This review presents a comprehensive overview of global sensitivity analysis (SA) methods for model output. The paper discusses various approaches to determine which input variables most significantly influence an output quantity, such as the variance of an output variable. Three main categories of methods are identified: screening methods for identifying influential inputs, importance measures for quantitative sensitivity indices, and deep exploration of model behavior to assess input effects across their entire variation range. The paper illustrates these methods using a river height simulation model and compares results from different techniques. Screening methods, such as the Morris method, are used to quickly identify non-influential inputs. The Morris method classifies inputs based on their effect on the output, distinguishing between inputs with negligible effects, linear effects, and non-linear or interaction effects. The results from the Morris method on the river model show that certain inputs have significant influence on the output, while others do not. Importance measures, including linear and rank-based methods, are used to quantify the contribution of each input to the output variance. These methods include Pearson correlation coefficients, standard regression coefficients, and partial correlation coefficients. The results from these methods on the river model indicate that certain inputs have a significant impact on the output, while others do not. The Sobol' indices, a variance-based method, are used to decompose the output variance into contributions from individual inputs and their interactions. These indices provide a detailed understanding of the model's behavior and are used to estimate the total effect of each input on the output. The results from the Sobol' indices on the river model show that certain inputs have a significant impact on the output, while others do not. The paper also discusses other sensitivity measures, such as entropy-based and distribution-based indices, which provide complementary information to Sobol' indices. These methods are useful when the output distribution is not well represented by variance. The paper concludes that SA methods are essential for understanding the behavior of complex models and for making informed decisions in various fields. The methods discussed provide a range of tools for analyzing model outputs and identifying the most influential inputs. The paper also highlights the importance of considering the computational cost and model complexity when selecting an SA method. The review emphasizes the need for a clear understanding of the study objectives and the importance of using appropriate methods for different types of models and applications.This review presents a comprehensive overview of global sensitivity analysis (SA) methods for model output. The paper discusses various approaches to determine which input variables most significantly influence an output quantity, such as the variance of an output variable. Three main categories of methods are identified: screening methods for identifying influential inputs, importance measures for quantitative sensitivity indices, and deep exploration of model behavior to assess input effects across their entire variation range. The paper illustrates these methods using a river height simulation model and compares results from different techniques. Screening methods, such as the Morris method, are used to quickly identify non-influential inputs. The Morris method classifies inputs based on their effect on the output, distinguishing between inputs with negligible effects, linear effects, and non-linear or interaction effects. The results from the Morris method on the river model show that certain inputs have significant influence on the output, while others do not. Importance measures, including linear and rank-based methods, are used to quantify the contribution of each input to the output variance. These methods include Pearson correlation coefficients, standard regression coefficients, and partial correlation coefficients. The results from these methods on the river model indicate that certain inputs have a significant impact on the output, while others do not. The Sobol' indices, a variance-based method, are used to decompose the output variance into contributions from individual inputs and their interactions. These indices provide a detailed understanding of the model's behavior and are used to estimate the total effect of each input on the output. The results from the Sobol' indices on the river model show that certain inputs have a significant impact on the output, while others do not. The paper also discusses other sensitivity measures, such as entropy-based and distribution-based indices, which provide complementary information to Sobol' indices. These methods are useful when the output distribution is not well represented by variance. The paper concludes that SA methods are essential for understanding the behavior of complex models and for making informed decisions in various fields. The methods discussed provide a range of tools for analyzing model outputs and identifying the most influential inputs. The paper also highlights the importance of considering the computational cost and model complexity when selecting an SA method. The review emphasizes the need for a clear understanding of the study objectives and the importance of using appropriate methods for different types of models and applications.