Research Article

Assessment of Water Scarcity Across Pakistan Using Machine Learning Techniques and Novel Water Stress Indicators

Mustajab Ali1*, Usman Ali1, Hammad Akhtar1,2, Muhammad Bilal1, Muhammad Imtinaan Ibraaz1, Muhammad Abad Arif1, Momna Asghar3, Tasleem Kausar4, Syed Hasnain Gillani1,5 and Hassam Ahmed1,6

1Department of Civil Engineering, Mirpur University of Science & Technology (MUST), Mirpur, AJK, 10250, Pakistan; 2Department of Built Environment, University of Derby, Kedleston Rd, Derby, DE22 1, UK; 3Department of Electrical Engineering, University of Azad Jammu and Kashmir (UAJK), Muzaffarabad, AJK, 13100, Pakistan; 4Department of Electrical Engineering, Mirpur University of Science & Technology (MUST), Mirpur, AJK, 10250, Pakistan; 5Department of Civil Engineering, University of Bologna, Viale del Risorgimento 2, 40136 Bologna, Italy; 6School of Business, Leeds Trinity University, 1 Trevelyan Square Leeds LS1 6AE, United Kingdom.

Received | March 21, 2025; Accepted | August 20, 2025; Published | March 11, 2026

*Correspondence | Mustajab Ali, Department of Civil Engineering, Mirpur University of Science & Technology(MUST), Mirpur, AJK,10250, Pakistan; Email: [email protected]

Citation | Ali, M., U. Ali, H. Akhtar, M. Bilal, M.I. Ibraaz, M.A. Arif, M. Asghar, T. Kausar, S.H. Gillani and H. Ahmed. 2026. Assessment of water scarcity across Pakistan using machine learning techniques and novel water stress indicators. Sarhad Journal of Agriculture, 42(1): 411-420.

DOI | https://dx.doi.org/10.17582/journal.sja/2026/42.1.411.420

Keywords | Falkenmark indicator, Machine learning, Pakistan, Water GAP2, Water withdrawal to availability ratio, Water scarcity

Copyright: 2026 by the authors. Licensee ResearchersLinks Ltd, England, UK.

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

Introduction

Water resources are essential for human existence, economic advancement, and ecological balance. Over the past several decades, human-induced water withdrawal has significantly increased, driven mainly rapid global population growth and improved living standards (Huang et al., 2021). The urbanization process is continuing at an unprecedented rate globally, with projections indicating an escalation to 6.7 billion (68.4%) by the year 2050 (He et al., 2021).

During the last century, human water demand has surged eightfold, a trend closely associated with the fourfold increase in global population. Pakistan’s population is expected to rise to 400 million by 2040 as a consequence of rapid urbanization with major cities expanding at an unprecedented speed (Arif et al., 2019). As concluded by Huang et al., (2021), major cities like Karachi and Lahore have become seasonal and perennial water scarce regions because of population growth. Additionally, due to the prevailing water scarcity situation, only 29 million hectares (Mha) of land out of a total of 80 Mha is deemed arable; however, merely 17 Mha is currently under significant canal irrigation (Khan et al., 2022). Consequently, the necessity to import food grains exerts further pressure on the economy, as Pakistan, being primarily an agrarian nation, heavily depends on its agricultural sector (Ali and Usman, 2025). Together with increasing food import dependency this restriction undermines national goals aligned with UN SDG 01 (No Poverty) and SDG 02 (Zero Hunger). At the same time, it is anticipated that climate change will significantly influence both the spatial and temporal distribution of water resources (Rahman et al., 2023; Wu et al., 2020). Consequently, urban water scarcities along with groundwater depletion are expected to emerge as a pressing concern in the near future (Ashraf et al., 2022; Liu et al., 2017). Hence, the concept of water scarcity warrants analysis through both physical and economic perspectives (Rosa et al., 2020).

In the Global Risks 2015 Report disseminated by the World Economic Forum (WEF, 2015), the crisis regarding freshwater supply was recognized as the foremost high-impact risk of contemporary significance. In this study, our primary focus is centered on blue water scarcity. Given that the phenomenon of water scarcity represents a substantial threat to food security, public health, socioeconomic progress, and ecosystem services. It is imperative to comprehend the spatial and temporal dynamics of water scarcity (Liu et al., 2017; Rosa et al., 2020). Physical water scarcity can be further categorized into green water scarcity and blue water scarcity. Green water scarcity denotes the deficiency of root-zone soil moisture that is essential for agricultural output (Gao et al., 2024). On the other hand, blue water scarcity refers to the lack of freshwater resources within both surface water and groundwater systems necessary to satisfy human consumption requirements (Liu et al., 2017). Economic water scarcity is characterized by a scenario in which renewable freshwater resources are accessible. However, the absence of adequate socioeconomic and institutional frameworks obstructs public access to these water resources (Yin et al., 2020). In the past, indicators like Falkenmark Indicator (Falkenmark et al., 1989) and Withdrawal to Availability Ratio (Alcamo and Henrichs, 2002) have widely been used to access water stress at both regional and global scales.

Machine-learning techniques have been proven to be satisfactory (Fatima and Rahimi, 2024; Kausar et al., 2023; Nguyen et al., 2023). The future climatic impacts of water scarcity in a data-scarce region like Pakistan can be efficiently modeled using advanced machine-learning techniques like KNN (Asghar et al., 2016). Both algorithms provide a powerful approach for water scarcity modeling, with KNN being better at short-term and polynomial regression being better at long-term forecasting (Hussain and Jan, 2019). Moreover, KNN has been found highly efficient in data-scarce regions (Zhang and Liu, 2018) and possesses the capacity to discern patterns and generate forecasts based on both historical data and latent trends that may not be readily distinct through conventional statistical approaches (Eed et al., 2024; Shaheen et al., 2024). Hence, the KNN algorithm along with polynomial regression was applied in this study owing to his robust performance in earlier studies (Ali et al., 2024; Zhang and Liu, 2018).

Prior to our research most of the studies were at a global scale (coarser scale)(Modi et al., 2022; Huang et al., 2021; Liu et al., 2017) and none of them conducted detailed investigations in water stress analysis. Therefore, using novel water stress indicators, this study aims to assess and forecast water scarcity in Pakistan at a much finer scale, i.e., 0.5-degrees, with the aim to provide a deeper insight into the water scarcity situation. In addition, climate change (RCP) and socio-economic scenarios (SSPs) are involved as such factors drive water stress trends at national, regional, and local scales to facilitate the sustainable management of water resources are considered (Sheikh et al., 2024; Yin et al., 2020), including sustainable practices (Al-Musabbir et al., 2022; Falkenmark and Rockström, 2004). Predicting how this blue water scarcity will evolve over the period of time is essential for proactive planning and such understanding is crucial, as it may potentially undermine the achievement of the United Nations Sustainable Development Goals (SDGs), particularly SDG 01 (No Poverty), SDG 02 (Zero Hunger) and SDG 11, which addresses Sustainable Cities and Communities (UN, 2015).

Materials and Methods

Study area

Pakistan is situated within the latitudinal coordinates of 23.5◦–37.5◦ N and longitudinal coordinates of 62◦–75◦ E, encompassing a total terrestrial expanse of 803,940 km² (Figure 1). The extensive climatological profile of the country is primarily characterized by regions of aridity and semi-aridity. This diversity in climatology can be attributed to the prevailing monsoonal patterns and various seasonal phenomena, leading to chilly and wet conditions in December, extremely arid and high-temperate weather in April, and hot, excessively humid conditions during September (Hussain et al., 2023; Almazroui et al., 2021).

 

Methodology

This study utilizes the hydrological simulation data from The Inter-Sectoral Impact Model Intercomparison Project (ISIMIP 2a and 2b, Gosling et al., 2017) for 2005–2040 which was obtained using the Global Hydrological Model (GHM) Water Global Assessment and Prognosis (WaterGAP2, Kupzig et al., 2024), which features a distinctive spatial resolution of 0.5° × 0.5°. WaterGAP2 climate change scenario RCP 8.5 along with socioeconomic data for SSP2 scenarios was employed to evaluate water usage in different economic sectors in Pakistan. Current GHM simulations incorporated hydrological (surface water discharge) and socioeconomic (water withdrawal) variables from the Hadley Centre Global Environment Model version 2 (HadGEM2-ES) and Model for Interdisciplinary Research on Climate (MIROC5). Global Climate Models (GCMs) are acknowledged for their effectiveness in climate science and in assessing the consequences of climate change on water resource systems (de Souza Costa et al., 2021). The preprocessing of the dataset included managing temporal alignment to bring down the population data and water availability variables to a common annual interval (2005-2040); spatial alignment to ensure consistent resolution (0.5° × 0.5°) for hydrological and population data. Furthermore, gap-filling or interpolation was performed to address missing data. In the absence of in-situ discharge or withdrawal records across Pakistan’s basins, model calibration and validation were performed using indirect yet robust approaches using coefficient of determination as a main statistical tool for the calibration of model. Although there exists a few global and regional water stress studies by Modi et al., 2022, Huang et al., (2021) and Liu et al., (2017), but we were unable to validate our results owing to difference in spatial-temporal scales as well as datasets used.

A multitude of indices have been formulated and employed to evaluate the current and prospective global water resources across catchment, national, and international frameworks (Liu et al., 2017). This study utilized two novel water stress indicators, namely the Falkenmark indicator (FI) and the Water Withdrawal to Availability Ratio (WTA). Details of both water stress indicators are provided below:

Falkenmark indicator

The FI is a widely utilized method for measuring water scarcity. It requires data on the population and the volume of available blue water in a specific area. Per capita water availability is expressed in (m3/cap/year) (Falkenmark et al., 1989).

Where, Surface runoff is measured annually in (m3/year) and population in numbers.

Water withdrawal to availability ratio

The WTA is another recognized metric for evaluating water scarcity, linking water usage to renewable resources. This ratio effectively measures the amount of water extracted from available resources.

A criticality ratio exceeding 40% indicates significant water stress as defined by Alcamo and Henrichs, (2002).

In this context, water withdrawal (m3/day) is defined as the volume of water extracted for residential consumption, agricultural irrigation, and commercial applications. On the other hand, water availability (m3/day) encompasses the total quantity of water present in a specified region, which includes both surface and groundwater sources.

For future analysis, we adopted the KNN technique due to its effectiveness in handling large datasets and its ability to model nonlinear relationships between predictor and response variables. It was also found useful in classifying the areas into different water stress categories based on the historical patterns (Eed et al., 2024). Hence, it can effectively model both the current and projected patterns of water scarcity in Pakistan. By combining the KNN and polynomial regression, we validated the predicted water stress levels across Pakistan with respect to time and space. Furthermore, the statistical parameter, coefficient of determination (R2) was used to assess the accuracy for both algorithms, predicting the observed and simulated water stress levels across Pakistan. Remote sensing imagery was then processed and analyzed using ArcGIS to visualize spatial and temporal trends. Such high-quality remote sensing spatial maps provide valuable insights into water scarcity dynamics across Pakistan (Aslam et al., 2024, Zahoor et al., 2023).

Results and Discussion

Spatial analysis

Falkenmark indicator

Spatial plots depict the progression of water stress throughout Pakistan spanning the years 2005 to 2040. The FI, derived from surface discharge datasets produced by the MIROC5 and HadGEM2-ES climate models in conjunction with population data from ISIMIP2a and 2b (RCP 8.5 and SSP2)(Schewe et al., 2014), elucidates the changing spatial configuration of water deficiency.

 

Spatial analyses from Figure 2(a to f) indicate an alarming trend; from 2005 onwards (Figure 2(a)) the per capita availability of freshwater is diminishing as Pakistan’s population continues to increase. Figure 2(c and d) also illustrates similar patterns of increasing water stress. Currently, a considerable proportion of the country faces severe or absolute water scarcity. The water scarcity situation is expected to worsen as evidenced by the FI dropping below the threshold of 500m3/person/year as illustrated in Figure 2(e and f) (Table 1, Falkenmark et al., 1989). Major urban areas such as Lahore, Rawalpindi, Faisalabad, Karachi, Sukkur, Thatta, Peshawar, Nowshera, Quetta, Karachi, Chaghi, and Panjgur are especially susceptible to water stress and diminished access to freshwater resources which align with the earlier findings of Huang et al., (2021) and Aslam et al., (2024).

 

 

Water withdrawal to availability ratio

Spatial analyses reveal a concerning trend of increasing water scarcity in Pakistan from 2005 to 2040. The WTA, derived from MIROC5 and HadGEM2-ES climate models, highlights the spatial distribution of water stress nationally. Initial studies identify eastern urban areas as critical hotspots, while the frequency of flash droughts has increased nationwide (Ali et al., 2024). The increase in withdrawal to availability ratio beyond 2015 period coincides with rapid urbanization, increased irrigation demands, and higher industrial consumption (Ashraf et al., 2022). This trend, combined with rising water withdrawals and population growth (Dalstein et al., 2022), has intensified water scarcity in various regions as evident from Figure 3 (c and d). Major urban centers such as Karachi, Thatta, Sukkur, Lahore, Rawalpindi, Faisalabad, Dera Ghazi Khan, Quetta, Panjgur, Gwadar, and Peshawar are notably vulnerable to these challenges as the values of WTA are approaching absolute scarcity (Table 2, Liu et al., 2017) and are frequently situated in areas characterized by aridity and semi-aridity (Ashraf et al., 2022), typically defined by low rainfall and high evaporation rates, and face heightened water stress as a consequence of limited water supply and not that fails to meet demand. Our findings corroborate prior research by Veldkamp et al., (2017) and Wada et al., (2011), emphasizing the significant impact of increasing water withdrawals and urban population growth on water scarcity in Pakistan.

 

Time series analysis

Falkenmark indicator

Figure 4 delineates the progression of water scarcity phenomena in Pakistan from 2005 to 2040. FI is a widely recognized metric employed to evaluate water scarcity and is quantified in cubic meters per capita annually. This time series encompasses both empirical and projected values of the FI, in addition to thresholds for various degrees of water scarcity: water scarcity (<1000 m³/capita/year), water stress (1000-2000 m³/capita/year), and physical water scarcity (500-1000 m³/capita/year) (Table 1, Falkenmark et al., 1989; Halder et al., 2024).

 

Table 1: Threshold for Falkenmark indicator (Falkenmark et al., 1989).

Sr. No	FI (m3/capita/year)	Stress Level
1	>1,700	No Stress
2	1,000-1,700	Stress
3	500-1,000	Scarcity
4	<500	Absolute Scarcity

 

This time series depicts an alarming trajectory of diminishing water availability in Pakistan during the past decade (Aslam et al., 2024).The actual values of the FI have persistently remained beneath the threshold indicative of water stress, signifying a continual condition of water scarcity. The projected values, derived from the analytical model, imply that the scenario is poised to deteriorate in the subsequent years. The FI is anticipated to decline further below the threshold of water scarcity. This suggests an escalating risk of water scarcity and its associated repercussions especially after 2030, ultimately causing diminished agricultural yield, restricted access to potable water, and heightened susceptibility to flash droughts and floods (Ali et al., 2025; Ali et al., 2024). A slight difference observed in the current and future forecasts in case of FI might be due to the lack of a detailed census in the country for many years. Since population stats from government agencies are reliable and are a very important parameter in computing the FI, the results show a strong sensitivity to the accuracy and constancy of population data.

Water withdrawal to availability ratio

Figure 5 illustrates the evolution of water scarcity in Pakistan from the year 2005 to 2040, with a particular emphasis on the WTA ratio, which is a widely recognized metric for evaluating water stress. This time series encompasses both empirical and projected figures, thereby illuminating the changing landscape of water scarcity over the specified timeframe. The overall trend shows an increasing trend, which means water scarcity is going to worsen in the future (Aslam et al., 2024; Veldkamp et al., 2017). WTA value is continuously increasing showing heightened water stress level in the country and such phenomenon is significant after 2020, i.e., where WTA>0.6. The reason could be an increase in population as predicted by FAO and many other studies (Ahmad et al., 2025; Van Dijk et al., 2021). Besides changes in water availability (m3) owing to variability in hydrological cycles could be a reason to such increased WTA values in the country (Habib, 2021).

Between 2005 and 2015, the WTA exhibits variability within the range of 0.3 to 0.4 (Table 2, Liu et al., 2017), signifying a state of moderate water stress. This interval is characterized by relatively stable hydrological conditions, but still under pressure. From 2015 to 2035, the ratio escalates to a range of 0.4 to 0.8, indicating a shift towards acute water stress, attributable to escalating demand, demographic expansion, and environmental challenges. After 2035, the WTA ratio attains a level of 1.0 (Table 2), signifying a condition of exceedingly high water stress. This upward trend indicates the cumulative repercussions of human activities, rapid urban development, and the adverse effects of climate change (Huang et al., 2021; Modi et al., 2022).

 

Table 2: Threshold for water withdrawal to availability ratio (Liu et al., 2017).

Sr No.	Water Withdrawal to availability ratio	Stress Level
1	>0-0.2	Low water stress
2	>0.2-0.4	Moderate water stress
3	>0.4-0.8	High water stress
4	>0.8-1.0	Very high –water stress

 

FI and WTA were examined on a grid cell level (0.5-degrees) for Pakistan using outputs provided by WaterGAP2 hydrological model. To determine high–risk zones, e.g., where FI < 1,700 m³/cap/year; WTA > 40% within administrative or catchment boundaries; both indices were computed annually using future projections, i.e., 2025–2040 under RCP 8.5 and SSP2. This allowed tracking the changes over time in FI, and WTA with the help of polynomial regression models.

 

Conclusions and Recommendations

This study provides the first fine-scale (0.5-degrees) machine learning-based forecast that reveals a notable escalation in water scarcity within Pakistan, attributed to urbanization, climate change, and population growth. The application of the Falkenmark Index and Water Withdrawal to Availability combined with the climate change scenario RCP 8.5 and socioeconomic data for SSP2 scenarios indicates that water stress is expected to escalate by 2040, especially in regions like Baluchistan and major urban areas in Punjab, Sindh, and Khyber Pakhtunkhwa. These results highlight the immediate necessity for integrated and sustainable water management approaches to alleviate the projected consequences on food security, public health, and economic viability. This study uniquely employed machine learning algorithms to predict water scarcity in Pakistan. The KNN algorithm achieved 71% accuracy in predicting the Falkenmark Index. For the Water Withdrawal to Availability ratio, both KNN and polynomial regression effectively projected the upward trend and escalating water stress. In WTA time series, polynomial regression was observed to be smooth which could imply that it can be employed for long-term forecasting. The KNN model, with its higher R-squared value, captured the short-term fluctuations better and can be considered more effective for near-term predictions, presenting an innovative framework for policymakers and researchers confronting this critical issue. The implementation of proactive strategies based on our results is essential for promoting resilient water resource management and achieving the objectives outlined in the UN SDGs, particularly those related to clean water and sustainable cities.

For future research, areas identified as high-risk by spatial plots should prioritize immediate groundwater recharge. Population and runoff data with a spatial resolution of 0.5o x 0.5o does not comprehensively describe the water variability in heterogeneous regions like mountains and urban zones. Hence, we recommend using a higher-resolution datasets in the future analysis. Moreover, inclusion of climate change and socio-economic scenarios can lead to detailed investigations. Lastly, looking for field data for model validation is important. Despite limitations regarding spatial scale, datasets, and indices, our study offers a pathway for improved water resources management in Pakistan. It emphasizes methods such as efficient irrigation, community-based water conservation, rainwater harvesting, and urban infrastructure development to mitigate future resource scarcity.

Acknowledgements

The authors express their gratitude to Mirpur University of Science and Technology (MUST), Mirpur,kind for their support.

Novelty Statement

High resolution water scarcity mapping in Pakistan using novel water stress indicators and in-corporating SSP and RCP scenarios.

Author’s Contribution

Mustajab Ali and Usman Ali: Conceptualization, methodology, data collection, re-sources, writing, review and editing, supervision.

Hammad Akhtar and Muhammad Bilal: Conceptualization, methodology, data collection, re-sources, software, writing.

Muhammad Imtinaan Ibraaz and Muhammad Abad Arif: Conceptualization, methodology, data collection, re-sources.

Momna Asghar, Syed Hasnain Gillani, Hassam Ahmed and Tasleem Kausar: Writing, review and editing.

Generative AI or AI assisted technology statement

Generative AI or AI assisted tools are not use while preparing this document.

Conflict of interest

The authors have declared no conflict of interest

References

Ahmad, I., M. Ali, H. Asghar, M. Okamoto, Y. Shirayama, Z. Talha, A.U. Orozalievna and M. Yuasa. 2025. Impact of Temperature, Humidity and Precipitation on the COVID-19 Cases: A Study across all Provinces in Pakistan.J., 8(2): p 21 https://doi.org/10.3390/j8030021

Alcamo, J. and T. Henrichs. 2002. Critical regions: A model-based estimation of world water resources sensitive to global changes, Aquat. Sci., 64: 352-362, https://doi.org/doi: 10.1007/PL00012591

Ali, M., M. Taha, M.S Aziz, H. Ahmed and H. Ahmed. 2024. Flash Flood Prediction of Panjkora River, KPK, Using Artificial Neural Networks (ANN) and Support Vector Machine (SVM). Techn. J., 3(ICACEE), 758-769

Ali, M., S.H. Gillani, and U. Ali. 2024. Flash Drought Monitoring in Pakistan Using Machine Learning Techniques and Multivariate Drought Indices. Techn. J., 3(ICACEE): 717-729.

Ali, M. and U. Ali. 2025. Remote sensing of seasonal variation of LAI and fAPAR in Pakistan under changing climate using machine learning. Fusion Journal of Engineering and Sciences. https://doi.org/10.64615/fjes.1.SpecialIssue.2025.24

Ali, M., U. Raziq, A. Qureshi, M. Zaid, U. Qaseem, and U. Ali. 2025. Integrating SWAT and HEC-RAS models for enhanced flood risk mapping in Swat River, KPK, Pakistan. Fusion J. Engineer. Sci. https://doi.org/10.64615/fjes.1.SpecialIssue.2025.58

Almazroui, M., M. Ashfaq, M.N Islam, I.U. Rashid, S. Kamil, M.A. Abid, and M. B. Sylla. 2021. Assessment of CMIP6 performance and projected temperature and precipitation changes over South America. Earth Sys. Environ., 5(2): 155-183. https://doi.org/10.1007/s41748-021-00233-6

Al-Musabbir A., M.A. Rahman, N. Anjum and M. Ali. 2022. Performance evaluation of new rotary blades and roller cutter of a versatile multi-crop planter on residue management. Sar. J. Agric., 38(5): 211-221. https://doi.org/10.17582/journal.sja/2022/38.5.211.221

Arif, G.M., M. Riaz, N. Faisal, M.J.K. Khattak, Z. Sathar, M. Khalil, M. Sadiq, S. Hussain and K. Khan. 2019. Climate, population, and vulnerability in Pakistan: Exploring evidence of linkages for adaptation: Population Council: New York, NY, USA. https://doi.org/10.31899/pgy10.1022

Ashraf, S., M. Ali, S. Shrestha, M.A. Hafeez, A. Moiz and Z.A. Sheikh. 2022. Impacts of climate and land-use change on groundwater recharge in the semi-arid lower Ravi River basin, Pakistan. Groundwat. Sustain. Develop., 17: 100743. https://doi.org/10.1016/j.gsd.2022.100743

Asghar, M., I. Arshad, I.A. Taj, G. Raja and A.K. Khan. 2016. Palm and Finger Segmentation of High-Resolution Images using Hand Shape and Texture: Palm and Finger Segmentation of High-Resolution Images using Hand Shape and Texture. Proceedings of the Pakistan Academy of Sciences: Phys. Computat. Sci., 53(4): 401-416.

Aslam, M.N., S. Ashraf, S. Shrestha, M. Ali, and N.C. Hanh. 2024. Climate change impact on water scarcity in the Hub River Basin, Pakistan. Groundwat. Sustain. Develop., 27: 101339. https://doi.org/10.1016/j.gsd.2024.101339

Dąbrowska. J., A.E.M. Orellana, W. Kilian, A. Moryl, N. Cielecka, K. Michałowska, and A. Włóka. 2023. Between flood and drought: How cities are facing water surplus and scarcity. J. Environ. Manag., 345: 118557. https://doi.org/10.1016/j.jenvman.2023.118557

Dalstein. F. and A. Naqvi. 2022. 21st Century water withdrawal decoupling: A pathway to a more water-wise world?. Water Resour. Econom., 38: 100197. https://doi.org/10.1016/j.wre.2022.100197

de Souza Costa C.E.A., C.J.C. Blanco and J.F. de Oliveira-Júnior. 2021. Impact of climate change on the flow regimes of the Upper and Middle Amazon River. Clim. Chan., 166(3): 45. https://doi.org/10.1007/s10584-021-03141-w

Eed. M., A.A. Alhussan, A.S.T. Qenawy, A.M. Osman, A.M. Elshewey, and R. Arnous. 2024. Potato consumption forecasting based on a hybrid stacked Deep Learning Model. Potato Res., 1-25. https://doi.org/10.1007/s11540-024-09764-7

Falkenmark. M. and J. Rockström. 2004. Balancing water for humans and nature: The new approach in ecohydrology, Earthscan, London.

Falkenmark. M., J. Lundqvist and C. Widstrand. 1989. Macro-scale water scarcity requires micro-scale approaches, Nat. Resour. Forum., 13: 258-267, doi: https://doi.org/10.1111/j.1477-8947.1989.tb00348.x

Fatima S.S.W. and A. Rahimi. 2024. A review of time-series forecasting algorithms for industrial manufacturing systems. Machin., 12(6): 380. https://doi.org/10.3390/machines12060380

Gao. H., M. Hrachowitz, L. Wang-Erlandsson, F. Fenicia, Q. Xi, J. Xia, and H.H. Savenije. 2024. Root zone in the Earth system. Hydrol. Earth Sy. Sci., 28(19): 4477-4499. https://doi.org/10.5194/hess-28-4477-2024

Gosling. S., H. Müller Schmied, R. Betts, J. Chang, P. Ciais, and F. Zhao. 2017. ISIMIP2a simulation data from the water (global) sector, available at: http://doi.org/10.5880/PIK .2017.010, GFZ Data Services, 2017. 

Habib, Z. 2021. Water availability, use and challenges in Pakistan-Water sector challenges in the Indus Basin and impact of climate change. Food Agriculture Organisation.

Halder, R.K., M.N. Uddin, M.A. Uddin, S. Aryal and A. Khraisat. 2024. Enhancing K-nearest neighbor algorithm: a comprehensive review and performance analysis of modifications. J. Big Data., 11(1): 113. https://doi.org/10.1186/s40537-024-00973-y

Hanasaki, N., S. Kanae, T. Oki, and K. Tanaka. 2008a. An integrated model for the assessment of global water resources–Part 1: Model description and input meteorological forcing. Hydrol. Earth Syst. Sci., 12 (4): 1007–1025. https://doi.org/10.5194/hess-12-1007-2008

He, C., Z. Liu, J. Wu, et al. 2021. Future global urban water scarcity and potential solutions. Nature Commun., 12: 4667 (2021). https://doi.org/10.1038/s41467-021-25026-3

Huang. Z., X. Yuan and X. Liu. 2021. The key drivers for the changes in global water scarcity: Water withdrawal versus water availability. J. Hydrol., 601: 126658. https://doi.org/10.1016/j.jhydrol.2021.126658

Hussain, A., K.Z. Jadoon, K.U. Rahman, and H. Khan. 2023. Analyzing the impact of drought on agriculture: evidence from Pakistan using standardized precipitation evapotranspiration index. Nat. Hazar., 115(1): 389-408. https://doi.org/10.1007/s11069-022-05559-6

Hussain, I. and N. Jan. 2019. Modeling the Impact of Climate Change on Water Resources in Pakistan using Machine Learning Techniques.” J. Hydrol., 577: 124-136.

Kausar, T., Y. Lu, A. Kausar, M. Ali and A. Yousaf. 2023. SD-GAN: A style distribution transfer generative adversarial network for Covid-19 detection through X-ray images. IEEE Access, 11, 24545-24560. https://doi.org/10.1109/ACCESS.2023.3253282

Khan, I., A. Akram, S. Fatima, and Z. Ahmad. 2022. Problems of agriculture in Pakistan: an insight into their solution. Pak. J. Biotechnol., 19(02): pp.73-83. https://doi.org/10.34016/pjbt.2022.19.2.73

Kupzig, J., N. Kupzig, and M. Flörke. 2024. Regionalization in global hydrological models and its impact on runoff simulations: a case study using WaterGAP3 (v 1.0. 0). Geosci. Mod. Develop., 17(17): 6819-6846. https://doi.org/10.5194/gmd-17-6819-2024

Liu, J., H. Yang, H.S.N. Gosling, M. Kummu, M. Flörke, and T. Oki. 2017. Water scarcity assessments in the past, present, and future. Earth future., 5(6): 545-559. https://doi.org/10.1002/2016EF000518

Modi, P., N. Hanasaki, D. Yamazaki, et al. 2022. Sensitivity of subregional distribution of socioeconomic conditions to the global assessment of water scarcity. Commun Ear. Environ., 3: 144. https://doi.org/10.1038/s43247-022-00475-w

Nguyen, D.T., S. Ashraf, M. Le, L.Q. Trung and M. Ali. 2023. Projection of climate variables by general circulation and deep learning model for Lahore, Pakistan. Ecol. Info., 75: 102077. https://doi.org/10.1016/j.ecoinf.2023.102077

Rahman, M.A., M. Ali, M.A. Mojid, and K.D.C.R. Dissanayaka. 2023. Crop coefficient, reference crop evapotranspiration and water demand of dry-season Boro rice as affected by climate variability: A case study from northeast Bangladesh. Irrigat. Drain., 72(1): 148-165. https://doi.org/10.1002/ird.2754

Rosa, L., D.D. Chiarelli, M.C. Rulli, J. Dell’Angelo and P. D’Odorico. 2020b. Global agricultural economic water scarcity. Sci. Adv., 6(18): eaaz6031. https://doi.org/10.1126/sciadv.aaz6031.

Schewe, J., J. Heinke, D. Gerten, I. Haddeland, N.W. Arnell, D.B. Clark, and P. Kabat. 2014. Multimodel assessment of water scarcity under climate change. Proceed. Nation. Acad. Sci., 111(9): 3245-3250. https://doi.org/10.1073/pnas.1222460110

Shahin, M., F.F. Chen, M. Maghanaki, S. Firouzranjbar and A. Hosseinzadeh. 2024. Evaluating the fidelity of statistical forecasting and predictive intelligence by utilizing a stochastic dataset. Int. J. Advan. Manufact. Technol., 1-31. https://doi.org/10.1007/s00170-024-14505-8

Sheikh, Z.A., S. Ashraf, S. Weesakul, M. Ali and N.C. Hanh. 2024. Impact of climate change on farmers and adaptation strategies in Rangsit, Thailand. Environ. Challen., 15: p.100902. https://doi.org/10.1016/j.envc.2024.100902

United Nations (UN). Transforming Our World: The 2030 Agenda for Sustainable Development. Seventieth Session of the United Nations General Assembly Resolution A/RES/70/1. (UN, New York, 2015).

Van Dijk, M., T. Morley, M.L. Rau and Y. Saghai. 2021. A meta-analysis of projected global food demand and population at risk of hunger for the period 2010–2050. Nature food., 2(7): pp.494-501. https://doi.org/10.1038/s43016-021-00322-9

Veldkamp, T.I.E., Y. Wada, J.C.J.H. Aerts, P. D¨oll, and P.J. Ward. 2017. Water scarcity hotspots travel downstream due to human interventions in the 20th and 21st centuries. Nat. Commun., 8(1): https://doi.org/10.1038/ncomms15697

Wada, Y., L.P.H. van Beek and M. F.P. Bierkens. 2011. Modeling global water stress of the recent past: on the relative importance of trends in water demand and climate variability. Hydrol. Earth Syst. Sci., 15(12): 3785–3808. https://doi.org/10.5194/hess-15-3785-2011

World Economic Forum. 2015. Global Risks 2015 10th Edition, Geneva, Switzerland.

Wu, W.Y., M.H. Lo, Y. Wada, J.S. Famiglietti, and Z.L. Yang. 2020. Divergent effects of climate change on future groundwater availability in key mid-latitude aquifers. Natur. Commun., 11(1): 3710. https://doi.org/10.1038/s41467-020-17581-y

Yin, Y., L. Wang, Z. Wang, Q. Tang, S. Piao, and J. Zhang. 2020. Quantifying Water Scarcity in northern China within the context of Climatic and Societal Changes and South-to-North Water Diversion. Earth. Futur., 8(8). https://doi.org/ 10.1029/2020EF001492

Zahoor, A.R., S.N. Khan, A. Arshad and R.A. Aslam. 2023. Remote Sensing and GIS Based Techniques for Monitoring and Conserving Water on Newly Developed Farmlands. In Emerging Technologies for Water Supply, Conserv. Manag., (pp. 83-107). Cham: Spring. International Publishing. https://doi.org/10.1007/978-3-031-35279-9_5

Zhang, L. and Y. Liu. 2018. Machine Learning Techniques for Predicting Water Availability in Data-Sparse Regions: A Case Study of Pakistan. Water Resour. Manage., 32(10): 3203-3216.