-intensive hotspots. By integrating village-level and basin-level data, we assess the impacts of both small-scale (farm-level irrigation practices) and large-scale (basin command areas) interventions on the WEFE nexus. Specifically, we aim to carry out the following:

• Map the factors influencing groundwater and energy use in irrigated agriculture.
• Quantify the spatial distribution of the energy consumed for groundwater pumping.
• Estimate the environmental impact of irrigation-related energy use, using carbon emissions as an indicator.

2. Study Area

2.1. Cauvery River Basin

Originating in the Brahmagiri hills of the Western Ghats, the Cauvery River flows eastward for approximately 800 km before draining into the Bay of Bengal (Figure 1). Its extensive 85,626 km² basin spans Karnataka (40.9%), Tamil Nadu (55.48%), Puducherry (0.18%), and Kerala (3.44%), exhibiting significant variations in precipitation, geology, water availability, and agro-economy [33]. The basin is over 66% agricultural land, with 5.8 million ha of cultivable land, which is 3% of India’s total cultivable land [34]. Its agricultural output, particularly rice and millets, plays a vital role in India’s food production [35]. While India has three main cropping seasons (Rabi, Kharif, and summer), cropping patterns in the CRB vary by state due to differing agro-climatic conditions and traditional practices.

2.2. Water Resources Development

The CRB relies on diverse irrigation infrastructure, including river-fed canals, multi-purpose dams, traditional rain-fed tanks, and groundwater systems like dug wells and tube wells. Surface-lift systems pump water from downstream surface sources (tanks, rivers, and canals) to areas where gravity flow is not feasible. The basin contains 45 major command areas (each exceeding 10,000 ha of cultivable land) and an equal number of medium and minor command areas (2000–10,000 ha), collectively irrigating 2.4 million ha (28% of the basin) across 25 districts [33] (Figure 1). Major dams in Karnataka and Tamil Nadu, such as Krishnaraj Sagar and the Mettur Dam, play a vital role in inter-basin water-sharing, as their discharge influences crop cultivation timing in Tamil Nadu. Multiple state agencies handle water allocation, while the Cauvery Water Management Authority (CWMA), established after the 2018 Cauvery Water Disputes Tribunal (CWDT) ruling, oversees inter-state water-sharing [37].

2.3. Geopolitical Context

Decades of conflict over Cauvery River water stem from historical, political, and socio-economic factors [10,38]. Political divisions hinder national–state coordination, while fragmented state-level water administration and information sharing obstruct effective management [10,22]. These divisions have led to fluctuating water allocations, culminating in the CWDT awards of 2007 and 2018. The dispute centres on Tamil Nadu’s reliance on surface water for delta paddy cultivation and Karnataka’s demand for equitable upstream irrigation rights, particularly amid monsoonal variability and growing demand from urban centres like Bangalore. While the tribunal rulings sought to balance irrigation needs, hydrological realities, and evolving demands, they have been criticised for overlooking groundwater depletion, ecosystem impacts, and inefficiencies in water distribution [19,24]. Consequently, inconsistent enforcement of water releases and decision-making delays have disrupted cropping patterns, increased groundwater reliance, intensified basin tensions, and left tail-end farmers in Tamil Nadu with unreliable supplies [22].

3. Methods and Data

3.1. Methodology

Using a spatially explicit approach, our study followed five steps (Figure 2). First, we digitised Minor Irrigation Census Data (MICD) in ArcGIS version 10.6 to map irrigation regimes using a grid cell of 0.125° × 0.125° (192.5 km²), capturing energy sources, borewell depths, seasonal groundwater levels, and urban/industrial water demand hotspots. Second, we estimated theoretical energy consumption for groundwater pumping using MICD data on pump efficiency, pumping hours, and pump power. Third, we calculated theoretical water productivity under conditions of unconstrained water availability. Fourth, we compared the theoretical water volume with crop water requirements simulated by the GWAVA crop model [28] to assess the gap between energy needs and actual consumption. Finally, we estimated CO₂ emissions associated with groundwater pumping.

3.2. Data

The core analysis including the estimation of irrigation energy use and the spatial variability of pumpset distribution focuses on the year 2013–14, which corresponds to the most recent year for which detailed Minor Irrigation Census data were available. However, to provide important contextual insights such as average seasonal groundwater depths and the long-term correlation between groundwater availability and crop production, we analysed extended time series data from 1996 to 2016. A summary of the data used is provided in

Table 1. Data used in this chapter and data sources.

Data	Spatial scale	Period	Used for	Source
Number of pumpsets under different irrigation sources	Village scale	2013–2014	Spatial mapping of irrigation systems and estimating energy consumed	[36]
Rice production	District scale	1997–2014	Estimating correlation between rice production and groundwater availability	[39]
Groundwater table	CGWB borewells	1997–2014	Estimating correlation between rice production and groundwater availability, estimating groundwater abstraction	[33]
Hours of pumping and pump horsepower	Village scale	2013–2014	Estimating energy consumed	[36]
Irrigation potential under different irrigation sources and water distribution methods	Village scale	2013–2014	Estimating irrigation efficiency	[36]
Pumpsets under different sources of energy and depth of borewells	Village scale	2013–2014	Spatial mapping of irrigation systems	[36]
Pumpset efficiency of 30%	-	-	Estimating groundwater abstraction	[3]
Irrigation source per farmers’ economic class	District scale	2014	Assessing economic factor in access to irrigation equipment	[39]

.

i

Farmer status data—We classified district-wide pumpset ownership data by farmer type according to their economic status in the agriculture census [39]. These data were aggregated for the whole basin and used to analyse the relationships between irrigation sources, access to groundwater pumping equipment, farmers’ economic status, and farm size.

ii

Groundwater table data—We obtained groundwater table data from the Central Groundwater Board (CGWB) portal (India-WRIS) [33] for all observation wells in basin districts from 1996 to 2016. Since the number of wells increased annually and seasonal data gaps were common, we calculated district-level averages to analyse 20-year groundwater trends. To approximate missing values, we used observations from the previous or following year for the same season.

iii

Rice Production data—We obtained district-scale rice production data from the National Agriculture Statistics archives [39] for the years 1997–2014 to estimate the correlation between rice production and groundwater availability.

iv

Minor irrigation data—We obtained village-wise groundwater irrigation data for the year 2013–14 from the Minor Irrigation (MI) Census database, available through the Indian Statistical Data Dissemination Portal. This dataset included information on groundwater irrigation structures such as deep, medium and, shallow tube wells, dug wells, and minor surface water irrigation systems including surface-lift pumps and surface flow schemes. For each MI structure type, the dataset provided the number of structures, their operational status, average daily pumping hours during Rabi and Kharif seasons, the distribution of structures by pump horsepower, energy sources, and water distribution system (Table 1). The data were manually resampled for villages within the basin boundary and converted to geospatial format by joining tabular data with village shapefiles in ArcGIS 10.6 software.

v

Major and medium irrigation projects—We obtained data on the area of major and medium irrigation projects from India-WRIS web-map services [33]. To analyse the correlation between groundwater pumping and the availability of surface irrigation sources, we manually digitised the command areas of these projects in ArcGIS software.

3.3. Estimating Energy Consumption and Groundwater Productivity of Pumps

The irrigation energy consumption can be estimated using the following equation [40]:

$$\begin{array}{cc}\hfill \mathrm{Energy}\left[\mathrm{kWh}\right]=& 0.7457\times \mathrm{Average}\mathrm{pump}\mathrm{horsepower}\left(HP\right)\hfill \\ \hfill & \times \mathrm{Daily}\mathrm{average}\mathrm{pumping}\mathrm{hours}\left(PH\right)\hfill \\ \hfill & \times \mathrm{Number}\mathrm{of}\mathrm{pumps}\mathrm{in}\mathrm{use}\left(N\right)\hfill \end{array}$$

(1)

Alternatively, the energy consumption can be estimated from the physical principles governing pump power, groundwater hydraulics, pump efficiency, and discharge [8,41,42,43,44,45]:

$$\begin{array}{cc}\hfill \mathrm{Energy}\left[\mathrm{kWh}\right]& ={\displaystyle \frac{9.8{\mathrm{ms}}^{-2}\times \mathrm{Lift}\left[\mathrm{m}\right]\times 1000{\mathrm{kgm}}^{-3}\times Q\left[{\mathrm{m}}^3\right]}{3.6\times {10}^6\times PE[\%]}}\hfill \\ \hfill & ={\displaystyle \frac{2.73\times \mathrm{Lift}\times Q}{1000\times PE}}\hfill \end{array}$$

(2)

where Energy [kWh] is the electricity drawn from the grid, Lift [m] is the vertical distance from the water table to the pump, Q [m³] is the volume of water extracted, and $PE$ [%] is the pump efficiency. Due to limited public data, we adopted the widely used estimate of 30% for electric pumpset efficiency in India [3].

We used Equation (2) to estimate the theoretical volume of groundwater Q that can be abstracted based on the estimated energy consumption from Equation (1), assuming no geological constraints. Estimating the groundwater depth accurately is essential for this approach, because lift head directly affects energy input (a 1 m decline in the water table increases energy input by 0.0091 kWh per m³ of water lifted). Our calculations assumed a stable intra-seasonal groundwater table but allowed for inter-seasonal variability. We extracted the groundwater table values for each village during Kharif and Rabi seasons by interpolating the groundwater table contour maps generated in ArcGIS using Central Ground Water Board (CGWB) well data.

4. Results

4.1. Spatial Mapping of Irrigation Regimes

Figure 1 gives an overview of the density of active groundwater pumps, state boundaries, the major irrigation project command areas, key dams, and data gaps.

4.1.1. Density of Minor Irrigation Pumpsets

As of 2014, CRB housed approximately 1.4 million groundwater pumpsets, accounting for about 3% of the global total [46]. Spatial maps of active groundwater sources reveal significant variation in pump density, ranging from 1 to over 50 pumps per km² (Figure 3). Among 14,260 villages in the CRB, the majority (4915) have fewer than 10 pumps/km², while 3069 have 10–20 pumps/km². However, a small fraction of villages (113) show extremely high densities, exceeding 100 pumps/km², indicating localised hotspots of intense pump activity.

Pump density differs based on water availability and land use. Tamil Nadu has a higher average pump density than Karnataka (26 vs. 15 pumps/km²), and a higher standard deviation (53.7 vs. 21.6), which we attribute to its larger village areas, averaging 8.9 km² compared to 3.5 km² in Karnataka. Within irrigation command areas, Tamil Nadu has a mean density of 23 pumps/km² and Karnataka has 16 pumps/km². Outside command areas, pump density rises to 29 pumps/km² in Tamil Nadu, driven by unreliable upstream water discharge and lower surface water productivity, whereas Karnataka’s decreases to 14 pumps/km².

Groundwater reliance is particularly pronounced in the Cauvery Delta region, where pump density averages 23 pumps/km², indicating its critical role in paddy irrigation amid surface water scarcity (Figure 3). In contrast, pump density remains low in high-altitude regions like the Mahadeswaramalai and Biligirirangan Ranges, and across the Western Ghats, although data coverage is limited, but the recorded pump density is generally low.

By 2013–14, there were 660,714 active groundwater pumps, with 45.8% in Karnataka, 43.4% in Tamil Nadu (excluding the delta), and 10.8% in the Cauvery Delta (Figure 3). Additionally, 5259 pumps were used for surface water lifting under MI schemes. The data highlight Tamil Nadu’s intensive groundwater, particularly in the delta region, while surface-lift pumps are more common than in Karnataka’s northern sub-catchments.

4.1.2. Seasonality of Pumpset Use

In 2013–14, groundwater pumps irrigated 1.16 million hectares, accounting for 93% of the total irrigated area in the CRB. Surface water sources, including lift and flow schemes, covered only 88,000 hectares (7%), mostly in the upper catchments of Karnataka’s Chikmagalur district. Seasonally, groundwater pump use peaked in the Kharif season, irrigating 47% of the total groundwater-irrigated area, followed by Rabi (29.3%) and summer (5.1%). This trend aligns with paddy cultivation, particularly in Karnataka and Tamil Nadu [47].

In Karnataka, 89.6% of the 284,000 hectares irrigated during Kharif relied on groundwater. In Rabi season, the irrigated area dropped by 57% to 122,000 hectares, with 93.5% of this area irrigated with groundwater. The decline in irrigated area during Rabi is linked to cooler temperatures, a shift to less water-intensive crops like wheat, and soil moisture retention from the monsoon [48]. In Tamil Nadu, the irrigation patterns during Kharif are similar to those in Karnataka, but the irrigated area increases in the Rabi season, particularly in the Cauvery Delta, where year-round paddy cultivation persists as the saline soils are unsuitable for other crops.

Groundwater plays a crucial role in the delta, where it irrigates nearly all cultivated land. Delays in Mettur Dam’s water release for Kuruvai crops often force heavy groundwater use during Kharif. Additionally, 18.3% of perennial crops area, such as areas with dryland fruits in higher altitudes, depends on groundwater.

Spatially, groundwater reliance intensifies from the middle catchment to the delta, particularly where canal water is insufficient. Groundwater use is lower in canal-irrigated zones in central Karnataka and the delta but rises sharply outside these zones. In northeastern and northwestern Karnataka, over 50% of irrigated land during Kharif relies on groundwater pumps, highlighting the importance of timely and adequate irrigation projects.

4.2. Drivers Influencing Water–Energy Dynamics in Irrigation Sector

4.2.1. Sources of Energy for Irrigation Pumps

In 1993–94, irrigation pumps in the CRB were powered by electricity (72%), diesel (15%), or animal labour (11%) [36]. By 2014, the basin had around 700,000 groundwater pumpsets and 5602 surface-lift pumps, together accounting for about 3% India’s total [46]. The majority of these pumps were electrified (89%), followed by mixed-energy pumps (7%), diesel pumps (3%), wind (0.03%), manual/animal (0.03%) and solar (0.01%). The negligible adoption of solar pumps highlights the challenges in scaling up this irrigation technology.

Spatial mapping confirms that electricity is the dominant pump energy source in the Cauvery region, indicating nearly all agriculture villages are connected to the grid. Diesel pumps are concentrated in the delta region and areas reliant on dug wells, such as parts of Nagapattinum, Thriruvarur, and Chikmaglur. Wind- and solar-powered pumps are scattered across the basin with no significant spatial clustering (Figure 4).

Expanding the use of solar pumps from 0.01% to 1% would require installing approximately 7000 solar pumps, each linked to a 20 kW photovoltaic panel system, adding up to 70 MW of solar capacity to the grid. However, widespread adoption faces significant barriers, including subsidised electricity, inadequate groundwater regulations, and high upfront and maintenance costs. While solar pumping can reduce carbon emissions and fossil fuels dependency, it may accelerate groundwater depletion if not integrated with water-saving technologies like micro-irrigation. The shift to solar-powered irrigation pumps must be accompanied by policies that ensure efficient water use, preventing the risk of groundwater overextraction as energy becomes more accessible and affordable. Integrating energy and water management strategies—such as Morocco’s approach, which links solar subsidies with micro-irrigation—could provide a model for balancing the energy–water nexus [49].

A major challenge for irrigation is the unreliable electricity supply. In 2013, 10% of wells were inactive due to power shortages. Pumping patterns have remained spatially consistent during the last 20 years, but average pump use has declined uniformly by 2–5 h a day since 1993 (see district-level data in Figure 5). This reduction may stem from the shift from diesel to electric pumps, as power companies use outages to manage electricity consumption, counterproductive to energy subsidies. Our field visits found that unannounced power cuts often force farmers to leave their pumps running continuously to secure water for irrigation, leading to an overconsumption of electricity. Additionally, deepening borewells due to declining water tables require more energy for pumping, further straining electricity supply [50].

4.2.2. Deepening Borewells and Groundwater Level

Groundwater ownership in India is tied to land ownership under the Indian Easement Act of 1882, granting landowners rights to access groundwater [1,51,52]. Across the CRB, farmers use a variety of groundwater structures of varying depth. We analysed the spatial variability of these structures to explore borewell depth and its correlation with external factors. The spatial distribution of the average groundwater structure depth is shown in Figure 6, and its relationship to groundwater table depths from CGWB is mapped in Figure 7. Figure 8 shows the proportion of gross irrigated area that uses tube wells of depth $<35$ m, 35 to 70 m, 80 to 150 m; dug wells; surface-lift schemes; and surface flow schemes.

According to CGWB data (Figure 7), groundwater tables are deeper (up to 64 m) in the upper catchments, yet the average well depth remains relatively shallow (Figure 6). The use of shallow wells in the upper catchments is likely influenced by the major and medium irrigation (MMI) command area, which should result in augmentation of groundwater. We noticed a similar discrepancy in the Dharmapuri district (Tamil Nadu) near the Mahadeswaramalai–Biligirirangan ranges along the Karnataka–Tamil Nadu border, where the groundwater table is shallow (up to 15 m), yet irrigation wells extend to depths of around 100 m. These discrepancies highlight the variability and uncertainty of groundwater availability, suggesting that farmers’ decisions to deepen wells are influenced by factors including economic ability and access to reliable surface water like canals. The deepest wells are prominently located outside the MMI command areas.

The spatial distribution of irrigation structures shows distinct patterns across the basin (Figure 8). Deep tubewells are widespread, but are particularly concentrated in parts of Chamarajnagar (Karnataka) and Dharmapuri (Tamil Nadu) near the Mahadeswaramalai–Biligirirangan ranges, as well as the Thanjavur and Ariyalur districts in the delta. Medium tubewells, machine-drilled borewells 35 to 70 m deep, are mainly used in Karnataka and in the command areas of the delta. Dug wells, hand-dug wells typically less than 30 m deep, predate the invention of drilled boreholes but remain common in Tamil Nadu due to lithological challenges that limit deeper drilling. Nonetheless, farmers are investing regularly in deepening these wells (as noted during our field trips). Shallow tubewells are mostly used in the upper catchments and the delta region.

Groundwater remains the most reliable source of irrigation in the CRB. Despite the expansion of tube well irrigation, reports of well failures and groundwater depletion have risen since the 2000s, often linked to the low storage capacity of the region’s hard rock aquifers [53]. By 2013, poor water yields had rendered 71% of total wells inactive, contributing to widespread abandonment of dug wells and deepening of borewells (Figure 9).

4.2.3. Competition for Water and Crop Production

Urban expansion in cities like Bangalore, coupled with the growth of industrial clusters and agricultural demands, has intensified competition for water resources (Figure 10). This competition is a major driver of the Cauvery conflict, exacerbating the WEF nexus challenges [54,55]. Overabstraction of groundwater in economic development zones has disrupted catchment water balance, contributing to well failure, reduced crop production, and declining farmer livelihoods downstream [7,28,53].

In groundwater-irrigated districts like Bangalore Rural and Bangalore Urban, rice production during the Rabi season is notably low. In Bangalore Rural, Rabi rice cultivation declined from over 200 ha in 2004 to below 20 ha in 2014, while Kharif rice production fell from 14,828 ha in 2000 to 598 ha. Similar trends are seen in Bangalore Urban, where Kharif rice reduced from 4,483 ha in 2000 to 950 ha, with a Rabi crop appearing only in 2012 and 2013. These declines reflect increasing water demand for urban and industrial growth, as well as shifts in employment preferences.

Across much of the basin, Rabi rice has been replaced by less water-intensive crops, like millets and non-food crops [39]. In Tumkur, a key groundwater-irrigated district and the basin’s largest rice producer, we found that while the Kharif and summer production remained stable, Rabi rice production has declined. In fully canal-irrigated districts, rice cultivation is confined to the Kharif season, aligning with scheduled dams releases in late June. In Mysuru and Mandya, where both groundwater and canal irrigation are common, the Kharif and summer crops are more productive than the Rabi crop, with no clear link to groundwater levels. In Pudukkottai, an arid district where tank irrigation is prevalent, rice cultivation is similarly limited to the Kharif season.

Access to reliable irrigation sources strongly influences cropping patterns across the Cauvery region. We observed a strong linear relationship between groundwater levels and rice production in Bangalore Rural, and a weaker but similar trend in Bangalore Urban, where rice production has gradually declined since 2001 due to socio-economic shifts driven by industrialisation. In Tumkur, Mandya, and Mysore, where surface water is the primary irrigation source, no significant relationship was found. In Erode, where surface and groundwater-irrigated areas are comparable, rice production shows a weak negative relationship with groundwater levels. In the delta, Pudukkottai relies on tanks and groundwater pumps for irrigation, whereas Thiruvarur falls under the delta irrigation project command area. Both districts exhibit a strong correlation between rice production and groundwater levels, reflecting their reliance on groundwater in a semi-arid climate with less than 1000 mm annual rainfall.

4.3. Spatial Analysis of Estimated Energy Consumption

In this section, we use energy consumption data to map groundwater withdrawal using Equation (2). We then compare our estimated withdrawal with actual crop water demand. If the estimated water withdrawal exceeds demand, it indicates a local overconsumption of energy.

4.3.1. Verification of Estimated Energy Consumption

To validate our energy consumption estimates, we performed a linear regression analysis using district-scale data for energy consumption data and groundwater draft data from the CGWB. Our analysis showed a strong positive correlation (p-value < 0.05, $r^2=0.74$) between estimated energy consumption (MWh) and groundwater draft (MCM) (Figure 11), verifying the reliability of census data in capturing irrigation-related energy demand. Uncertainties may stem from standardising pumping hours and horsepower values, as we used village-level averages.

4.3.2. Energy Consumption in Irrigated Agriculture

To analyse spatial patterns of agricultural energy use, we mapped the average daily energy consumed by groundwater pumps and surface-lift schemes during peak Kharif and Rabi seasons. We found that the spatial distribution of both groundwater and surface water energy consumption remained broadly consistent across seasons (Figure 12). Interestingly, the upper catchments of Karnataka’s Hassan district accounted for a significant proportion of surface water energy consumption, likely due to the region’s higher altitude and the need to transport water to elevated agricultural areas.

However, a contrast emerged in the overall groundwater energy consumption. Although the spatial distribution was similar, Rabi consumption reached 349.9 GWh per day, an approximately 2400% increase compared to Kharif’s 9.6 GWh per day. This substantial rise in Rabi energy consumption occurs despite a 40% reduction in irrigated land and significantly lower pumping hours, as confirmed by a two-sample t-test (

Table 2. Average daily pumping hours of pumpsets associated with different irrigation sources during peak Kharif and Rabi seasons within CRB (2013–14).

	Pumping Hours/Day (Kharif)			Pumping Hours/Day (Rabi)
G.W Source	Mean	Standard Error 1	Maximum	Mean	Standard Error 1	Maximum	p Value
Deep tubewells	4	±0.68	22	3.6	±0.47	22	≪0.05
Medium tubewells	4.4	±0.69	22	3.7	±0.5	22	≪0.05
Shallow tubewells	3.9	±0.46	22	3.4	±0.33	22	≪0.05
Dug wells	4	±0.4	22	3.5	±0.35	14	≪0.05
Surface-lift schemes	3	±0.48	22	2.7	±0.35	22	≪0.05

Note: ¹ 95% confidence interval.

). The key driver is the significant decline in the groundwater table during Rabi, which increases the energy required per pump (Equation (2)), ultimately leading to greater overall energy consumption. Energy consumption by pumpsets is highly sensitive to the depth of the groundwater table and the energy efficiency of the pumpsets. Our estimates indicate that a one-meter drop in the groundwater table can increase energy cost for lifting 1000 m³ of groundwater by $15.36\pm 0.17\%$ during Kharif and by $14.07\pm 0.21\%$ during Rabi. Conversely, a 10% increase in pump efficiency could reduce energy demand by 25% during both seasons.

Contrary to expectations, pumpset use remained consistent inside and outside MMI command areas, despite their additional surface water supply. Seemingly, groundwater availability, pump access, and farmers’ economic status play a larger role than surface water supply. This pattern aligns with broader trends in India, where agricultural energy consumption does not directly follow monsoon patterns [54]. In the CRB, surface water availability varies significantly between upstream and downstream regions. Upstream in Karnataka, farmers prefer groundwater pumping despite stable irrigation systems and high annual rainfall, likely due to its convenience and reliability. Conversely, Tamil Nadu’s irrigation is heavily influenced by uncertain upstream reservoir discharges, driving farmers to rely more on groundwater pumps for traditional rice crops.

We also found that energy consumption increased as annual rainfall fell. During Kharif, the median daily energy consumption was 213.7 kWh/ha in the highest rainfall zone (3000–2000 mm), 745.7 kWh/ha in the moderate zone (2000–1000 mm), and 914.4 kWh/ha in the lowest zone (<1000 mm). During Rabi, these medians increased to 2237 kWh/ha, 6271 kWh/ha, and 2117 kWh/ha, respectively.

4.3.3. Estimating Groundwater Withdrawal

Using Equation (2) and CGWB ground level data, we estimated the groundwater withdrawal during peak Kharif and Rabi seasons (Figure 13). These calculations assumed unlimited groundwater availability and unrestricted tubewell yield, providing a theoretical maximum. Under these conditions, the estimated average daily water abstraction was 12,775 m³/ha during Kharif and 311,753 m³/ha during Rabi.

The estimated 2400% increase in Rabi groundwater pumping energy, despite a 40% reduction in irrigated land highlights the complex and interlinked drivers within the WEF nexus, with broader implications for national and state resource security. The Rabi energy surge creates power deficits in non-agricultural sectors, resulting in power cuts that further disrupt farmers’ operations (e.g., pumpset failure). This situation raises a critical question: do farmers truly require free electricity?

To address this question, we compared our estimated groundwater withdrawals with irrigation water demand simulated by the Global Water Availability (GWAVA) model, a semi-distributed hydrological model that estimates irrigation demand by considering irrigation efficiency and crop water requirements based on physical variables such as precipitation, temperature, land use, and soil characteristics [56]. We calculated the total irrigation water demand for the basin, excluding the delta, setting the irrigation efficiency parameter at 44.11%, the India average estimated by FAO. Remarkably, we found that crop water demand is less than 1% of the theoretical water withdrawal shown in Figure 13, indicating significant overconsumption of energy for groundwater pumping. This discrepancy can be attributed to factors such as poor pump efficiency, declining groundwater tables, inefficient irrigation practices, inadequate regulatory oversight, and the absence of comprehensive water and energy accounting systems.

The spatial overlay of urban hubs, industrial zones, and high pumping intensity areas like Bangalore Rural and Urban underscores the growing pressure on local aquifers. These trends reflect a shift in surface water allocation driven by socio-economic priorities rather than agricultural suitability, leading to reduced resilience in food production. The spatial heterogeneity observed also illustrates how water competition manifests differently across the basin suggesting the need for geographically tailored but coordinated policies.

5. Discussion

5.1. The Nexus in Irrigated Agriculture

Within the WEFE framework, the intricate interdependencies between water use, energy for extraction and distribution, and agricultural productivity create complex feedback loops that influence resource availability and sustainability [13,16,17,57,58,59,60,61] (Figure 14). Insights generated in this paper contribute to the understanding that the WEFE nexus is not just a technical challenge, but a political and institutional one rooted in competing land and water use regimes [62,63,64,65].

In the CRB, the nexus is driven by factors such as groundwater regulations and free electricity policies, which have incentivised deeper borewell drilling and accelerated groundwater depletion [25,55]. Despite this depletion, irrigation pumping hours have remained consistent, indicating a continued reliance on groundwater. Yet, deeper groundwater extraction has not improved agricultural productivity, as rice production has declined, potentially due to competing industrial water demands amongst other bio-physical and socio-economic factors.

Policies intended to ease the energy burden, such as promoting solar energy pumps and drip irrigation systems, have seen limited adoption, likely due to economic, cultural, or operational barriers surrounding new technology [66,67]. These interdependencies highlight how changes in one sector, such as energy subsidies, can cascade across the nexus, amplifying resource stresses [12]. Addressing these challenges requires integrated policy interventions that align incentives across water, energy, and agriculture.

5.2. Challenges and Uncertainties in the Nexus

The CRB faces significant challenges related to both surface and subsurface water depletion, leading to inter-sectoral competition over scarce water resources [68,69]. These challenges are exacerbated by trans-boundary water-sharing disputes and spatial variations in rainfall. Within irrigated agriculture, surface water scarcity and poor management have increased reliance on groundwater pumping. A major obstacle is the lack of reliable, spatially resolved data on groundwater and energy consumption. Inconsistent and fragmented government data hinders comprehensive nexus assessments, making it difficult to quantify trade-offs, identify resource-intensive systems, and design region-specific management strategies [70].

Our study represents an effort to quantify the nexus within India’s agricultural sector by analysing socio-economic factors and policy drivers. Our findings reveal significant spatial variability in irrigation sources, with the delta region emerging as a hotspot for energy-intensive irrigation due to its year-round paddy cultivation and unreliable canal irrigation. The basin’s history of water conflicts, particularly the Cauvery water dispute, has compelled farmers to rely heavily on groundwater, especially for paddy cultivation [69]. This increased reliance has worsened inequality in water resource access, especially for farmers outside MMI command areas, who must deepen borewells up to 130 m to access groundwater, compared to 70 m required inside MMI command areas.

Historically, dam water releases are scheduled for June 12th to support the cultivation of Kuruvai (June–Sept) and Taladi (Oct–Jan) paddy crops. However, delays in these releases have compelled farmers below the Lower Anicut (major and ancient water control structure directing water through canals into the delta, Figure 1) to cultivate only the longer-duration, less productive, and less economically preferred Samba crop (Aug–Jan). Previous research revealed a strong correlation between the delay of dam releases and harvested crop area [71], confirmed by our stakeholder interactions. Consequently, a policy initially designed to enhance irrigation resources has paradoxically increased farmers’ reliance on groundwater even further.

Groundwater access poses a significant economic challenge, particularly for marginal farmers. Despite free electricity, marginal farmers face substantial costs related to pump purchase and maintenance and well deepening. Our analysis of pump ownership reveals a stark disparity in access to electrified pumpsets: only 41% of marginal farmers’ wells and 82% of their tubewells are electrified, compared to 83%, 86%, and 81% of wells, and 96%, 98%, and 98% of tubewells owned by semi-medium, medium, and large farmers. This economic disadvantage is further underscored by the 51% abandonment rate of wells owned by marginal farmers (see Supplementary Materials).

Although reliance on electrified groundwater pumps has grown substantially over the past two decades, their use has not improved food crop production. In Tumkur, for instance, annual Kharif production halved in the last decade, from an average of 42 million kg between 1997 and 2002 to 21 million kg between 2010 and 2014, with similar declines for Rabi (33 to 16 million kg) and summer (43 to 16 million kg) production. These statistics suggest that free electricity policies have failed to enhance food or water security and have instead complicated the challenges of the WEF security nexus. The financial burden of groundwater pumps, driven by poor management, unequal surface water distribution, and declining groundwater productivity, remains a primary cause of distress for farmers in the region [58,72].

The water–energy nexus in the CRB also has significant environmental impacts, particularly concerning greenhouse gas (GHG) emissions. The estimated daily energy consumption for water pumping during peak Kharif (9.65 GWh) and Rabi (345.9 GWh) results in substantial CO₂-eq emissions of 8 million kg in Kharif and 294 million kg in Rabi, based on an emission factor of 0.85 kg CO₂-eq per kWh [73]. Alarmingly, the Cauvery region accounts for 16% of India’s agricultural energy consumption, despite representing only 4% of the country’s net irrigated area. This disproportionate energy use underscores the urgent need to reassess free electricity policies and strengthen groundwater governance. While efforts have been made to mitigate these challenges, including the introduction of solar pumps and water-efficient irrigation technologies, such as drip and sprinkler systems, the limited reach and low adoption rate of these resource-efficient solutions continue to pose significant obstacles.

The major water withdrawals as comprehensively explained in this article have made far-reaching alterations—not only to the water quantity in the river channels but also to other essential river flow and quality dynamics that are essential to support the ecosystem services. These changes have disrupted ecological regimes, degraded habitats, and affected migratory species due to altered flow from large and check dams. Inadequate consideration of biodiversity in water management has led to depleted flows, particularly in hyporheic zones, impacting both aquatic and riparian ecosystems. Additional pressures from the agriculture sector such as discharge of agrochemicals and pesticides have further contributed to the decline in fish stocks and overall biodiversity [74].

5.3. Limitations of This Study

While our spatially explicit nexus approach generates basin-wide insights into irrigation energy use and its underlying drivers and consequences, there are a few limitations. First, the theoretical estimates rely on assumptions around pump efficiency and pump yield; consequently, the results do not capture the variability in equipment performance. Second, assuming stable groundwater tables and uniform pump operations within cropping seasons may oversimplify the temporal dynamics of groundwater use, introducing uncertainties into energy and abstraction estimates. Third, extrapolating groundwater table data using observation well data underestimates the effect of ‘cone-of-depression’ on groundwater level in irrigation pumps. Fourth, although the influence of industrial and urban growth is clear from our analysis, lack of high-resolution sector-specific groundwater withdrawal data limits the deeper analysis of these dynamics. Finally, the lack of primary data on farmers’ decision-making processes limits the ability to contextualise irrigation patterns as adaptive responses to socio-economic or institutional triggers. These limitations present important opportunities for future research that can integrate high-resolution behavioural, institutional, and bio-physical data.

6. Conclusions

The CRB exemplifies the intricate and often conflicting dynamics between economic growth, water governance, resource sustainability, and food productivity. The unsustainable practices of free electricity provision, poor surface water management, and the escalating water demands of urban and industrial centres like Bangalore have driven energy inefficiency and groundwater depletion, exacerbating socio-economic distress. Although the immediate removal of subsidies is impractical since millions of farmers depend on them, there is an urgent need to transition from traditional flood irrigation to more efficient methods like drip irrigation and expand on-farm rainwater storage. Integrated policies that provide financial and technical support to farmers could create widespread benefits, improving agricultural resilience while alleviating pressure on power utilities, state resources, and the environment.

Addressing these challenges requires a multifaceted approach. At the village scale, practices like aligning irrigation schedules with electricity supply, investing in pipeline infrastructure to reduce water loss, and developing strategic water allocation systems are essential for optimising crop yields and generating revenue from water resources. With climate change expected to increased intense rainfall events, leading to greater run-off and further strain on water resources, there is a critical need for region-specific resource management strategies. Robust monitoring of groundwater abstractions and energy consumption is crucial for understanding the groundwater–energy nexus and developing predictive models for future demand. Furthermore, metering groundwater abstraction can enhance the economic value of groundwater, curb over abstraction, and improve energy management.

Therefore, we propose a comprehensive strategy including the following:

• Accelerating the adoption of water-efficient irrigation methods by promoting drip irrigation and rainwater harvesting.
• Reforming electricity subsidies to reduce energy consumption while maintaining support for farmers.
• Strengthening water governance by improving inter-state water-sharing agreements and promoting integrated basin-level management.
• Investing in data-driven decision-making by building a robust data.
• Empowering farmers by providing financial and technical support for adopting sustainable agricultural practices.

By implementing these strategies, the CRB can move towards a more sustainable and resilient future. Continued research, monitoring, and adaptive management are essential to ensure the long-term effectiveness of these interventions in the face of evolving environmental and socio-economic challenges.