Advancing global water use research

Understanding global water use is essential to address critical challenges in the management and sustainability of water resources. With sectoral water demands projected to increase and hydrological cycles undergoing significant changes due to climate change, advanced modeling tools are increasingly being used to guide decisions on water management, infrastructure investments, and policy making. Despite notable progress, the current generation of Earth System Models (ESMs) often falls short in accurately representing human-water interactions. This thesis seeks to bridge this gap by developing novel methodologies to advance global water use research, focusing on enhancing modeling capabilities, harmonizing existing water use datasets, and proposing innovative frameworks for water allocation.

A key contribution of this work is the development of a sectoral water use module integrated into the Community Earth System Model (CESM2). This module is among the first in Earth System Models to comprehensively represent all major sectoral water uses, including domestic, livestock, thermoelectric, manufacturing, mining, and irrigation. It incorporates sector-specific withdrawals and consumptions, features a two-way coupling with an internal routing model, and employs a basic sectoral competition algorithm. This enables detailed assessments of global water scarcity and the interactions between human-water dynamics and climate. The validation results demonstrate the ability of the model to capture global and regional historical patterns of water scarcity, setting a new benchmark for integrated water resource modeling in ESMs. For the first time, we show that non-irrigative sectoral water consumption has a very limited impact on local climate at scales larger than 100 km, while highlighting the critical importance of comprehensive sectoral inclusion in ESMs for accurate water scarcity assessments.

A critical challenge in global water use research is the widespread presence of inconsistencies in datasets at the transition between historical and future periods, which undermines the credibility of model simulations and analyses. These inconsistencies manifest themselves as discontinuities in mean annual values, abrupt shifts in seasonality, and spatial mismatches caused by differing spatial downscaling methods between periods. To address these issues, we introduce a novel harmonization algorithm that is applied to correct an ensemble of water use scenarios. We demonstrate the algorithm’s effectiveness and versatility in smoothing temporal transitions and eliminating spatial artifacts across all sectors, thereby establishing a robust foundation for more reliable global water use datasets.

The widespread magnitude of inconsistencies in global water use data, observed across all datasets analyzed, suggests that many studies published today may have drawn conclusions based on flawed data and simulations. Human water use is a critical modulator of variables such as river discharge, groundwater depletion, water quality indicators, environmental flow requirements, and flood risks. To evaluate the degree to which these data errors propagate in related research, we conducted a case study on global water deficit calculations. Our findings reveal that unharmonized water use data can introduce substantial errors, often exceeding the impacts of socio-economic and climate scenarios at both local and national scales. In some cases, trend sign reversals were detected, where the use of harmonized versus original data altered the direction of projected changes in water deficits for certain countries. These results highlight the urgent need for improved quality control
of model inputs and outputs at the transition between historical and future periods. They also call into question the validity of analyses based on these inconsistent datasets.

Finally, this thesis proposes a new theoretical framework for sectoral water allocation in global hydrological models (GHMs). The framework introduces a distinction between essential and prosperity (non-essential) water demands for each sector. Essential demands represent baseline requirements necessary to prevent severe socio-economic and ecological impacts, while prosperity demands encompass discretionary uses that can be reduced during periods of scarcity. The framework is integrated with a ’traffic light’ system, inspired by the drought management practices of the Catalan Water Agency, to guide allocation decisions dynamically based on water availability. These innovations aim to provide a more realistic and actionable representation of sectoral water demands and competition under scarcity, serving as a blueprint for advancing future model development.

Overall, this thesis contributes to the field by advancing the representation of global water use in ESMs, addressing critical water use data inconsistencies while highlighting their potential impacts, and introducing a novel framework for modeling sectoral water use and competition in GHMs. Future research could build on this work by exploring the influence of land-atmosphere feedbacks on sectoral water scarcity — an area uniquely accessible through the two-way coupling between land and atmosphere in an ESM. In addition, we recommend to further advance the proposed harmonization strategies and implement better data quality protocols to ensure robust scientific results in the future. Finally, operationalizing the proposed allocation framework in global hydrological models could provide actionable insights for water resource management under scarcity conditions.