Model Structure Uncertainty Research Articles

Assessing GFDL‐ESM4.1 Climate Responses to a Stratospheric Aerosol Injection Strategy Intended to Avoid Overshoot 2.0°C Warming

AbstractIn this work, we apply the GFDL Earth System Model (GFDL‐ESM4.1) to explore the climate responses to a stratospheric aerosol injection (SAI) scenario that aims to restrict global warming to 2.0°C above pre‐industrial levels (1850–1900) under the CMIP6 overshoot scenario (SSP5‐34‐OS). Simulations of this SAI scenario with the CESM Whole Atmosphere Community Climate Model (CESM2‐WACCM6) showed nearly unchanged interhemispheric and pole‐to‐Equator surface temperature gradients relative to present‐day conditions around 2020, and reduced global impacts, such as heatwaves, sea ice melting, and shifting precipitation patterns (Tilmes et al., 2020, https://doi.org/10.5194/esd‐11‐579‐2020). However, model structural uncertainties can lead to varying climate projections under the same forcing. Implementing identical stratospheric aerosol radiative properties in GFDL‐ESM4.1, which has a much lower Effective Climate Sensitivity compared to CESM2‐WACCM6, resulted in a decrease in global‐mean surface temperature by more than 1.5°C and a corresponding reduction in precipitation responses. Two major reasons contribute to the different temperature response between the two models: first, GFDL‐ESM4.1 has less warming in the SSP534‐OS scenario; second, GFDL‐ESM4.1 has shown more pronounced cooling in response to the same stratospheric AOD perturbation. Notably, the Southern Hemisphere experiences substantial cooling compared to the Northern Hemisphere, accompanied by a northward shift of the Intertropical Convergence Zone (ITCZ). Furthermore, our analysis reveals that spatially heterogeneous forcing within the SAI scenario results in diverse climate feedback parameters in the GFDL‐ESM4.1 model, through varying surface warming/cooling patterns. This research highlights the importance of considering model structural uncertainties and forcing spatial patterns for a comprehensive evaluation of future scenarios and geoengineering strategies.

On the importance of discharge observation uncertainty when interpreting hydrological model performance

Abstract. For users of hydrological models, the suitability of models can depend on how well their simulated outputs align with observed discharge. This study emphasizes the crucial role of factoring in discharge observation uncertainty when assessing the performance of hydrological models. We introduce an ad hoc approach, implemented through the eWaterCycle platform, to evaluate the significance of differences in model performance while considering the uncertainty associated with discharge observations. The analysis of the results encompasses 299 catchments from the Catchment Attributes and MEteorology for Large-sample Studies Great Britain (CAMELS-GB) large-sample catchment dataset, addressing three practical use cases for model users. These use cases involve assessing the impact of additional calibration on model performance using discharge observations, conducting conventional model comparisons, and examining how the variations in discharge simulations resulting from model structural differences compare with the uncertainties inherent in discharge observations. Based on the 5th to 95th percentile range of observed flow, our results highlight the substantial influence of discharge observation uncertainty on interpreting model performance differences. Specifically, when comparing model performance before and after additional calibration, we find that, in 98 out of 299 instances, the simulation differences fall within the bounds of discharge observation uncertainty. This underscores the inadequacy of neglecting discharge observation uncertainty during calibration and subsequent evaluation processes. Furthermore, in the model comparison use case, we identify numerous instances where observation uncertainty masks discernible differences in model performance, underscoring the necessity of accounting for this uncertainty in model selection procedures. While our assessment of model structural uncertainty generally indicates that structural differences often exceed observation uncertainty estimates, a few exceptions exist. The comparison of individual conceptual hydrological models suggests no clear trends between model complexity and subsequent model simulations falling within the uncertainty bounds of discharge observations. Based on these findings, we advocate integrating discharge observation uncertainty into the calibration process and the reporting of hydrological model performance, as has been done in this study. This integration ensures more accurate, robust, and insightful assessments of model performance, thereby improving the reliability and applicability of hydrological modelling outcomes for model users.

Comparative Analysis of Traditional and Pharmacometric-Based Pharmacoeconomic Modeling in the Cost-Utility Evaluation of Sunitinib Therapy.

Cost-utility analyses (CUAs) increasingly use models to predict long-term outcomes and translate trial data to real-world settings. Model structure uncertainty affects these predictions. This study compares pharmacometric against traditional pharmacoeconomic model evaluations for CUAs of sunitinib in gastrointestinal stromal tumors (GIST). A two-arm trial comparing sunitinib 37.5 mg daily with no treatment was simulated using a pharmacometric-based pharmacoeconomic model framework. Overall, four existing models [time-to-event (TTE) and Markov models] were re-estimated to the survival data and linked to logistic regression models describing the toxicity data [neutropenia, thrombocytopenia, hypertension, fatigue, and hand-foot syndrome (HFS)] to create traditional pharmacoeconomic model frameworks. All five frameworks were used to simulate clinical outcomes and sunitinib treatment costs, including a therapeutic drug monitoring (TDM) scenario. The pharmacometric model framework predicted that sunitinib treatment costs an additional 142,756 euros per quality adjusted life year (QALY) compared with no treatment, with deviations - 21.2% (discrete Markov), - 15.1% (continuous Markov), + 7.2% (TTE Weibull), and + 39.6% (TTE exponential) from the traditional model frameworks. The pharmacometric framework captured the change in toxicity over treatment cycles (e.g., increased HFS incidence until cycle 4 with a decrease thereafter), a pattern not observed in the pharmacoeconomic frameworks (e.g., stable HFS incidence over all treatment cycles). Furthermore, the pharmacoeconomic frameworks excessively forecasted the percentage of patients encountering subtherapeutic concentrations of sunitinib over the course of time (pharmacoeconomic: 24.6% at cycle 2 to 98.7% at cycle 16, versus pharmacometric: 13.7% at cycle 2 to 34.1% at cycle 16). Model structure significantly influences CUA predictions. The pharmacometric-based model framework more closely represented real-world toxicity trends and drug exposure changes. The relevance of these findings depends on the specific question a CUA seeks to address.

PyCHAMP: A crop-hydrological-agent modeling platform for groundwater management

The Crop-Hydrological-Agent Modeling Platform (PyCHAMP) is a Python-based open-source package designed for modeling agro-hydrological systems. The modular design, incorporating aquifer, crop field, groundwater well, finance, and behavior components, enables users to simulate and analyze the interactions between human and natural systems, considering both environmental and socio-economic factors. This study demonstrates PyCHAMP's capabilities by simulating the dynamics in the Sheridan 6 Local Enhanced Management Area, a groundwater conservation program in the High Plains Aquifer in Kansas. We highlight how a model, empowered by PyCHAMP, accurately captures human-water dynamics, including groundwater level, water withdrawal, and the fraction of cropland dedicated to each crop. We also show how farmer behavior, and its representation, drives system outcomes more strongly than environmental conditions. The results indicate PyCHAMP's potential as a useful tool for human-water research and sustainable groundwater management, offering prospects for future integration with detailed sub-models and systematic evaluation of model structural uncertainty.

Coastal generalized ecosystem model (CGEM) 1.0: Flexible model formulations for simulating complex biogeochemical processes in aquatic ecosystems

The Coastal Generalized Ecosystem Model (CGEM) is a biogeochemical model developed to study regulating processes of water-column optical properties, water-column and benthic carbon, oxygen, and nutrient cycles, and phytoplankton and zooplankton dynamics. CGEM offers numerous formulations for important rate processes, providing users flexibility in altering model structure. This flexibility also provides a means for evaluating model structural uncertainty and impacts on simulations, which are rarely evaluated with numerical ecosystem models. As an open-source model, CGEM also offers users the option to implement new formulations or modify existing routines. We also provide a full description of the model formulations, state variables, and model parameters in CGEM. Using two published case studies, we explore how different formulations for light attenuation, phytoplankton temperature growth response, and sediment processes impact simulations. We discuss CGEM's role as a new ecosystem model within the modeling community and opportunities to address current and future water quality issues.

Examining the Impact of Structural Uncertainty Across 10 Type 2 Diabetes Models: Results From the 2022 Mount Hood Challenge

ObjectivesThe Mount Hood Diabetes Challenge Network aimed to examine the impact of model structural uncertainty on the estimated cost-effectiveness of interventions for type 2 diabetes. MethodsTen independent modeling groups completed a blinded simulation exercise to estimate the cost-effectiveness of 3 interventions in 2 type 2 diabetes populations. Modeling groups were provided with a common baseline population, cost and utility values associated with different model health states, and instructions regarding time horizon and discounting. We collated the results to identify variation in predictions of net monetary benefit (NMB) and the drivers of those differences. ResultsOverall, modeling groups agreed which interventions had a positive NMB (ie, were cost-effective), Although estimates of NMB varied substantially—by up to £23 696 for 1 intervention. Variation was mainly driven through differences in risk equations for complications of diabetes and their implementation between models. The number of modeled health states was also a significant predictor of NMB. ConclusionsThis exercise demonstrates that structural uncertainty between different health economic models affects cost-effectiveness estimates. Although it is reassuring that a decision maker would likely reach similar conclusions on which interventions were cost-effective using most models, the range in numerical estimates generated across different models would nevertheless be important for price-setting negotiations with intervention developers. Minimizing the impact of structural uncertainty on healthcare decision making therefore remains an important priority. Model registries, which record and compare the impact of structural assumptions, offer one potential avenue to improve confidence in the robustness of health economic modeling.

Uncertainty analysis on long-term runoff projection from the Budyko framework and a conceptual hydrological model

ABSTRACT Runoff projections are subject to uncertainties related to model structure and parameters. This study aims to analyze uncertainties in long-term runoff estimations from an empirical (Budyko framework) and a conceptual hydrological model (MHD-INPE). Results indicate that both MHD-INPE and Budyko estimations tend to overestimate long-term runoff during years of recurring droughts. Pareto front solutions in MHD-INPE exhibited small uncertainties in long-term runoff estimations regarding parameter calibration (bias between 5 and 7%); differences were observed in low (bellow 5% variation) and high (bellow 10% variation) daily runoff. Related to model structure uncertainties, both models follow similar patterns and performance for a qualitative analysis. Budyko's future projections tend to exceed MHD-INPE's during high precipitation estimates, where at 2000 mm yearly precipitation the estimated runoff from Budyko tends to be 100 mm greater than the hydrological model. Under arid conditions Budyko tends to estimate smaller runoff than MHD-INPE due to variations in soil moisture and water storage not properly represented in Budyko's parameter. Although uncertainties were identified related to model complexity and calibrated parameters, higher uncertainties were identified as related to the climate models. Therefore, the Budyko method is a viable alternative for first-order analysis of long-term impacts.

Modelling the economy-wide effects of unilateral CO2 pricing under different revenue recycling schemes in Austria – Searching for a triple dividend

In this paper we identify policy implications for implementing carbon pricing in Austria, taking into account model structure uncertainty. Methodologically, we compare the results of two macroeconomic models, DYNK (a New Keynesian model) and WEGDYN-AT (a neoclassical general equilibrium model), to evaluate the effects of carbon pricing under different revenue recycling options. Specifically, we examine the model results with respect to their findings regarding a potential triple dividend, that we define as a simultaneous materialization of i) a reduction in CO2 emissions, ii) positive effects on GDP and employment, and iii) positive and distributionally progressive effects in household consumption possibilities. Accounting for model structure uncertainty improves the robustness of our results; at least within the macroeconomic frameworks applied here. Our results confirm the tradeoff between equity and efficiency identified in previous analyses. In the New Keynesian model, but not in the general equilibrium model, we find evidence for a triple dividend for a combination of the recycling options of non-wage labor cost reductions and lump-sum per-capita transfers (“climate bonus payments”).

Development of a multi-layer canopy model for E3SM Land Model with support for heterogeneous computing

The vertical structure of vegetation canopies creates micro-climates. However, the land components of most Earth System Models, including the Energy Exascale Earth System Model (E3SM), typically neglect vertical canopy structure by using a single layer big-leaf representation to simulate water, CO2, and energy exchanges between the land and the atmosphere. In this study, we developed a Multi-Layer Canopy Model for the E3SM Land Model to resolve the micro-climate created by vegetation canopies. The model developed in this study re-implements the CLM-ml_v1 to support heterogeneous computing architectures consisting of CPUs and GPUs and includes three additional optimization-based stomatal conductance models. The use of Portable, Extensible Toolkit for Scientific Computation provides a speedup of 25–50 times on a GPU relative to a CPU. The numerical implementation of the model was verified against CLM-ml_v1 for a month-long simulation using data from the Ameriflux US-University of Michigan Biological Station site. Model structural uncertainty was explored by performing control simulations for five stomatal conductance models that exclude and include the control of plant hydrodynamics (PHD) on photosynthesis. The bias in simulated sensible and latent heat fluxes was lower when PHD was accounted for in the model. Additionally, six idealized simulations were performed to study the impact of three environmental variables (i.e. air temperature, atmospheric CO2, and soil moisture) on canopy processes (i.e. net CO2 assimilation, leaf temperature, and leaf water potential). Increasing air temperature reduced net CO2 assimilation and increased air temperature. Net CO2 assimilation increased at higher atmospheric CO2, while decreasing soil moisture resulted in lower leaf water potential.

Research on Quantification of Structural Natural Frequency Uncertainty and Finite Element Model Updating Based on Gaussian Processes

During bridge service, material degradation and aging occur, affecting bridge functionality. Bridge health monitoring, crucial for detecting structural damage, includes finite element model modification as a key aspect. Current finite element-based model updating techniques are computationally intensive and lack practicality. Additionally, changes in loading and material property deterioration lead to parameter uncertainty in engineering structures. To enhance computational efficiency and accommodate parameter uncertainty, this study proposes a Gaussian process model-based approach for predicting structural natural frequencies and correcting finite element models. Taking a simply supported beam structure as an example, the elastic modulus and mass density of the structure are sampled by the Sobol sequence. Then, we map the collected samples to the corresponding physical space, substitute them into the finite element model, and calculate the first three natural frequencies of the model. A Gaussian surrogate model was established for the natural frequency of the structure. By analyzing the first three natural frequencies of the simply supported beam, the elastic modulus and mass density of the structure are corrected. The error between the corrected values of elastic modulus and mass density and the calculated values of the finite element model is very small. This study demonstrates that Gaussian process models can improve calculation efficiency, fulfilling the dual objectives of predicting structural natural frequencies and adjusting model parameters.

A hybrid model coupling process-driven and data-driven models for improved real-time flood forecasting

Accurate and reliable incoming flood forecasting is an important prerequisite for flood warning, flood risk analysis and reservoir flood control operation. This paper proposes a hybrid model for real-time flood forecasting that couples process-driven hydrological models (HMs) with data-driven models (DDMs). The generic hybrid model framework adds DDMs as the post-processing procedure for residual correction to the original results of HMs, and considers multiple uncertainties in input data, parameter and model structure simultaneously. The performance of the hybrid model is evaluated comprehensively in terms of deterministic forecast accuracy, interval forecast reliability, and the reliability and sharpness of probabilistic forecast. Taking the multireservoir system at the east Pi River as a study case, the results indicate that: (1) Compared to the benchmark model (ensemble XAJ model), the hybrid model with additional residual analysis show a significant improvement. The average continuous ranked probability score (CRPS) metric values calculated by the Stacking-Hybrid model improved by 71.5 %, 67.0 % and 38.1 % in the three data samples. Furthermore, the adaptability of the Stacking-Hybrid model for residual correction during short-duration intense rainfall events has been validated, with the relative error of the peak discharge improved to within ± 10 %. (2) The Stacking-Hybrid model, which also takes into account structure uncertainty, is able to better exploit the combined advantages and improve the stability of the model performance compared to those that only apply a single DDM. (3) When the number of iterations within the BOA reaches 300, the parameter optimization process is capable to search for the hyperparameters that bring out the best performance of the DDMs. (4) When the ensemble size reaches 200, the uncertainty of HM parameters can be fully defined, and the consumed computational resources can be controlled within an acceptable bound while ensuring stable model performance. Overall, the hybrid model that takes into account multiple sources of uncertainty generates both interval and probabilistic forecast in addition to deterministic forecast, which can provide richer risk information for subsequent flood warning and reservoir operation, making the flood prevention decisions more reliable and scientific.

Simulating the Role of Biogeochemical Hotspots in Driving Nitrogen Export From Dryland Watersheds

AbstractClimate change and nitrogen (N) pollution are altering biogeochemical and ecohydrological processes in dryland watersheds, increasing N export, and threatening water quality. While simulation models are useful for projecting how N export will change in the future, most models ignore biogeochemical “hotspots” that develop in drylands as moist microsites in the soil become hydrologically disconnected from plant roots when soils dry out. These hotspots enable N to accumulate over dry periods and rapidly flush to streams when soils wet up. To better project future N export, we developed a framework for representing hotspots using the ecohydrological model RHESSys. We then conducted a series of virtual experiments to understand how uncertainties in model structure and parameters influence N export to streams. Modeled N export was sensitive to three major factors (a) the abundance of hotspots in a watershed: N export increased linearly and then reached an asymptote with increasing hotspot abundance; this occurred because carbon and N inputs eventually became limiting as hotspots displaced vegetation cover, (b) the soil moisture threshold required for subsurface flow from hotspots to reestablish: peak streamflow N export increased and then decreased with an increasing threshold due to tradeoffs between N accumulation and export that occur with increasingly disconnected hotspots, and (c) the rate at which water diffused out of hotspots as soils dried down: N export was generally higher when the rate was slow because more N could accumulate in hotspots over dry periods, and then be flushed more rapidly to streams at the onset of rain. In a case study, we found that when hotspots were modeled explicitly, peak streamflow nitrate export increased by 29%, enabling us to better capture the timing and magnitude of N losses observed in the field. N export further increased in response to interannual precipitation variability, particularly when multiple dry years were followed by a wet year. This modeling framework can improve projections of N export in watersheds where hotspots play an increasingly important role in water quality.

Methodology to generate earthquake-tsunami fragility surfaces for community resilience modeling

Community resilience analysis and performance-based tsunami design methods typically use separate fragility curves for earthquake and near-field tsunami hazards to predict damage, neglecting the direct cascading effects of the earthquake damage on the subsequent tsunami damage analysis for coastal communities. This paper presents an improved two-phase successive analysis that can incorporate the most significant sources of uncertainties in structural modeling and hazards. However, generating multi-hazard 3D fragility surfaces that fully incorporate uncertainties through Monte Carlo Simulation requires millions of successive nonlinear analyses, which can be computationally intensive. To address this issue, a multi-distributed and parallel processing architecture was used and summarized herein. This methodology has been applied to a 4-story RC building as an illustrative example and demonstrated to be highly efficient and scalable in generating near-continuous fragility surfaces for earthquake and tsunami hazards, reducing the expected computational time by over 98%. Due to their near-continuous characteristic, these surfaces are easier to implement in resilience frameworks and applications. The fragility surfaces also feature earthquake-only and tsunami-only 2D fragility curves along the intensity measure axis boundaries, making them versatile for use in resilience frameworks that focus on a single hazard or both earthquake and tsunami hazards. Therefore, the proposed method is scalable making it helpful in developing fragility models for building portfolios to be used at the community resilience modeling of full coastal cities.

Collaborative Parameter Estimation of Multiple Unmanned Surface Vessels: A Robust Distributed Estimator-Based Approach

In this paper, the collaborative parameter estimation of multiple unmanned surface vessels (USVs) with model structure uncertainties is studied. The considered parameter estimation problem is first converted into a distributed state and parameter joint estimation problem. Then, the robust distributed estimator is constructed to handle the inevitable model structure uncertainties, and the upper bounds of prediction and estimation error covariance matrices are derived respectively. Additionally, the upper bound of the estimation error covariance matrix is minimized by designing appropriate estimator gains. Finally, the advantages of the proposed distributed parameter estimation approach from the perspectives of information interaction and consideration of model structure uncertainties are verified via simulations and practical experiments.

Synthesis of Evidence Yields High Social Cost of Carbon Due to Structural Model Variation and Uncertainties

Synthesis of Evidence Yields High Social Cost of Carbon Due to Structural Model Variation and Uncertainties

Synthesis of Evidence Yields High Social Cost of Carbon Due to Structural Model Variation and Uncertainties

Synthesis of Evidence Yields High Social Cost of Carbon Due to Structural Model Variation and Uncertainties

Bayesian Spatial Models for Projecting Corn Yields

Climate change is predicted to impact corn yields. Previous studies analyzing these impacts differ in data and modeling approaches and, consequently, corn yield projections. We analyze the impacts of climate change on corn yields using two statistical models with different approaches for dealing with county-level effects. The first model, which is novel to modeling corn yields, uses a computationally efficient spatial basis function approach. We use a Bayesian framework to incorporate both parametric and climate model structural uncertainty. We find that the statistical models have similar predictive abilities, but the spatial basis function model is faster and hence potentially a useful tool for crop yield projections. We also explore how different gridded temperature datasets affect the statistical model fit and performance. Compared to the dataset with only weather station data, we find that the dataset composed of satellite and weather station data results in a model with a magnified relationship between temperature and corn yields. For all statistical models, we observe a relationship between temperature and corn yields that is broadly similar to previous studies. We use downscaled and bias-corrected CMIP5 climate model projections to obtain detrended corn yield projections for 2020–2049 and 2069–2098. In both periods, we project a decrease in the mean corn yield production, reinforcing the findings of other studies. However, the magnitude of the decrease and the associated uncertainties we obtain differ from previous studies.

Model-to-model Bayesian calibration of a Chemical Reactor Network for pollutant emission predictions of an ammonia-fuelled multistage combustor

Low-fidelity, cost-effective, physics-based models are useful for assessing the environmental performance of novel combustion systems, especially those utilizing alternative fuels, like hydrogen and ammonia. However, these models require calibration and quantification of their limitations to be reliable predictive tools. This paper presents a framework for calibrating a simplified Chemical Reactor Network model using higher-fidelity Computational Fluid Dynamics data from a micro-gas-turbine-like combustor fuelled with pure ammonia. A Bayesian inference strategy that explicitly accounts for model error is used to calibrate the most relevant CRN parameters based on NO emissions data from CFD simulations and to estimate the model's structural uncertainty. The calibrated CRN model accurately predicts NO emissions within the design space and can extrapolate reasonably well to conditions outside the calibration range. By utilizing this framework, low-fidelity models can be employed to explore various operating conditions during the preliminary design of innovative combustion systems.

A Review of Computational Modeling in Wastewater Treatment Processes.

Wastewater treatment companies are facing several challenges related to the optimization of energy efficiency, meeting more restricted water quality standards, and resource recovery potential. Over the past decades, computational models have gained recognition as effective tools for addressing some of these challenges, contributing to the economic and operational efficiencies of wastewater treatment plants (WWTPs). To predict the performance of WWTPs, numerous deterministic, stochastic, and time series-based models have been developed. Mechanistic models, incorporating physical and empirical knowledge, are dominant as predictive models. However, these models represent a simplification of reality, resulting in model structure uncertainty and a constant need for calibration. With the increasing amount of available data, data-driven models are becoming more attractive. The implementation of predictive models can revolutionize the way companies manage WWTPs by permitting the development of digital twins for process simulation in (near) real-time. In data-driven models, the structure is not explicitly specified but is instead determined by searching for relationships in the available data. Thus, the main objective of the present review is to discuss the implementation of machine learning models for the prediction of WWTP effluent characteristics and wastewater inflows as well as anomaly detection studies and energy consumption optimization in WWTPs. Furthermore, an overview considering the merging of both mechanistic and machine learning models resulting in hybrid models is presented as a promising approach. A critical assessment of the main gaps and future directions on the implementation of mathematical modeling in wastewater treatment processes is also presented, focusing on topics such as the explainability of data-driven models and the use of Transfer Learning processes.

Contrasting hydrological responses to climate change in two adjacent catchments dominated by karst and nonkarst

Karst systems are well-known for their special features, such as epikarst zone and quick flow, compared to other hydrological systems. However, few studies explored the difference in hydrological responses to climate change between karst and nonkarst catchments. In this paper, two adjacent catchments dominated by karst and nonkarst in Southwest China (subtropical monsoon climate) were selected to investigate their different hydrological behaviors under future climates. To account for potential model structure uncertainties, three different karst hydrological models and three traditional catchment models were used to simulate these two catchments. The results demonstrate that different models yield comparable streamflow simulations within each catchment, allowing for the focus to be on the impact of climate change and catchment hydrological characteristics. The karst-dominated catchment is modeled to have larger actual evapotranspiration (AET) than the nonkarst-dominated catchment under similar climate conditions, with the AET of the former being more sensitive to changes in precipitation and potential evapotranspiration. These differences are primarily due to the presence of the epikarst in the unsaturated zone of the karst system, which can store more infiltration rainwater for evapotranspiration. Both catchments are expected to experience a wetter projection in the wet season and a drier projection in the dry season in the future. However, the karst-dominated catchment will face a drier climate with a significant increase in precipitation in just two months (July and August). Additionally, because of a higher sensitivity of monthly AET to climate change, the karst-dominated catchment will experience a greater monthly discharge reduction in the first two wet months (April and May), a relatively lower discharge increase in other wet months, and a lower monthly discharge in dry months, while the monthly discharge of the nonkarst-dominated catchment increases in the wet season and decrease in the dry season. Furthermore, the karst-dominated catchment will have a much lower annual discharge than the nonkarst-dominated catchment, particularly during extremely dry years in the future period due to its larger AET consumption. These all suggest that the karst-dominated catchment will be more vulnerable to droughts. Overall, due to the drier climate and the different hydrological backgrounds of the karst-dominated catchment, it is expected to have much drier hydrological conditions in the future compared to the nonkarst-dominated catchment.