Orals: Tue, 5 May, 14:00–15:45 | Room 2.15
Just in the last four decades, ~5 million sq.km of humid land has transformed into a dryland. A rapidly warming climate is expected to further accelerate this dryland expansion. Consequently, not only will societies face permanent water insecurity, but plant and animal biodiversity will be under threat. Despite these consequences, very little is known about the physical mechanisms which cause irreversible drying of humid regions. As a result, our ability to predict the future expansion of drylands and its impact remains limited. Learning how drylands have expanded historically from humid regions could hold the key to predicting how they might expand in the future. So far, dryland expansion has been attributed to shifts in atmospheric circulation, topography, and orbital cycles. However, these changes occur at timescales reaching up to millions of years and thus do not fully explain the current rate of expansion. In this regard, the role of vegetation response to atmospheric drying and the consequent changes in land–atmosphere interactions have been largely ignored as possible mechanistic pathways of dryland expansion. Here, by tracking the air flowing over drylands, I show that the warming and drying of that air by changes in dryland vegetation-driven land–atmosphere feedback contributes to dryland expansion in the downwind direction. As they dry, drylands contribute less moisture and more heat to downwind humid regions, reducing precipitation and increasing atmospheric water demand, which ultimately causes their aridification. In ~40% of the land area that recently transitioned from a humid region into a dryland, self-expansion accounted for >50% of the observed aridification. Our results highlight the urgent need for climate change mitigation measures in drylands and provide a scientific basis for land-based interventions to prevent irreversible drying of humid regions
How to cite: Koppa, A.: Land–atmosphere feedbacks a mechanism of dryland expansion, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4955, https://doi.org/10.5194/egusphere-egu26-4955, 2026.
Drought propagation in dryland environments is often conceptualised as a sequential transfer of meteorological deficits into soil moisture, streamflow, and groundwater drought with the propagation rate controlled by catchment and aquifer properties. However, quantifying the spatial and temporal propagation of rainfall deficits remains challenging because few hydrological models explicitly connect climate forcing to all relevant hydrological stores and fluxes, or represent their interactions across spatial scales. Here, we investigate how meteorological drought propagates through overland flow, streamflow, and soil moisture into groundwater storage across dryland catchments in the Horn of Africa. We hypothesise that drought propagation will be related to aridity, catchment scale, and lateral groundwater connectivity. To investigate these ideas, we use the 1 km resolution DRYland water Partition (DRYP) hydrological model to simulate daily water-balance components across the Horn of Africa from 2000–2023. To examine the influence of basin size, catchments are delineated using a range of stream orders (4–8); for example, stream order 5 yields ~1,300 basins with a mean area of ≈900 km² spanning hyper-arid to humid conditions. We focus on regionally widespread meteorological drought events defined using SPEI-12 (derived from CHIRPS rainfall and hPET); catchment-mean SPEI is used to identify basins experiencing prolonged meteorological drought (> 12 months). We then use basin-scale total water storage anomaly (TWSA) and the standardised total water storage index (STWSI) to explore groundwater dynamics in basins experiencing prolonged drought. Our analysis reveals substantial spatial heterogeneity in groundwater response to meteorological drought. Even during prolonged drought events, many basins exhibit neutral or increasing TWSA trends, with coherent spatial clusters of both declining and increasing storage observed within the same drought period. These results demonstrate that meteorological drought does not consistently translate into groundwater drought, even over multi-year timescales. We explore how aridity, basin size, and lateral groundwater contributions can decouple groundwater dynamics from atmospheric water deficits, leading to enhanced drought resilience in some regions (and longer recovery in others). This work highlights the limitations of assuming fixed drought-propagation pathways in drylands and demonstrates the value of high-resolution modelling for improving drought monitoring and water-resource management under increasing climate variability.
How to cite: Blake, G., Michaelides, K., Singer, M., Quichimbo, E. A., and Cuthbert, M.: Controls on drought propagation in drylands, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7899, https://doi.org/10.5194/egusphere-egu26-7899, 2026.
Dryland catchments are highly sensitive to climatic variability, with evapotranspiration dominating the water balance and strongly constraining water availability. In irrigated drylands, understanding how climate change and land-use transformations jointly affect crop water consumption is critical for sustainable water management. This study analyses long-term changes in agricultural water use in the Amu Darya Basin, the largest transboundary river basin in Central Asia and one of the world’s most water-stressed regions.
We classified annual crop cover and cropping practices from 1987 to 2019 using the Landsat archive at 30 m spatial resolution. These crop cover maps served as a consistent input for estimating crop water consumption using satellite-based estimates of actual evapotranspiration, again derived from Landsat imagery, and computed with the Operational Simplified Surface Energy Balance (SSEBop) model, a water–energy balance approach well suited for dryland environments. A decomposition approach was applied to disentangle the relative contributions of climate change and land-use change to observed evapotranspiration dynamics.
Results show that total crop water consumption increased by about 10% over the study period, while average water use per unit area rose by 18%. Rising temperatures and increasing atmospheric evaporative demand alone would have pushed up water consumption by 21%. In contrast, shifts toward less water-intensive cropping practices, most notably from water-intensive summer cotton to winter wheat, offset only around 3% of this increase. Climate-driven effects intensified after the early 2000s and were strongest in downstream areas, where water stress, salinity, and ageing irrigation infrastructure limit adaptive capacity.
The findings demonstrate that, in irrigated dryland catchments, land-use change and cropping adjustments alone cannot counteract the accelerating impacts of climate change on evapotranspiration. All evapotranspiration and land-use datasets generated in this study are openly accessible, supporting transparency, reproducibility, and future research in data-scarce dryland regions. Our results underscore the need to combine improvements in water-use efficiency with climate mitigation and basin-scale management to strengthen hydrological resilience under continued warming.
How to cite: Müller, D., Senay, G., Umirbekov, A., Tarasova, L., Rufin, P., Pulatov, B., and Peña-Guerrero, M. D.: Climate-driven increases in crop water consumption in a Central Asian dryland catchment despite less water-intensive cropping, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21807, https://doi.org/10.5194/egusphere-egu26-21807, 2026.
Global climate change is reshaping precipitation regimes worldwide and intensifying aridity across many dryland regions. These changes also impact the dry subtropics of northwestern Australia. In this region, the strong spatial and temporal variability of rainfall complicates the characterisation of hydrologic processes, making it challenging to assess groundwater recharge, manage water resources effectively, and protect vulnerable ecosystems.
The region is marked by a distinct dry winter season (April-October) and a wet summer season (November-March), and it receives a far greater proportion of cyclone-driven rainfall than most other parts of Australia. These cyclonic systems contribute to pronounced seasonal and interannual variability, producing intense but short-lived flash-flooding events separated by extended droughts. During the study period (2015–2024), more than 80 % of total rainfall occurred between December and March across the five monitored weather stations. Mean annual rainfall ranged from 288 to 366 mm, while mean relative humidity remained low (30-37 %). Consequently, potential evaporation rates were extremely high, often exceeding 3000 mm/y.
To better understand the atmospheric processes governing precipitation formation and moisture sourcing in this environment, we analysed ten years of rainfall stable hydrogen and oxygen isotope compositions (δ¹⁸O and δ²H) from five sites. We also analysed 1,101 air parcel trajectories corresponding to rain events at five sites over the study period. Rainfall with low δ²H and δ¹⁸O values occurred predominantly during high-rainfall months, demonstrating a strong ‘amount effect’ at most locations. Both, stable hydrogen and oxygen isotope compositions were positively correlated with the stratiform fraction of total precipitation, indicating substantial sub‑cloud evaporation during stratiform events. On average, ~30 % of rainfall was lost to sub‑cloud evaporation, and back‑trajectory analysis showed that up to 47 % of wet-season moisture originated from recycled land evapotranspiration.
Differences between arithmetic and volume-weighted monthly isotope means highlight the seasonal importance of small-volume rainfall events. To address this bias, we introduce a new “cut‑off” method designed to reduce the disproportionate influence of low‑volume rainfall on monthly isotope compositions and on the construction of Local Meteoric Water Lines.
How to cite: Skrzypek, G., Wan, C., Dogramaci, S., Gleeson, J., Hedley, P., Grierson, P., and Gibson, J.: What the stable isotope composition of precipitation reveals when it rarely rains - a decade of observations from northwestern Australia, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16315, https://doi.org/10.5194/egusphere-egu26-16315, 2026.
Ephemeral river systems are key hydrological components of dryland catchments, particularly in semi-arid regions where surface water availability is highly episodic and closely linked to short-lived rainfall events. Shallow channel depressions temporarily store water and extend its availability beyond individual flow events. However, the extent to which sedimentation affects their long-term storage capacity remains poorly constrained, especially in data-scarce regions of southern Africa. This knowledge gap is increasingly relevant under rising water demand, pronounced climate variability, and more frequent droughts.
This study focuses on the Iishana system, a transboundary network of ephemeral channels and shallow depressions in northern Namibia and southern Angola. The system constitutes the primary rural water resource for a densely populated semi-arid region, with surface water availability largely restricted to short periods following rainfall events. Consequently, the capacity of channel depressions to retain water over extended periods is critical for domestic use, livestock, and small-scale agriculture. Understanding sediment accumulation processes in these depressions is therefore essential for improving water resource management in dryland catchments. The Iishana system is characterized by very low gradients, episodic runoff, and highly variable hydrological connectivity, making it representative of semi-arid environments where event-driven processes, storage, and transmission losses dominate.
The objective of this study is to characterize sediment properties and quantify sedimentation rates in selected depressions in order to assess their influence on surface water storage. Sediment cores were collected from multiple depressions and analyzed using physical and geochemical methods. Chronologies were established using 210Pb and 137Cs radionuclides, supported by radiocarbon dating. Sedimentation rates were calculated using Constant Flux-Constant Sedimentation (CFCS) and Constant Rate of Supply (CRS) models.
Sediments are predominantly fine-grained sand and silt, with weak pedogenic development, indicating limited and discontinuous deposition. The CFCS model results show low accumulation rates ranging from 0.017 to 0.12 cm yr-1, while CRS-derived mass accumulation rates range between 0.05 and 0.07 g cm-2yr-1. Below approximately 10 cm depth, sediment ages commonly exceed 150 years. In contrast, 137Cs activities were very low and lacked identifiable peaks, rendering this radionuclide unsuitable for chronologies in the study area.
The consistently low sedimentation rates indicate that natural sediment infill currently plays a negligible role in reducing the surface water storage capacity of ephemeral depressions within the Iishana system. From a hydrological perspective, this suggests that storage limitations are primarily controlled by hydrological connectivity, event-driven runoff generation, and infiltration rather than by progressive sediment accumulation. The results provide an empirical basis for evaluating the potential of selected depressions for targeted deepening or lateral expansion to enhance short-term surface water storage during episodic flow events.
Furthermore, spatial variability in sediment properties and accumulation rates highlights the importance of site-specific characteristics such as channel connectivity, flow velocity, and local catchment conditions, which are key controls on hydrological processes in dryland systems. By linking sediment dynamics with surface water storage, this study contributes to hydrological modeling, scenario-based planning, and sustainable water management in semi-arid catchments under increasing climatic stress.
How to cite: Fahrenberg, M., Reinhardt-Imjela, C., Katte, V., Pires, E., Jüpner, R., and Schulte, A.: Do sediments matter? Assessing sedimentation effects on surface water storage in a semi-arid ephemeral river system, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11411, https://doi.org/10.5194/egusphere-egu26-11411, 2026.
Climate seasonality, characterized by alternating drought and rainfall, drives the degradation and recovery of soil properties that shape hydrological functioning. These shifts can trigger ecohydrological responses such as increased runoff, reduced soil water availability, and vegetation decline, while subsequent recovery may enable partial or full restoration of soil hydraulic properties. The objective of this study was to determine how the interaction of these mechanisms, driven by stochastic climatic inputs, shapes the soil water balance, with particular emphasis on runoff generation.
We developed a conceptual model in which soil hydraulic properties are represented as a dynamic variable subject to degradation, recovery, or no change, formulated using discrete-state logic. In a specific implementation, state-dependent thresholds determine system behavior: degradation occurs as a discrete jump triggered by abrupt wetting of the soil, characterized by rapid wetting from a dry state to an upper moisture threshold; recovery follows a time-dependent trajectory when soil moisture is maintained within this threshold for a sufficient duration; and outside these conditions, no change occurs. This variable dynamically scales infiltration capacity, allowing rainfall under degraded conditions to generate surface runoff.
Resulting runoff time series were normalized by annual evaporation and classified using agglomerative hierarchical clustering with dynamic time warping. Three characteristic runoff patterns emerged: (i) rapid recovery with sub-decadal runoff decline; (ii) an intermediate transitional pattern with slower recovery and moderate runoff persistence; and (iii) permanent degradation associated with multi-decadal runoff regimes. To identify the drivers of these patterns, we analyzed a dimensionless set of parameters using linear discriminant analysis. The dominant control leading to degradation was the temporal clustering of rainfall, defined as the relative duration of rainfall events compared to the expected interarrival time between events.
Using dimensionless parameter combinations that express state-shifting forcings associated with each regime, we investigated clusters of runoff patterns. The results indicate that temporally isolated rainfall events can trigger irreversible state shifts leading to runoff-dominated regimes, whereas sustained or closely spaced rainfall events have the potential to initiate recovery processes spanning multiple decades.
How to cite: Hinz, C., Monhasser, A., and Wachsmut, G.: Event-based changes in hydraulic properties of surface soils in semi-arid regions may generate surface runoff regimes at decadal timescales, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9834, https://doi.org/10.5194/egusphere-egu26-9834, 2026.
Groundwater recharge is one of the least constrained hydrological fluxes in dryland catchments, particularly where thick vadose zones, ephemeral streamflow, and high evaporative demand decouple rainfall from aquifer replenishment. Recharge is commonly attributed to rare, high-magnitude floods, yet this perspective rarely accounts for event sequencing and vadose-zone memory. Here, we synthesise multi-year hydrometric observations and process-based simulations from an ephemeral dryland stream in the arid zone of Australia (Fowlers Gap, NSW) to show that temporal clustering of moderate streamflow events can enable focused recharge where thick vadose zones impose strong percolation thresholds. Field data indicate that several historic floods produced substantial vadose-zone wetting but no sustained groundwater response, whereas sequences of closely spaced, moderate flows generated delayed, yet persistent water-table rise beneath the channel.
Numerical simulations demonstrate that event clustering progressively wets the vadose zone, suppresses evapotranspiration losses, and non-linearly increases unsaturated hydraulic conductivity, resulting in sufficient flow via the streambed to measurably recharge the aquifer. These results show that groundwater recharge in dryland catchments emerges from the interaction between event sequencing, vadose-zone properties, transmission losses, and evapotranspiration, rather than from rainfall or streamflow magnitude alone. Under projected climate change, shifts toward more intense but less frequent rainfall may reduce recharge by disrupting event clustering, even where total precipitation remains unchanged. Explicitly accounting for vadose-zone memory and event sequencing is therefore essential for recharge estimation, model calibration, and dryland water-resource assessments.
How to cite: Rau, G. C., Bastías Espejo, J., Acworth, I., Andersen, M. S., Irvine, D., Bernardi, T., and Cuthbert, M. O.: Streamflow event clustering and vadose-zone memory as coupled controls on focused groundwater recharge in dryland catchments, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5983, https://doi.org/10.5194/egusphere-egu26-5983, 2026.
Flow Duration Curves (FDCs) are used to describe streamflow variability and support water-resources planning. In Mediterranean climates, intermittent rivers challenge FDC regionalization because zero-flow periods influence the lower part of the curve and discharge observations are scarce or fragmented in gauged basins. Southern Italy is a representative case where historical hydrological information is discontinuous and uncertain, so regional tools are needed to infer flow behavior in unmonitored catchments [1].
This study utilizes historical daily discharge from 58 gauging stations across three regions in southern Italy, covered by the Departments of Napoli (1925–1994), Bari (1950–1996), and Catanzaro (1950–1984). Stations were retained if they provide at least 15 years of records. Streamflow intermittency is quantified using the Intermittency Ratio (τ), defined as the fraction of non-zero daily discharge observations over the total number of daily observations. In the selected dataset, 20 stations exhibit τ < 1, indicating zero-flow days and intermittent behavior; these stations are considered in the intermittency analysis and regionalization.
For each basin, a hydrologically connected DEM was built by integrating an authoritative 20 m × 20 m DTM with the available river network and basin boundaries, enabling extraction of physiographic and morphologic descriptors in a GIS environment, supported by land-cover and geological and rainfall information [2,3]. A set of physiographic, topographic, and climatic covariates was analyzed, starting with collinearity assessment to reduce redundancy among predictors. The dependence of τ on catchment descriptors was investigated through stepwise regression using a power-law formulation. Results show that a parsimonious model based on two predictors—catchment area (A) and a catchment-shape descriptor (SF)—is sufficient to describe a significant portion of the observed variability in τ across the study region (as seen in figure 1).
Model performance is evaluated using standard statistical indices, including R², PBIAS, RMSE, MAE, and MAPE, which summarize explained variance, bias, and absolute/relative errors. The approach supports practical estimation of intermittency in ungauged Mediterranean catchments and provides a basis to incorporate intermittency into FDC regionalization, improving low-flow and zero-flow representation in data-constrained basins.
Figure 1. Observed vs predicted intermittency ratio (τ). Labels: B = Bari, N = Napoli, C = Catanzaro. Dashed line = 1:1.
Keywords: Intermittency ratio, Flow Duration Curve, regionalization, regression, stepwise selection, Mediterranean rivers, Southern Italy, GIS
References:
[1] Viola, F., Noto, L. V., Cannarozzo, M., & La Loggia, G. (2011). Regional flow duration curves for ungauged sites in Sicily. Hydrology and Earth System Sciences, 15(1), 323–331.
[2] Mendicino, G., & Senatore, A. (2013). Evaluation of parametric and statistical approaches for the regionalization of flow duration curves in intermittent regimes. Journal of Hydrology, 480, 19–32.
[3] Burgan, H. I., & Aksoy, H. (2022). Daily flow duration curve model for ungauged intermittent subbasins of gauged rivers. Journal of Hydrology, 604, 127249.
How to cite: Moradi, F., Padulano, R., and Del Giudice, G.: Regionalization of streamflow intermittency in Mediterranean catchments (Southern Italy), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11211, https://doi.org/10.5194/egusphere-egu26-11211, 2026.
Posters on site: Tue, 5 May, 16:15–18:00 | Hall A
Mediterranean coastal areas are prone and increasingly exposed to hydrological stress driven by water scarcity, climate change, increasing agricultural pressure, groundwater exploitation, and water quality degradation. These drivers manifest differently across regions but often result in common challenges and issues such as salinization, saltwater intrusion, and competition between agricultural and domestic or industrial water uses, thus impacting water management in irrigation districts.
In this context, the AI4Water PRIMA project investigates hydrological issues and water management challenges in four Mediterranean coastal irrigation districts and aims to apply Artificial Intelligence optimization and prediction techniques to improve water management. The four study areas are the Ras Jebel Coastal area (Tunisia), the Coastal Constantinois and Seybouse Basins (Algeria), the Capitanata Irrigation District (Italy), and the Nile Delta Basin (Egypt). Although these areas are all located in the Mediterranean region, they differ in population (from 50 thousands up to 3.5 million people), hydrogeological characteristics, water sources, irrigation practices, and water management policies.
This contribution, after introducing the AI4Water project, presents a preliminary comparison of the main hydrological and irrigation issues in the selected case studies, with a broader Mediterranean perspective. The comparison highlights both shared vulnerabilities and site-specific drivers of hydrological stress, emphasizing the need for context-dependent management strategies. By framing the different case studies within a common perspective, the project provides a basis for cross-district comparison and discussion, supporting the development of adaptive and transferable water management approaches for Mediterranean coastal systems. This comparative approach is intended to stimulate discussion and critical feedback from the scientific community working on similar or related case studies.
How to cite: Colosio, P., Bouguerra, H., Elsayed, S., Hanif, M. F., Khlifi, S., Mohammad, H., Onaindia, E., Ounaies, S., Peli, M., Ranzi, R., Serina, I., Signoroni, R., Tachi, S. E., Trablesi, F., and Barontini, S.: Hydrological issues in Mediterranean coastal semiarid irrigation districts in the context of the AI4Water PRIMA project, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-21452, https://doi.org/10.5194/egusphere-egu26-21452, 2026.
Rapid urban expansion in arid cities is transforming surface hydrology and amplifying flood risks under increasingly variable rainfall conditions. This study examines land use and land cover (LULC) change and its hydrologic implications in Al Ain, UAE, a transboundary basin influenced by orographic forcing from the Oman Mountains. Multi-temporal Landsat imagery from 1987–2023 was classified into five dominant surface types-sand, compacted sand, rocky terrain, built-up, and green areas using supervised Maximum Likelihood classification. Accuracy assessment yielded overall accuracies of 68–86% and kappa values up to 0.81, consistent with regional arid-environment benchmarks.
LULC analysis revealed a substantial shift from natural to managed surfaces, with built-up and green areas increasing by 274% and 1,667%, respectively, at the expense of rocky and sandy terrains. These changes were incorporated into a distributed GSSHA model to simulate rainfall–runoff responses across five major storm events between 2007 and 2024. Model results show that progressive urbanization markedly increased peak discharge (up to 78%) and runoff volume, particularly under low- to moderate-intensity storms where infiltration-excess (Hortonian) processes dominate. Under extreme events, the flood response became primarily governed by rainfall intensity, diminishing the relative impact of LULC change.
Spatial analysis emphasized strong localization of flood hazards within newly urbanized areas, with flood depths intensifying along residential and roadway corridors. While the expansion of irrigated green spaces enhanced infiltration locally, their spatial distribution limited broader runoff mitigation. The findings highlight storm condition dependent urban flood response in arid environments and emphasize the need for hydrologically informed urban design, including permeable pavements, vegetated buffers, and managed aquifer recharge systems.
This integrated approach, combining multi-temporal remote sensing and distributed hydrologic modeling, offers a transferable framework for evaluating urban flood dynamics in data-scarce arid regions, supporting policy efforts toward climate-resilient urban planning in rapidly developing desert cities.
How to cite: Bose, A. and Sharif, H.: Hydrologic consequences of rapid urbanization in an arid environment: Multi-temporal remote sensing and distributed flood modeling, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14306, https://doi.org/10.5194/egusphere-egu26-14306, 2026.
Streamflow forecasting is an essential component of effective water resources management, particularly in regions highly vulnerable to extreme hydroclimatic events, such as the Brazilian semi-arid region, which is characterized by pronounced spatial and temporal variability of precipitation, frequent droughts, and occasional flood events. The scarcity, irregularity, and limited duration of hydrological data in many watersheds of this region pose significant challenges to traditional hydrological modeling approaches, restricting the ability to make informed decisions in water resources planning and operational management. In recent years, machine learning–based models, particularly Long Short-Term Memory (LSTM) recurrent neural networks, have shown considerable potential for streamflow modelling due to their ability to capture complex nonlinear relationships and long-term temporal dependencies between precipitation, catchment storage, and runoff generation processes. However, the modelling performance is highly dependent on the availability of extensive and continuous historical records, which limits their direct applicability in data-scarce watersheds. In this context, transfer learning has emerged as a promising strategy to overcome these limitations by enabling the transfer of knowledge learned in well-monitored source sub-basins to improve predictions in target watersheds with limited data availability. This study aims to evaluate the transferability of deep learning models for streamflow modelling among watersheds of the Brazilian semi-arid region, considering different scenarios of data availability. The study also seeks to identify the main physical and hydrological parameters that influence both the performance and transferability of the models. LSTM models were initially pre-trained on watersheds with longer historical records and subsequently fine-tuned for watersheds with varying levels of available local data. Performance evaluation, conducted using widely adopted hydrological metrics, demonstrated that knowledge transfer is effective, allowing significant gains in predictive accuracy even when local datasets are limited. Furthermore, it was observed that certain hydrological and physiographic attributes exert a direct influence on the models’ ability to generalize to new basins. The application of eXplainable Artificial Intelligence (XAI) techniques further reinforced the physical consistency of the streamflow modelling, enhancing both interpretability and reliability of the results. Overall, the use of transfer learning proved to be a highly promising strategy for improving hydrological modelling in data-scarce semi-arid regions, reducing dependence on long-term monitoring, supporting more effective water resources management, and contributing to risk mitigation and sustainability in these vulnerable environments.
How to cite: Fonte Boa de Campos Maia, T., Marcela de Paula Kolanski, M., Rodrigues, A., and Brentan, B.: Transfer Learning for Streamflow Modelling Among Sub-Basins of the Brazilian Semi-Arid Region, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5965, https://doi.org/10.5194/egusphere-egu26-5965, 2026.
Irrigated agricultural farms are crucial for ensuring food security, but they affect the global water cycle through freshwater withdrawals. In this work, the irrigation water requirement in Al-Ahsa region of Saudi Arabia is estimated by combining the satellite-based soil moisture active and passive (SMAP) product by the National Aeronautics and Space Administration (NASA), evapotranspiration products by the Global Land Evaporation Amsterdam Model (GLEAM) and the European Reanalysis-5 (ERA5), and the precipitation product by Saudi Rainfall (SaRa) with the SM2RAIN algorithm. The study underlines the restricted ability of coarse-resolution satellite products to capture small agricultural farms and stresses selecting the proper evapotranspiration dataset for irrigation water estimation in arid regions. For analyzing the future changes in IWR under climate change, the research also generated multiple linear regression (MLR) models utilizing climate model data from General Circulation Models under SSP 1-2.6, SSP 2-4.5, SSP 3-7.0, and SSP 5-8.5 scenarios as predictor variables and SM2RAIN-estimated IWR as the target variable. The statistical evaluation of MLR models revealed that evapotranspiration was a significant predictive variable that allowed the models to account for 55% of the variation in future IWR. In summary, by estimating annual IWR and its change under various scenarios and baselines in Al-Ahsa Oasis, the results from this study offer comprehensive information and a point of reference for future research and highlight the factors that require further investigation for sustainable and dynamic water resource planning and agricultural management in arid environments under climate change scenarios.
How to cite: Al-Areeq, A. and Abdullah, M.: Application of SM2RAIN algorithm and linear regression model for future estimation of irrigation water use in an arid region, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2260, https://doi.org/10.5194/egusphere-egu26-2260, 2026.
How to cite: Onglassynov, Z., Absametov, M., Alimgazina, A., Muratova, M., and Domej, G.: Regional Assessment and Mapping of Groundwater Runoff in the Aral–Syrdarya Water Management Basin, Republic of Kazakhstan, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6797, https://doi.org/10.5194/egusphere-egu26-6797, 2026.
Hydrological rainfall-runoff models are essential tools for simulating and predicting watershed responses to meteorological forcings by linking precipitation and climate inputs with catchment characteristics such as land cover, topography, and soil properties. By representing key hydrological processes, these models support water resources management, extreme event forecasting, and infrastructure design. In semi-arid and data-scarce environments, modelling becomes particularly challenging due to high hydroclimatic variability and ephemeral flow regimes, often associated with limited or discontinuous observational records. Under these conditions, reliable rainfall-runoff simulations are essential not only for understanding catchment dynamics but also for supporting water resource management, including reservoir operation, allocation strategies, and drought risk mitigation. Regardless of their structure, all models require parameter calibration using observed data to ensure reliable reconstruction of hydrological response; this calibration represents a critical step, especially in data-limited contexts, where parameter uncertainty and equifinality pose significant challenges.
This study investigates the impact of different calibration strategies on model performance and parameter estimation under conditions of limited and incomplete observational data in a semi-arid region. The analysis was carried out using the IHACRES model (Identification of unit Hydrographs And Components from Rainfall, Evaporation and Streamflow; Jakeman, 1990), a parsimonious conceptual rainfall-runoff model specifically designed for applications in data-scarce environments. IHACRES consists of a non-linear loss module that converts rainfall into effective precipitation and a linear routing module that simulates both fast and delayed runoff components. The model was slightly modified and applied to several gauged catchments in Sicily (Italy), encompassing a wide range of climatic conditions and including many ephemeral streams. Calibration experiments were performed using a Monte Carlo approach and evaluated using both single- and multi-objective frameworks. Four complementary performance metrics were adopted as objective functions: Nash-Sutcliffe Efficiency (NSE), Kling-Gupta Efficiency (KGE), relative cumulative volume error (RVE), and an error metric based on flow duration curves signatures (D*). Single-objective calibration optimized individual metrics, whereas multi-objective configurations combined time series accuracy, water balance consistency, and flow regime representation in bi-, tri-, and tetra-objective setups. Multi-objective calibration explicitly incorporated equifinality through Pareto dominance theory, identifying non-dominated parameter sets and quantifying trade-offs among competing objectives.
Results indicate that single-objective calibration may reproduce specific hydrograph features but can misrepresent overall water availability and flow regime characteristics. In contrast, multi-objective calibration approaches can jointly constrain hydrograph dynamics, cumulative water balance, and flow regime behavior as represented by flow duration curves, leading to more reliable estimates of both high and low flows. Pareto-optimal analysis also revealed functional relationships among model parameters, suggesting opportunities to reduce parameter dimensionality and derive empirical relationships to estimate one parameter from another. This study demonstrates that multi-objective approaches offer significant advantages in explicitly addressing equifinality-driven parameter uncertainty, and that integrating Pareto-based optimization with uncertainty quantification improves the robustness and interpretability of hydrological simulations.
How to cite: Alongi, F., Alonzo, C., Francipane, A., and Noto, L. V.: Calibration Strategies for IHACRES in Data-Scarce Environments: Addressing Equifinality and Parameter Uncertainty, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17863, https://doi.org/10.5194/egusphere-egu26-17863, 2026.
Runoff responses to precipitation can vary widely across catchments and climates, shaped by hillslope water storage and release dynamics and by the transmission of hydrological signals through channel networks. Understanding these controls is critical for interpreting hydrological behavior and informing water resource management.
Here, we apply ensemble rainfall–runoff analysis (ERRA) to characterize runoff responses across 189 Iranian catchments spanning diverse landscapes and climates. ERRA quantifies the increase in lagged streamflow attributable to each unit of additional precipitation while accounting for nonlinear catchment behavior.
Peak runoff response, as quantified by ERRA across Iran, is higher in more humid climates, in steeper and smaller catchments, and in catchments with shallower water tables. The direction and approximate magnitude of these effects persist after accounting for correlations among the drivers (e.g., deeper water tables are more common in more arid regions).
These findings highlight the importance of catchment attributes in shaping runoff behavior, particularly in arid and semi-arid regions, where climatic variability and groundwater dynamics play a crucial role in sustainable water resource management and effective flood risk mitigation.
How to cite: Eslami, Z., Seybold, H., and Kirchner, J. W.: Climatic, topographic, and groundwater controls on runoff response to precipitation: evidence from a large-sample data set, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9343, https://doi.org/10.5194/egusphere-egu26-9343, 2026.
Mediterranean basins are increasingly exposed to sequences of dry and wet years that challenge reliable estimates of water availability and ecosystem resilience1,2.In karst catchments, strong surface–groundwater coupling and large subsurface storage fundamentally alter how precipitation is partitioned and how long hydrological anomalies persist. Although the physical behavior of karst systems is well known, most hydrological assessments and management-oriented studies still rely on models that do not explicitly represent karst processes. Consequently, it remains unclear how much water-balance estimates and management-relevant fluxes would change if karst dynamics were properly accounted for under persistent climatic extremes.
Here we address this gap using a counterfactual modeling framework that directly compares karst and non-karst realizations of the same basin under identical climate and land-use forcing. Our central research question is how climatic persistence (consecutive dry and wet years) controls precipitation partitioning in a Mediterranean karst basin, and how different those responses would be if the basin behaved as non-karst. We hypothesize that explicit representation of karst processes increases hydrological memory, making multi-year climatic sequences more influential than isolated extreme years and leading to substantially different estimates of key water fluxes.
We applied a physically based, distributed hydrological model to the 4,818 km² Mijares River basin in eastern Spain, a heterogeneous karst system with strong surface–groundwater interaction. The model was forced with meteorological data for 2000–2014 and a land-use map derived from the SIOSE 2014 classification. Using the same forcing and experimental design, we generated a counterfactual non-karst scenario in which karst-specific subsurface processes were suppressed.
At the basin scale, results show that karst-induced subsurface storage and delayed water transfer strongly amplify the impact of climatic persistence. In the karst configuration, sequences of dry or wet years exert a stronger control on the partitioning of precipitation into evapotranspiration, infiltration, runoff, subsurface flow, and percolation than do isolated extreme years. In contrast, the non-karst scenario exhibits weaker hydrological memory, with more immediate and climate-proportional responses. In several flux components, the karst and non-karst simulations diverge not only in magnitude but also in their implied hydrological functioning.
At the land-use class level, forests and shrublands in karst terrain promote infiltration and evapotranspiration with negligible surface runoff, reinforcing delayed subsurface responses during dry periods. Agricultural areas show higher interannual variability, while artificial surfaces generate disproportionate increases in runoff, particularly during persistent wet sequences. These contrasts are markedly attenuated in the non-karst experiment.
Overall, this paired karst–non-karst modeling approach demonstrates that omitting or oversimplifying karst processes can lead to substantial errors in both the magnitude and interpretation of water fluxes under persistent climatic conditions, providing a robust basis for evaluating water-resource decisions in karst regions facing increasing climate variability.
References
1 Rivas-Tabares, D., Tarquis, A. M., Willaarts, B., & De Miguel, Á. (2019). An accurate evaluation of water availability in sub-arid Mediterranean watersheds through SWAT: Cega-Eresma-Adaja. Agricultural Water Management, 212, 211-225.
2 Rivas-Tabares, D. A., Saa-Requejo, A., Martín-Sotoca, J. J., & Tarquis, A. M. (2021). Multiscaling NDVI series analysis of rainfed cereal in Central Spain. Remote Sensing, 13(4), 568.
How to cite: Vivas Betancourt, L. D., Rivas Tabares, D., Neumann, M. B., Herrero, J., and Sanz, M. J.: Paired karst–non-karst modeling reveals hidden sensitivity to climatic persistence, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13964, https://doi.org/10.5194/egusphere-egu26-13964, 2026.
Shallow, temporary salt lakes, known as ephemeral playas, are considered among the hydrogeological most sensitive systems to climatic extreme perturbations and flood extent. With a drainage basin exceeding 1,000,000 km2 and a lake depth of less than 6.5 m, Kati Thanda-Lake Eyre (KT-LE) in Southern Australia is characterized by a highly variable water balance and large water level fluctuations. It has reached its maximum level only once in the past 150 years, during the 1974-1977 “Great Filling”, emphasizing its status as one of the most unpredictable systems.
During the project, we aim to understand how hydro-climate variability across event magnitudes drives episodic lake filling and drying, and which basin-scale processes (inflow generation, transmission losses, evaporation, and potential groundwater interactions) control the water-level dynamics of KT-LE. We use the one-dimensional General Lake Model (GLM), forced with hourly ERA5 meteorological and hydrological reanalysis inputs over the 1974–2022 period to simulate daily lake level variations through time.
Initially, in this study, our model, calibrated in terms of surface energy balance and driven by a basin-averaged surface runoff, produces overestimated lake levels compared to satellite measurements. This suggests that basin-scale precipitation signals, largely driven by the northern catchment, do not necessarily reflect hydrological conditions farther south near the lake. To quantify water losses and the processes controlling them, we first define the fraction of inflow reaching the lake that reproduces the observed lake-level evolution with satisfactory agreement. We find that losses are strongly non-linear through time, exceeding 70–90% during minor floods, when precipitation affects only limited portions of the Lake Eyre Basin, and decreasing toward 0% during major floods, indicating a saturated and fully interconnected basin state. To identify the processes driving these non-linear losses, accounting for where precipitation occurs and which sub-basins are active, a river-focused approach using GLOFAS v4.0 (a channel-routing system) suggests that transmission losses estimated from spatial-mean ERA5 runoff may be overestimated in years when not all sub-basins contribute simultaneously to downstream flow.
Given the large surface area of the basin and the limited in-situ monitoring, these multi-scale results underline the importance of assessing and continuously evaluating the limitations of reanalysis-based climatic and hydrological forcing when applied to arid environments.
How to cite: Mama, A. N., Fluteau, F., Le Hir, G., May, J.-H., Faraon, T., Storms, J., and Schuster, M.: A modeling perspective on hydro-climate variability in dryland lakes : What is the impact of low- to high-frequency and intensity hydro-climate variability on the rise, development, persistence and demise of Lake Eyre - Kati Thanda ?, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13776, https://doi.org/10.5194/egusphere-egu26-13776, 2026.
The Qaidam Basin, located on the northern margin of the Qinghai-Tibet Plateau in China, is the second-largest inland basin in the country. Hosting numerous salt lakes rich in mineral resources, with particularly prominent lithium reserves. The Nalenggele River, the largest inland river within the Qaidam Basin, hosts the most extensive brine-type lithium deposit in China in its terminal salt lake region. With lithium resources reaching 2.3 million tons, this deposit accounts for approximately 87% of Chinese total lithium reserves. The genetic mechanism of brine in this terminal salt lake area is closely linked to hydrogeochemical processes. Previous studies have primarily focused on the qualitative analysis of hydrogeochemical processes, while relatively few have quantitatively assessed the impact of different hydrogeochemical processes. In this study, river water and groundwater from the mountainous areas to the basin within the watershed are selected as the research objects. Multiple isotopic tracers (δ²H, δ¹⁸O, ⁸⁷Sr/⁸⁶Sr, δ11B and δ7Li) are employed to trace the material sources and evolutionary processes of elements. The material sources, controlling factors and evolutionary mechanisms of hydrochemistry in the Nalenggele River Basin are clarified. The Positive Matrix Factorization model is applied to quantitatively identify the recharge sources of different water bodies, trace the material sources of major ions, and elucidate the evolutionary processes of the watershed hydrological cycle. The results show that: (1) The hydrochemical compositions of both river water and groundwater are dominated by Na⁺ and Cl⁻. The hydrochemical types evolve from mixed Cl·HCO₃·SO₄-Na·Ca type in the upstream rivers to Cl-Na type in the downstream waters. Analyses of ionic ratios and strontium isotope data confirm that water-rock interaction is the primary controlling factor of hydrochemical compositions, which is characterized by silicate weathering as the dominant process, supplemented by carbonate weathering and evaporite (halite, gypsum, mirabilite) dissolution. Cation exchange exhibits spatial heterogeneity: forward exchange (Ca²⁺/Mg²⁺ vs. Na⁺/K⁺) occurs in the upstream and downstream areas, while reverse exchange takes place in the midstream area. (2) Evidence from δ²H and δ¹⁸O indicates that river water is mainly recharged by atmospheric precipitation from the southern mountainous areas with the elevation of 4700m. Groundwater has close hydraulic connectivity with river water, showing bidirectional recharge-discharge interactions. (3) The observed B and Li isotopic footprints in the Nalenggele River Catchment are significantly depleted in heavy isotopes compared with those in other geological systems dominated by natural weathering processes. In the upper reaches of the Nalenggele River, the concentrations of B and Li increase sharply, while the δ¹¹B and δ⁷Li values decrease gradually. The mechanism responsible for the B and Li enrichment is mainly associated with the Li-B supply potential of material sources, favorable tectonic conduits for water circulation, and high evaporation rates. (4) The Positive Matrix Factorization model quantitatively reveals the contribution rates of different hydrogeochemical processes during the hydrological cycle, specifically: evaporite mineral dissolution (28%), mixed evaporite dissolution (25%), agricultural activity (17%), and silicate weathering (30%). This study provides a comprehensive framework for integrating multi-isotope tracers and statistical models to quantify hydrochemical processes in arid inland basins.
How to cite: Zhang, S., Zhao, C., Liu, K., and Zhang, Y.: Revealing hydrochemical characteristics and evolution process of river and groundwater in the Nalenggele Basin, northwest China: insights from major ions, multi-isotopes (H、O 、Sr、B and Li) tracers, and positive matrix factorization, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8648, https://doi.org/10.5194/egusphere-egu26-8648, 2026.
The Caatinga, the largest tropical dry forest in South America, holds significant yet often unrecognized potential for carbon sequestration and ecosystem functioning despite its highly seasonal and water-limited environment. However, carbon dynamics in this biome remain poorly quantified, especially regarding how vegetation structure, climate variability, and land-use interventions influence net primary productivity (NPP). This knowledge gap is particularly concerning given that the Caatinga is the Brazilian biome most severely affected by land degradation. Approximately 12% of its territory is classified within the two highest degradation classes, characterized by vegetation loss, low productivity, and depleted soil organic matter. This degradation process disproportionately affects traditional populations, such as Indigenous Peoples, Quilombola communities, and smallholder farmers, who rely directly on natural resources for their livelihoods.
To address this gap, we integrate satellite-based remote sensing, eddy-covariance observations, and ecological modeling to investigate spatial and temporal patterns of NPP under contrasting vegetation conditions and management regimes. First, we estimate NPP using a Light Use Efficiency (LUE) model driven exclusively by remote sensing inputs and compare these outputs with flux tower-derived NPP calculated from flux measurements collected during both a dry year and a wet year. This comparison enables the assessment of how semi-arid constraints, such as recurrent droughts, elevated temperatures, and soil water scarcity, shape photosynthetic efficiency and biomass accumulation. Once validated, the LUE model is applied to characterize spatial and temporal patterns of NPP in two contrasting socio-ecological contexts: a degraded area and a recovering area. This approach allows us to evaluate how contrasting management conditions influence the capacity of Caatinga vegetation to assimilate carbon.
Preliminary results indicate a strong sensitivity of NPP to rainfall variability and canopy structure, with degraded areas showing reduced carbon sequestration compared to conserved areas. These findings contribute to broader discussions on sustainable land management in the Brazilian Semi-Arid region and the urgent need for inclusive public policies to mitigate land degradation, protect ecosystems, and support the livelihoods of local communities.
How to cite: Oliveira, S., Bezerra, U., Lourenço, A., Valente, F., and Cunha, J.: Advancing NPP modelling to support sustainable management in Brazil’s semi-arid ecosystems, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-1220, https://doi.org/10.5194/egusphere-egu26-1220, 2026.
