Extractable Water Index (EWI): Towards a Universal Metric for Sustainable River Extraction

Abstract

Sustainable river management depends on indices that balance human water demands with ecological flow requirements while accounting for hydrological variability. Existing water scarcity and withdrawal indices are largely based on monthly or annual aggregates, often neglecting daily variability and the effects of drought buffering. This study introduces the Extractable Water Index (EWI), a novel, dimensionless metric that quantifies the sustainable potential for water extraction using daily flow records. The EWI integrates mean available flow, flow variability, low-flow thresholds, and storage contributions into a single expression, thereby capturing both hydrological dynamics and ecological protections. Two scenarios were evaluated, (i) no-storage and (ii) with-storage, with the latter employing a semi-analytical approximation to represent a reservoir or pond. The EWI was applied to 20 daily river flow series for 16 river basins in Sri Lanka. Under no-storage conditions, thresholds were defined as follows: EWI < 0.45 indicates low extraction potential; 0.45 < EWI < 0.60 indicates moderate extraction potential; and EWI > 0.75 indicates high extraction potential. The results demonstrate that even modest storage can substantially enhance sustainable withdrawals. The EWI provides a transparent, reproducible decision-support tool that complements environmental flow standards and prioritizes rivers based on extractability. The EWI provides a valuable tool for estimating water extraction potential within the Sri Lankan context. This index can be applied across diverse hydroclimatic regimes and, when combined with threshold validation, can predict extraction requirements under varying seasonal flow conditions.

1. Introduction

Watercourses meet the immediate needs of communities and play a crucial role in maintaining environmental balance [1]. River ecosystems, in particular, serve as essential habitats that support a diverse array of species [2,3]. They provide vital resources for numerous aquatic organisms, including fish, amphibians, and various invertebrates, which rely on water flow, temperature, and quality for their survival [4]. Humans gain significant advantages from river ecosystems, using them for drinking water, recreation, and economic pursuits such as industrial water, fishing, and tourism [1]. Freshwater sources, particularly river flow, are lifelines for communities, fulfilling essential functions such as providing potable water for domestic use, supporting agricultural practices, generating hydroelectric power, and sustaining diverse ecological systems [5,6,7]. The intricate web of life within these ecosystems emphasizes their importance for biodiversity and the overall health of the planet.

Urban expansion often leads to the development of impervious surfaces, such as asphalt roads, concrete sidewalks, and buildings, which impede the natural infiltration of rainwater into the ground [8,9]. As urban areas grow, large areas that previously allowed rainwater to infiltrate the soil are replaced by non-porous materials [10]. This alteration of the landscape increases surface runoff, which can overwhelm drainage systems and natural waterways, potentially leading to flooding [11]. Consequently, base flow is reduced, causing perennial rivers to become intermittent and contributing to water scarcity during drier months [9].

The removal of forests further exacerbates this issue by disrupting the watershed’s capacity to retain moisture, thereby altering precipitation patterns and reducing groundwater recharge [12]. As a result, these factors contribute to greater fluctuations in river flow, jeopardizing the ecological health of aquatic systems and impacting water availability for nearby communities [13]. Therefore, careful management of water extraction for essential needs, such as drinking water, is crucial to meet environmental flow requirements. During periods of low water flow, it is imperative to thoroughly analyze and adapt various requirements, particularly environmental flow standards [14,15].

Furthermore, a comprehensive assessment of environmental flow requirements requires an examination of seasonal variations in water levels [10,16]. By integrating these factors into the water extraction evaluation, it becomes possible to more precisely determine sustainable and responsible rates of water withdrawal from a river or other water source. This approach helps balance human water needs with the need to conserve the natural environment, thereby fostering long-term ecological health and resilience.

Industrialization and increased domestic demand due to population growth have contributed to water scarcity, leading to the development of quantifiable water indicators in the 1980s [17]. Falkenmark’s indicator was the first scale designed to quantify water scarcity using readily available national data [18]. Critical ratio indexing was subsequently introduced, employing the ratio of annual water use to total water resource availability [19]. The International Water Management Institute (IWMI) developed the IWMI indicator, which is comprehensive and sensitive to macro-level adaptive capacity [20]. Additionally, the water poverty index [21] and the Sustainable Development Goal 6 indicator are considered classic indicators of water scarcity, each addressing specific aspects of water scarcity [17]. In contrast, indicators such as the System for Environmental-Economic Accounting for Water, Water Exploitation Index (WEI), Water Footprint and Virtual Water, Green-Blue Scarcity Indicator, Life Cycle Assessment-Based Indicator, Water Stress Index (WSI), Environment Flow Requirement Indicators, Reliability Resilience Vulnerability (RRV) framework and Quantitative Quality Indicator are regarded as some of the holistic indicators that encompass multiple dimensions of water scarcity [17,22,23,24,25,26,27]. The WEI is an annual, ratio-based metric that compares total water abstraction to long-term renewable water resources [23]. Consequently, WEI primarily reflects aggregate human pressure on water resources but does not account for storage effects, short-term variability, ecological thresholds, or supply reliability. The WSI evaluates freshwater withdrawal relative to available water after accounting for environmental flow requirements, making it more environmentally sensitive than WEI [25]. Nevertheless, WSI remains predominantly annual in scope and does not explicitly address hydrological variability or storage buffering. EFR indicators focus exclusively on the volume of water that must remain in rivers to sustain ecosystem functions, often utilizing low-flow statistics such as Q₉₀ or fixed percentages of mean annual flow [25,26]. These indicators do not estimate the volume of water that can be safely extracted after ecological needs are satisfied. The RRV framework is more dynamic than WEI, WSI, and EFR, as it employs time-series analyses to assess water system performance in terms of reliability, recovery capacity, and failure severity and can explicitly incorporate storage and variability [27]. However, RRV necessitates a comprehensive supply–demand simulation and produces multiple performance metrics rather than a single extraction-oriented indicator.

Existing indicators have been developed to quantify water scarcity, stress, and withdrawal pressure at large spatial scales; however, these indicators demonstrate significant limitations when applied to daily hydrological processes. These indices do not simultaneously incorporate hydrological variability, low-flow conditions, environmental flow requirements, storage-buffering effects, and operational extraction conditions at the basin scale and daily temporal resolution within a single metric. To address this limitation, the Extractable Water Index (EWI) was developed. The EWI quantitatively estimates the potential volume of extractable flow using river-gauge flow data and incorporates essential characteristics of the river environment.

2. Materials and Methods

The novel Extractable Water Index (EWI) presented in this study aims to assess the sustainable extraction of water from a flow with storage capabilities. This index was calculated from flow series data, with particular focus on low-flow periods, and incorporates daily flow measurements while accounting for environmental flow requirements. Two distinct scenarios were analyzed in the study. The first scenario assumes no storage and relies solely on river flow, for which a specific formula has been established. In contrast, the second scenario applies a semi-analytical approximation that considers storage, such as that provided by a reservoir or pond. Ultimately, the EWI yields a single value that effectively encapsulates the insights from both scenarios.

2.1. No Storage Scenario

In the no-storage scenario, the condition of a typical river was assessed using available river-flow data. This assessment included a detailed examination of low-water periods and environmental flow ($Q_e$). The $Q_e$ was analyzed using the Flow–Duration Curve (FDC), focusing on the 10^th percentile flow. Historical river flow data were examined to gain insights into how natural variables, such as precipitation patterns and watershed characteristics, affect river health. The daily gauges flow ($Q\left(t\right)$ (m³/s), $t = 1, 2, \dots , n$) is used to determine the daily available instantaneous flow $\left(Q_{avail} \left(t\right)\right)$ is defined in Equation (1).

$$Q_{avail}\left(t\right)=max\left[Q\left(t\right)-Q_e,0\right]$$

(1)

Hence, the reliability for candidate ($D$) (no storage) is the fraction of days $\left(n\right)$, $Q_{avail}\left(t\right)\ge D$ is given in Equation (2).

$$R\left(D\right)={\displaystyle \frac{1}{n}}\sum _{t=1}^{n}I\left(Q_{avail}\left(t\right)\ge D\right)$$

(2)

where I is the indicator function and $R\left(D\right)$ is the reliable extraction. The maximum constant extraction ($D^{\mathrm{*}}$ in m³/s) that reliably achieves $R_{target}$ is the empirical quantile of the $Q_{avail}$ series (Equation (3)). F⁻¹ is the inverse cumulative distribution function and $R_{target}=0.95$, hence the $D^{\mathrm{*}}$ is the 5th percentile of the $Q_{avail}$.

$$D^{\mathrm{*}}=F_{Q_{avail}}^{-1}\left[1,-R_{target}\right]$$

(3)

Additionally, with no storage, the supply on any day cannot exceed instantaneous availability; therefore, meeting demand on $p\%$ of the days is similar to choosing $D^{\mathrm{*}}$ equal to the $(1-p)$ quantile of the available flow distribution.

2.2. With Storage Scenario

In this scenario, the analysis focuses on storage, with the reliable extraction capacity increasing due to deficits in the storage buffer. While an exact solution necessitates continuous simulation, having an interpretable approximation can serve as a useful reference or initial basis for a design formula. The estimation begins with the reservoir mass balance, as in Equations (4) and (5)

$$S\left(t+1\right)=S\left(t\right)+Q_{avail}\left(t\right)∆-D∆$$

(4)

$$0\le S\left(t\right)\le S_{max}$$

(5)

where S(t) is the storage at day t (m³), $Q_{avail}\left(t\right)$ is the available flow after environmental flow deduction (m³/s), D is the constant extraction rate (m³/s) and $∆$ is 86,400 s/day. During a drought of duration $L_d$ the total deficit is given by Equation (6) (explanations and further calculations are shown in Equations (7)–(9)).

$$Deficit=(D-{\mu }_a)∆L_d$$

(6)

To avoid failure:

$$Deficit\le S_{max}$$

(7)

Thus

$$(\mathrm{D}-{\mu }_a)∆L_d\le S_{max}$$

(8)

Solving for D:

$$D\le {\mu }_a+{\displaystyle \frac{S_{max}}{∆L_d}}$$

(9)

Hence, a pragmatic maximum constant extraction (m³/s) is given in Equation (10).

$$\tilde{D}={\mu }_a+{\displaystyle \frac{S_{max}}{∆L_d}}$$

(10)

where $L_d$ is the design drought duration estimate from the historical daily flow sequence and $S_{max}$ is the usable storage (m³). The mean available flow $\left({\mu }_a\right)$ is given in Equation (11).

$${\mu }_a={\displaystyle \frac{1}{n}}\sum _{t=1}^n\left(Q_{avail}\left(t\right)\right)$$

(11)

The parameter ${\mu }_a$ can be replaced with a lower quantile to introduce additional conservatism. Additionally, the available flow fluctuates around its mean, potentially leading to an underestimation of risk. For a random variable $Q_{avail}$ with the coefficient of variation ${CV}_a$, the probability of deficit increases as variance increases. To address this, a first-order variance-based risk correction factor ($f_{var}$) can be incorporated with $\tilde{D}$, as reliance solely mean conditions may underestimate risk. The $D_{adj}$ is determined from the reduced $\tilde{D}$ by applying factor ($f_{var}$), as defined in Equation (12). The first-order variance-based $f_{var}$ and ${CV}_a$ are specified in Equations (13) and (14).

$$D_{adj}=f_{var}\mathrm{x}\tilde{D}$$

(12)

$$f_{var}={\displaystyle \frac{1}{(1+{CV}_a)}}$$

(13)

$${CV}_a={\displaystyle \frac{{\sigma }_a}{{\mu }_a}}$$

(14)

where ${\sigma }_a$ is the standard deviation of $Q_{avail} \left(t\right)$. The precise approximate extraction $D_{approx}$ defined as shown in Equation (15).

$$D_{approx}={\displaystyle \frac{{\mu }_a+\frac{S_{max}}{(∆L_d)}}{1+{CV}_a}}$$

(15)

The $L_d$ examined historical sequences, such as the longest consecutive run of days with $Q<Q_e$ or days with a specific condition: $Q_{avail}=0$ and selected the design drought length based on criteria like the mean of the top five longest droughts or the 90th percentile drought length. To ensure that the approximation is conservative, it must be calibrated by conducting the full mass-balance simulation for one or two candidate designs. The dynamics of storage influence this approximation, and timing plays a critical role; if storage replenishes slowly, the effective contribution is diminished.

2.3. Extractable Water Index (EWI)

The Extractable Water Index (EWI) was developed as a dimensionless measure indicating the availability of water from a river or other water source, accounting for ecological requirements, natural flow variability, and potential storage buffering. The EWI was calculated as the ratio of average sustainable supply (or long-term extractable capacity), determined by combining surplus flow (after environmental flow releases) with any contributions from storage (Equation (16)). This storage includes a buffer and a low-flow cushion to account for variability. This represents hydrological stress and risk factors, integrating a low-flow statistic with a penalty for variability. The index can be computed from any gauged daily record and is based on time-series data that incorporates flow variability. It is, however, sensitive to record length and data quality. Environmental flows can be simplified using a fixed value as the probable flow of the flow duration curve $\left(Q_{90}=Q_e\right)$, without considering seasonal patterns, and storage is treated as a fundamental average contribution rather than a full reservoir operation. Consequently, high EWI values indicate that a river possesses ample and stable extractable water resources, whereas low values reflect limited or unreliable extraction potential. Hence, this indicator comprehensively addresses specific requirements for Sri Lanka that are not met by other commonly used water abstraction indicators [24,28,29,30,31,32].

$$\mathrm{E}\mathrm{W}\mathrm{I}={\displaystyle \frac{{\mu }_a+\frac{S_{max}}{(∆L_d)}}{Q_{avail}^{\left(p\right)}+{\alpha \sigma }_a}}$$

(16)

where $\alpha$ is the variability adjustment coefficient representing risk tolerance, and $Q_{avail}^{\left(p\right)}$ is the p-th percentile of $Q_{avail} \left(t\right)$. This index serves as a multi-criteria decision-support tool to enhance ecological sensitivity assessments. It also facilitates the prioritization of rivers based on their extractability. The EWI indicates that a higher value corresponds to greater potential for water extractions relative to low-flow risks.

2.4. Application of the EWI for Threshold Classification

Thresholds for the Extractable Water Index (EWI) were estimated from daily discharge data under no-storage conditions, as extracting river flow is challenging in this scenario. Twenty river flow time series distributed across Sri Lanka were selected to estimate the EWI, taking into account the representation of the three climatic zones and data availability. All river gauge data were obtained from the Irrigation Department of Sri Lanka, with the exception of Eluwankulama. For Eluwankulama, synthetic daily time series were generated using a calibrated hydrological model, as there is no existing river gauge. Details regarding model calibration, validation, and model simulation are available in Iresh et al. [33]. The EWI thresholds were defined empirically based on EWI values, mean daily flow, seasonal flow variation, and the climate zone of each river gauge. Figure 1 and

Table 1. Details of the River Stations Utilized for EWI Estimation.

River Basin	Station Name	Mean Daily Discharge (m3)	Catchment Area (km2)	Climate Zone
Kalu	Rathnapura	45.46	603	Wet
	Putupaula	173.73	2598	Wet
Kelani	Hanwella	88.97	1782	Wet
Gin	Baddegama	66.3	749	Wet
Nilwala	Pitabaddara	16.03	310	Wet
Walawe	Bolana	18.37	2432	Dry
Kirindi Oya	Kuda Oya	3.59	291	Dry
	Thanamalwila	8.07	749	Dry
Menik Ganga	Katharagama	7.38	787	Dry
Kumbukkan Oya	Nakkala	8.6	216	Intermediate
Heda Oya	Siyambalanduwa	4.18	295	Intermediate
Kala Oya	Eluwankulama	3.75	2532	Dry
Maduru Oya	Padiyatalawa	4.62	159	Intermediate
Mahaweli Ganga	Peradeniya	49.48	1168	Wet
	Manampitiya	94	7418	Dry
Mi Oya	Galgamuwa	2.18	299	Dry
Deduru Oya	Chilaw	76.54	2608	Intermediate
Maha Oya	Badalgama	42.41	1360	Wet
Attanagalu Oya	Dunamale	8.49	153	Wet
	Karasnagala	6.18	54	Wet

present detailed information on the daily time series data. Thresholds for the EWI were established through empirical analysis of data from 20 river gauges distributed throughout Sri Lanka. The overall methodology carried out is shown in Figure 2.

3. Results and Discussion

3.1. EWI Threshold Classification

Table 2. EWI for river basins.

River Basin	Station Name	EWI	River Basin	Station Name	EWI
Kalu	Rathnapura	0.816	Heda Oya	Siyambalanduwa	0.485
	Putupaula	0.720	Kala Oya	Eluwankulama	0.365
Kelani	Hanwella	0.727	Maduru Oya	Padiyatalawa	0.373
Gin	Baddegama	0.785	Mahaweli Ganga	Peradeniya	0.906
Nilwala	Pitabaddara	0.715		Manampitiya	0.485
Walawe	Bolana	0.374	Mi Oya	Galgamuwa	0.479
Kirindi Oya	Kuda Oya	0.575	Deduru Oya	Chilaw	0.508
	Thanamalwila	0.668	Maha Oya	Badalgama	0.550
Menik Ganga	Katharagama	0.492	Attanagalu Oya	Dunamale	0.690
Kumbukkan Oya	Nakkala	0.618		Karasnagala	0.647

and Figure 3 present EWI values for each station and illustrate how EWI varies with mean daily flow. The EWI was evaluated at stations representing major river basins within the country’s wet, intermediate, and dry climatic zones.

These zones display substantial differences in rainfall distribution, runoff generation, and river flow regimes, which directly influence the reliability of extractable water resources. The resulting EWI values, ranging from 0.365 to 0.906, demonstrate significant spatial variability in extractable water potential across diverse hydrological environments. Stations located in the wet climatic zone, including Peradeniya in the Mahaweli Ganga basin (EWI = 0.906; mean daily flow = 49.48 m³/s), Rathnapura in the Kalu basin (EWI = 0.816; mean daily flow = 45.46 m³/s), and Baddegama in the Gin basin (EWI = 0.785; mean daily flow = 66.3 m³/s), exhibit relatively high EWI values. These rivers receive annual rainfall exceeding 2500 mm and sustain perennial flow regimes, resulting in greater flow reliability and higher extractable water potential. EWI values above approximately 0.75 therefore indicate high extractable water availability, which is characteristic of wet-zone rivers. Moderate EWI values, ranging from approximately 0.60 to 0.75, are observed at several stations in the intermediate and wet climatic zones, such as Hanwella in the Kelani basin (EWI = 0.727; mean daily flow = 88.97 m³/s), Pitabaddara in the Nilwala basin (EWI = 0.715; mean daily flow = 16.03 m³/s), and Nakkala in the Kumbukkan Oya basin (EWI = 0.618; mean daily flow = 8.6 m³/s). These basins experience moderate rainfall or upstream flow regulation, which creates seasonal flow variability, resulting in sufficient but less stable extractable water availability compared to stations with EWI values greater than 0.75. Lower EWI values are predominantly observed in dry-zone basins, where rainfall is limited, and river flows are highly seasonal. Stations such as Eluwankulama in the Kala Oya basin (EWI = 0.365; mean daily flow = 3.75 m³/s), Bolana in the Walawe basin (EWI = 0.374; mean daily flow = 18.37 m³/s), and Padiyatalawa in the Maduru Oya basin (EWI = 0.373; mean daily flow = 4.62 m³/s) record the lowest EWI values. These rivers typically experience reduced baseflow and significant reductions in flow during the dry season, which constrain the reliability of extractable water. Although mean daily discharge reflects overall water availability, the results indicate that EWI is influenced by flow magnitude, hydrological variability, and environmental flow constraints. For example, Manampitiya station in the Mahaweli basin has a high mean daily flow (94 m³/s) but only a moderate EWI value (0.485), illustrating that variability and environmental flow requirements reduce the proportion of water that can be reliably extracted.

Figure 4 displays a heat map illustrating seasonal flow variation across 20 gauging stations. The normalized monthly flow data demonstrate pronounced seasonality, with distinct differences among stations in the timing and persistence of high-flow periods. Most stations experience reduced discharge during midyear months, followed by a significant intensification of flow toward the end of the year, indicating strong climatic control over runoff generation. Several stations exhibit a bimodal seasonal structure, suggesting multiple rainfall-driven runoff phases. In contrast, other stations exhibit prolonged periods of low flow, consistent with limited catchment storage and rapid recession

Analysis of the observed distribution of EWI values, in relation to climatic zones and mean daily flow characteristics, led to the development of the threshold classification presented in

Table 3. The EWI Threshold Classification.

EWI Range	Category	Hydrological Interpretation
EWI < 0.45	Low extractable water potential	Rivers with limited reliable extraction due to low flows or high variability
0.45 < EWI < 0.60	Moderate potential	Extraction possible but requires management and seasonal regulation
0.60 < EWI < 0.75	Good potential	relatively stable flows allowing moderate water extraction
EWI ≥ 0.75	High potential	high flow reliability and favorable conditions for water withdrawal

. The EWI thresholds established through this analysis indicate that the multi-basin approach effectively captures the combined effects of hydrological regime, climatic zone, and flow variability on extractable water potential. Wet-zone basins generally exhibit higher EWI values, which are associated with sustained baseflow and greater flow reliability. In contrast, dry-zone basins tend to show lower values, reflecting more pronounced seasonal constraints on water availability. Certain stations located in wet or intermediate zones exhibit low EWI values, even when the mean daily flow is high, due to upstream reservoir regulation. This situation introduces analytical complexity that the EWI method is equipped to address. The findings highlight the applicability of the EWI framework across a range of hydrological settings and establish a robust foundation for assessing extractable water potential in basin-scale water resource planning.

The computed EWI here uses $Q_{90}$, a fixed e-flow that represents more ecologically realistic seasonal environmental flows, or habitat-based rules may alter both $Q_e$ and the distribution of $Q_{avail}$. Additionally, the no-storage formulation provides a conservative lower bound on extractable capacity, the approximate with-storage formula, and a full, daily mass-balance simulation should be used to obtain final design values for infrastructure investments. The tuning parameter α, the percentile $p$ for $Q_{pl}$, and the design drought length $L_d$ are context-dependent; these should be reported and sensitivity-tested in policy applications.

Although the EWI thresholds were developed empirically, the index provides a clear metric for assessing water-extraction potential. The novel EWI and its associated thresholds establish a hydrologically meaningful classification system that accurately reflects the potential for water extraction in Sri Lankan river basins. This classification system facilitates the application of the EWI in national-scale water resource planning and basin-level water allocation assessments. However, these thresholds may not be reliable in climatic regions that differ substantially from tropical countries such as Sri Lanka. Despite this limitation, the method demonstrates robustness and can be adapted for application in other regions.

3.2. Limitations of the Case Study Application

The lack of storage capacity is the primary limiting factor in this analysis. Even a modest storage buffer ($S_{max}>0$) would help mitigate variability and enhance the EWI, thereby increasing reliability during dry spells. These findings underscore the importance of using indices such as the EWI for effective integrated water management. Unlike singular metrics such as mean annual flow or $Q_{90}$, the EWI synthesizes ecological thresholds, variability, and drought risks into a single, comprehensive, and interpretable figure, thereby providing more precise guidance for extraction licensing and allocation. This study did not include a comprehensive reservoir simulation; future research should address this limitation by incorporating such simulations.

In addition, this analysis assumes a fixed $Q_{90}$ environmental flow requirement, which may not adequately represent seasonal ecological needs. Implementing adaptive, seasonal environmental flows could yield more accurate estimates of extraction. Furthermore, using $Q_{pl}=0$ m³/s as the limiting denominator may underestimate the actual extraction potential. Future research should compare this approach with alternative indices, such as flow–duration-curve-based metrics or reliability–resilience–vulnerability measures [34].

3.3. Mapping to Sustainable Development Goals

The outcome of the presented work can be mapped to several United Nations Sustainable Development Goals. River water extraction for domestic use is intrinsically linked to several Sustainable Development Goals (SDGs) because it directly affects water security, ecosystem sustainability, food production, and climate resilience. Primarily, it advances SDG 6 (Clean Water and Sanitation) by providing households with safe and reliable water, thereby improving sanitation and public health outcomes. Nevertheless, SDG 6 also underscores the importance of sustainable withdrawals, requiring that extraction remains within renewable water limits and that environmental flows are preserved to avoid ecological degradation [35,36]. The research work is closely related to SDG 13 (Climate Action), as climate change modifies river flow regimes through increased variability, droughts, and extreme events, thereby increasing the vulnerability of water supply systems. Incorporating climate considerations into abstraction planning strengthens adaptive capacity and helps prevent over-extraction during periods of low flow [37].

River water extraction is also linked to SDG 2 (Zero Hunger) through the trade-offs involved in allocating water between domestic use and irrigation. Although domestic supply is prioritized for human well-being, excessive abstraction can diminish water availability for agriculture, thereby potentially reducing food production. Conversely, improved access to domestic water enhances food security by supporting better health and increased labor productivity [38]. Sustainable abstraction also advances SDG 3 (Good Health and Well-being) by reducing reliance on unsafe water sources and maintaining sufficient river flows to support water quality [38]. Furthermore, preserving environmental flows supports SDG 15 (Life on Land) and SDG 14 (Life Below Water) by safeguarding aquatic ecosystems, wetlands, and downstream habitats [39]. Furthermore, sustainable river water extraction represents a crucial intersection of human development, ecosystem protection, and climate resilience. Achieving balance in these areas requires integrated water resources management, and implementing EWI offers a promising approach to advancing the SDGs.

4. Conclusions

This study introduces the EWI, a simple, physically interpretable metric that synthesizes gauged daily hydrology, ecological flow protection, and drought buffering into a single index to indicate a river’s practical extraction potential. An EWI value below 0.45, in the absence of storage, indicates low potential extractability and significant exposure to low-flow days. This scenario necessitates conservative licensing and prioritization of storage or demand-management investments to ensure a reliable domestic supply and preserve ecological integrity. Although mean water availability is relatively high, substantial flow variability and days with zero surplus availability offset this, suggesting that extraction without buffering mechanisms would be susceptible to short-term drought events. When compared conceptually to existing basin-scale scarcity/pressure indicators (for example, WEI and WSI), the EWI offers two key advantages: (1) it is directly computed from the daily flow record and therefore captures intra-annual variability and flow sequencing that aggregate annual indicators; (2) it explicitly integrates a storage-equivalent term that quantifies how much usable storage would improve extractable capacity. These features address recent calls in the hydrological literature for metrics that reconcile ecological protection with operational water availability and allocation under non-stationarity and strong seasonality.

The EWI fills a practical gap between aggregated scarcity metrics and full simulation models: it is transparent, reproducible from daily gauged records, and readily extended to include storage and scenario testing. Unlike traditional indicators that rely on annual or monthly averages, the EWI is derived from a daily streamflow time series. As a result, the EWI is more sensitive to hydrological drought conditions and environmental flow requirements, and it incorporates storage dynamics and reliable extraction limits, which are essential for sustainable water resources management. The EWI therefore bridges the gap between large-scale water stress indicators and operational hydrological analysis. By integrating hydrological inflow variability, environmental flow requirements, storage dynamics, and reliable extraction limits, the index offers a more physically meaningful representation of sustainable water extraction capacity. This methodology is especially appropriate for highly seasonal river systems, such as those in Sri Lanka, where annual indicators may overestimate extraction potential by failing to capture dry-season deficits, monsoonal variability, or the buffering effects of reservoirs and tanks.