A Simple Time-Varying Sensitivity Analysis (TVSA) for Assessment of Temporal Variability of Hydrological Processes

Yelena Medina,Enrique Muñoz

doi:10.3390/w12092463

Yelena Medina, Enrique Muñoz

Open Access

https://doi.org/10.3390/w12092463

Copy DOI

Journal: Water	Publication Date: Sep 2, 2020
Citations: 3	License type: CC BY 4.0

Affiliation: Universidad Católica de la Santísima Concepción

Abstract

Time-varying sensitivity analysis (TVSA) allows sensitivity in a moving window to be estimated and the time periods in which the specific components of a model can affect its performance to be identified. However, one of the disadvantages of TVSA is its high computational cost, as it estimates sensitivity in a moving window within an analyzed series, performing a series of repetitive calculations. In this article a function to implement a simple TVSA with a low computational cost using regional sensitivity analysis is presented. As an example of its application, an analysis of hydrological model results in daily, monthly, and annual time windows is carried out. The results show that the model allows the time sensitivity of a model with respect to its parameters to be detected, making it a suitable tool for the assessment of temporal variability of processes in models that include time series analysis. In addition, it is observed that the size of the moving window can influence the estimated sensitivity; therefore, analysis of different time windows is recommended.

Highlights

Sensitivity analysis studies the impact of input factor variation on model results [1,2]
The first performs an identifiability analysis using the Regional SensitivityAnalysis (RSA) (DYNIA) method, the second using the PAWN, Fourier Amplitude Sensitivity Test (FAST), Sobol, and Morris methods in a moving window, and the third using the Sobol, Morris, and Variogram Analysis of Response Surfaces (VARS) methods via the Generalized Global Sensitivity Matrix (GGSM) [17]. These toolboxes were developed to facilitate the application of sensitivity analysis, one of the main drawbacks of Time-varying sensitivity analysis (TVSA) is its high computational cost, as the calculations are made repetitively for each time step. This cost is commonly assessed in terms of the quantity of model evaluations required for the method to generate acceptable results in terms of robustness; it is associated with a high execution time, as well as computer memory requirements that mean that TVSAs in some models with a large number of parameters cannot be run on a regular computer
In this study a simple function was developed that allows a TVSA with a low computational cost to be applied using the RSA method and maximum vertical distance (MVD) index

Summary

Introduction

Sensitivity analysis studies the impact of input factor variation on model results [1,2]. The first performs an identifiability analysis using the RSA (DYNIA) method, the second using the PAWN, FAST, Sobol, and Morris methods in a moving window, and the third using the Sobol, Morris, and VARS methods via the Generalized Global Sensitivity Matrix (GGSM) [17] These toolboxes were developed to facilitate the application of sensitivity analysis, one of the main drawbacks of TVSA is its high computational cost, as the calculations are made repetitively for each time step. This cost is commonly assessed in terms of the quantity of model evaluations required for the method to generate acceptable results in terms of robustness; it is associated with a high execution time, as well as computer memory requirements that mean that TVSAs in some models with a large number of parameters cannot be run on a regular computer. To support the analysis of the results, a comparison with the standardized precipitation index (SPI) is included

TVSA Function Description

Code Description

Objective Functions

Convergence Analysis

Study Area and Data

Study andabsence

TVSA Implementation

Computational Cost

Method

Conclusions

Size of the Window of Analysis