scholarly journals On the choice of calibration metrics for “high-flow” estimation using hydrologic models

2019 ◽  
Vol 23 (6) ◽  
pp. 2601-2614 ◽  
Author(s):  
Naoki Mizukami ◽  
Oldrich Rakovec ◽  
Andrew J. Newman ◽  
Martyn P. Clark ◽  
Andrew W. Wood ◽  
...  
Keyword(s):  
Performance Metrics ◽  
Peak Flow ◽  
Hydrologic Model ◽  
Flood Frequency ◽  
High Flow ◽  
Hydrologic Models ◽  
Peak Flows ◽  
Annual Peak ◽  
Performance Metric ◽  
Squared Error

Abstract. Calibration is an essential step for improving the accuracy of simulations generated using hydrologic models. A key modeling decision is selecting the performance metric to be optimized. It has been common to use squared error performance metrics, or normalized variants such as Nash–Sutcliffe efficiency (NSE), based on the idea that their squared-error nature will emphasize the estimates of high flows. However, we conclude that NSE-based model calibrations actually result in poor reproduction of high-flow events, such as the annual peak flows that are used for flood frequency estimation. Using three different types of performance metrics, we calibrate two hydrological models at a daily step, the Variable Infiltration Capacity (VIC) model and the mesoscale Hydrologic Model (mHM), and evaluate their ability to simulate high-flow events for 492 basins throughout the contiguous United States. The metrics investigated are (1) NSE, (2) Kling–Gupta efficiency (KGE) and its variants, and (3) annual peak flow bias (APFB), where the latter is an application-specific metric that focuses on annual peak flows. As expected, the APFB metric produces the best annual peak flow estimates; however, performance on other high-flow-related metrics is poor. In contrast, the use of NSE results in annual peak flow estimates that are more than 20 % worse, primarily due to the tendency of NSE to underestimate observed flow variability. On the other hand, the use of KGE results in annual peak flow estimates that are better than from NSE, owing to improved flow time series metrics (mean and variance), with only a slight degradation in performance with respect to other related metrics, particularly when a non-standard weighting of the components of KGE is used. Stochastically generated ensemble simulations based on model residuals show the ability to improve the high-flow metrics, regardless of the deterministic performances. However, we emphasize that improving the fidelity of streamflow dynamics from deterministically calibrated models is still important, as it may improve high-flow metrics (for the right reasons). Overall, this work highlights the need for a deeper understanding of performance metric behavior and design in relation to the desired goals of model calibration.

scholarly journals On the choice of calibration metrics for high flow estimation using hydrologic models

2018 ◽  
Author(s):  
Naoki Mizukami ◽  
Oldrich Rakovec ◽  
Andrew Newman ◽  
Martyn Clark ◽  
Andrew Wood ◽  
...  
Keyword(s):  
Performance Metrics ◽  
Peak Flow ◽  
Hydrologic Model ◽  
High Flow ◽  
Hydrologic Models ◽  
Peak Flows ◽  
Annual Peak ◽  
Performance Metric ◽  
Squared Error ◽  
Application Specific

Abstract. Calibration is an essential step for improving the accuracy of simulations generated using hydrologic models, and a key modeler decision is the selection of the performance metric to be optimized. It has been common to used squared error performance metrics, or normalized variants such as Nash–Sutcliffe Efficiency (NSE), based on the idea that their squared-error nature will emphasize the estimation of high flows. However, we find that NSE-based model calibrations actually result in poor reproduction of high flow events, such as the annual peak flows that are used for flood frequency estimation. Using three different types of performance metrics, we calibrate two hydrological models, the Variable Infiltration Capacity model (VIC) and the mesoscale Hydrologic Model (mHM) and evaluate their ability to simulate high flow events for 492 basins throughout the contiguous United States. The metrics investigated are (1) NSE, (2) Kling–Gupta Efficiency (KGE) and variants, and (3) Annual Peak Flow Bias (APFB), where the latter is an application-specific hydrologic signature metric that focuses on annual peak flows. As expected, the application specific APFB metric produces the best annual peak flow estimates; however, performance on other high flow related metrics is poor. In contrast, the use of NSE results in annual peak flow estimates that are more than 20 % worse, primarily due to the tendency of NSE to result in underestimation of observed flow variability. Meanwhile, the use of KGE results in annual peak flow estimates that are better than from NSE, with only a slight degradation in performance with respect to other related metrics, particularly when a non-standard weighting of the components of KGE is used. Overall this work highlights the need for a fuller understanding of performance metric behavior and design in relation to the desired goals of model calibration.

Predictably Predictable -  The Role of Catchment Characteristics and Complexity.

2021 ◽  
Author(s):  
Sophia Eugeni ◽  
Eric Vaags ◽  
Steven V. Weijs
Keyword(s):  
Water Resource Management ◽  
Performance Metrics ◽  
Complexity Analysis ◽  
Model Performance ◽  
Principal Component ◽  
The United States ◽  
Hydrologic Model ◽  
Hydrologic Models ◽  
Hydrologic Modelling

<p>Accurate hydrologic modelling is critical to effective water resource management. As catchment attributes strongly influence the hydrologic behaviors in an area, they can be used to inform hydrologic models to better predict the discharge in a basin. Some basins may be more difficult to accurately predict than others. The difficulty in predicting discharge may also be related to the complexity of the discharge signal. The study establishes the relationship between a catchment’s static attributes and hydrologic model performance in those catchments, and also investigates the link to complexity, which we quantify with measures of compressibility based in information theory. </p><p>The project analyzes a large national dataset, comprised of catchment attributes for basins across the United States, paired with established performance metrics for corresponding hydrologic models. Principal Component Analysis (PCA) was completed on the catchment attributes data to determine the strongest modes in the input. The basins were clustered according to their catchment attributes and the performance within the clusters was compared. </p><p>Significant differences in model performance emerged between the clusters of basins. For the complexity analysis, details of the implementation and technical challenges will be discussed, as well as preliminary results.</p>

scholarly journals Analysis of Peak Flow Distribution for Bridge Collapse Sites

Water ◽  
2019 ◽  
Vol 12 (1) ◽  
pp. 52 ◽  
Author(s):  
Fahmidah U. Ashraf ◽  
Madeleine M. Flint
Keyword(s):  
Peak Flow ◽  
Flow Distribution ◽  
Heavy Tail ◽  
Flood Frequency ◽  
Return Periods ◽  
Peak Flows ◽  
Tail Distribution ◽  
Bridge Collapse ◽  
Flood Magnitude ◽  
Distribution Parameters

Bridge collapse risk can be evaluated more rigorously if the hydrologic characteristics of bridge collapse sites are demystified, particularly for peak flows. In this study, forty-two bridge collapse sites were analyzed to find any trend in the peak flows. Flood frequency and other statistical analyses were used to derive peak flow distribution parameters, identify trends linked to flood magnitude and flood behavior (how extreme), quantify the return periods of peak flows, and compare different approaches of flood frequency in deriving the return periods. The results indicate that most of the bridge collapse sites exhibit heavy tail distribution and flood magnitudes that are well consistent when regressed over the drainage area. A comparison of different flood frequency analyses reveals that there is no single approach that is best generally for the dataset studied. These results indicate a commonality in flood behavior (outliers are expected, not random; heavy-tail property) for the collapse dataset studied and provides some basis for extending the findings obtained for the 42 collapsed bridges to other sites to assess the risk of future collapses.

Clustering model responses in the frequency space for improved flood risk analysis

2021 ◽  
Author(s):  
Anna E. Sikorska-Senoner
Keyword(s):  
Time Series ◽  
Flood Risk ◽  
Hydrologic Model ◽  
Flood Frequency ◽  
Prediction Intervals ◽  
Hydrologic Models ◽  
Frequency Space ◽  
Computational Costs ◽  
Multiple Parameter ◽  
Clustering Approach

<p>Hydrologic models are employed in the flood risk studies to simulate time series of model responses to given inputs. These simulated time series of pseudo-observations can be next statistically analysed and in this way they can extend existing observed records. Simulations of hydrologic models are however associated with modelling uncertainty, often represented through a simulation ensemble with multiple parameter sets. The need of using multiple parameter sets to represent uncertainty is linked however with increased computational costs that may become prohibitive for long-time series and many input scenarios to be analysed. Due to the non-linear input-output relationship in the hydrologic model, a pre-selection of parameter sets is challenging.</p><p>This work presents a clustering approach as a tool to learn about the model hydrologic responses in the flood frequency space from the training dataset. Based on this learning process, representative parameter sets are selected that can be directly used in other model applications to derive prediction intervals at much lower computational costs. The study is supported with sensitivity analysis to the number of clusters. Based on results from a small catchment in Switzerland and 10’000 years of streamflow pseudo-observations, it has been found that grouping the full simulation ensemble with 1000 members into 3 to 10 clusters is already suitable to derive commonly applied prediction intervals (90%, 95% or 98%) in the flood frequency space. The proposed clustering approach can be applied in any flood risk analysis to lower the computational costs linked with the use of a hydrologic model.</p>

Scaling Theory of Floods for Developing a Physical Basis of Statistical Flood Frequency Relations

2017 ◽  
Author(s):  
Vijay Gupta
Keyword(s):  
Continuity Equation ◽  
Peak Flow ◽  
Flood Frequency ◽  
Scaling Theory ◽  
Rainfall Runoff ◽  
Self Similarity ◽  
Channel Networks ◽  
Annual Peak ◽  
Streamflow Data ◽  
Homogeneous Regions

Prediction of floods at locations where no streamflow data exist is a global issue because most of the countries involved don’t have adequate streamflow records. The United States Geological Survey developed the regional flood frequency (RFF) analysis to predict annual peak flow quantiles, for example, the 100-year flood, in ungauged basins. RFF equations are pure statistical characterizations that use historical streamflow records and the concept of “homogeneous regions.” To supplement the accuracy of flood quantile estimates due to limited record lengths, a physical solution is required. It is further reinforced by the need to predict potential impacts of a changing hydro-climate system on flood frequencies. A nonlinear geophysical theory of floods, or a scaling theory for short, focused on river basins and abandoned the “homogeneous regions” concept in order to incorporate flood producing physical processes. Self-similarity in channel networks plays a foundational role in understanding the observed scaling, or power law relations, between peak flows and drainage areas. Scaling theory of floods offers a unified framework to predict floods in rainfall-runoff (RF-RO) events and in annual peak flow quantiles in ungauged basins. Theoretical research in the course of time clarified several key ideas: (1) to understand scaling in annual peak flow quantiles in terms of physical processes, it was necessary to consider scaling in individual RF-RO events; (2) a unique partitioning of a drainage basin into hillslopes and channel links is necessary; (3) a continuity equation in terms of link storage and discharge was developed for a link-hillslope pair (to complete the mathematical specification, another equation for a channel link involving storage and discharge can be written that gives the continuity equation in terms of discharge); (4) the self-similarity in channel networks plays a pivotal role in solving the continuity equation, which produces scaling in peak flows as drainage area goes to infinity (scaling is an emergent property that was shown to hold for an idealized case study); (5) a theory of hydraulic-geometry in channel networks is summarized; and (6) highlights of a theory of biological diversity in riparian vegetation along a network are given. The first observational study in the Goodwin Creek Experimental Watershed, Mississippi, discovered that the scaling slopes and intercepts vary from one RF-RO event to the next. Subsequently, diagnostic studies of this variability showed that it is a reflection of variability in the flood-producing mechanisms. It has led to developing a model that links the scaling in RF-RO events with the annual peak flow quantiles featured here. Rainfall-runoff models in engineering practice use a variety of techniques to calibrate their parameters using observed streamflow hydrographs. In ungagged basins, streamflow data are not available, and in a changing climate, the reliability of historic data becomes questionable, so calibration of parameters is not a viable option. Recent progress on developing a suitable theoretical framework to test RF-RO model parameterizations without calibration is briefly reviewed. Contributions to generalizing the scaling theory of floods to medium and large river basins spanning different climates are reviewed. Two studies that have focused on understanding floods at the scale of the entire planet Earth are cited. Finally, two case studies on the innovative applications of the scaling framework to practical hydrologic engineering problems are highlighted. They include real-time flood forecasting and the effect of spatially distributed small dams in a river network on real-time flood forecasting.

A Flood Potential Index for Effective Communication of Streamflow Forecasts at Ungauged Communities

2020 ◽  
Vol 21 (4) ◽  
pp. 807-814 ◽  
Author(s):  
Felipe Quintero ◽  
Witold F. Krajewski ◽  
Marcela Rojas
Keyword(s):  
Peak Flow ◽  
Hydrologic Model ◽  
Drainage Network ◽  
Relative Importance ◽  
Proof Of Concept ◽  
Streamflow Forecasting ◽  
Annual Peak ◽  
Potential Index ◽  
Streamflow Forecasts ◽  
Forecasting System

AbstractThis study proposes a flood potential index suitable for use in streamflow forecasting at any location in a drainage network. We obtained the index by comparing the discharge magnitude derived from a hydrologic model and the expected mean annual peak flow at the spatial scale of the basin. We use the term “flood potential” to indicate that uncertainty is associated with this information. The index helps communicate flood potential alerts to communities near rivers where there are no quantitative records of historical floods to provide a reference. This method establishes a reference that we can compare to forecasted hydrographs and that facilitates communication of their relative importance. As a proof of concept, the authors present an assessment of the index as applied to the peak flows that caused severe floods in Iowa in June 2008. The Iowa Flood Center uses the proposed approach operationally as part of its real-time hydrologic forecasting system.

scholarly journals Projected Changes in the Frequency of Peak Flows along the Athabasca River: Sensitivity of Results to Statistical Methods of Analysis

Climate ◽  
2019 ◽  
Vol 7 (7) ◽  
pp. 88 ◽  
Author(s):  
Yonas Dibike ◽  
Hyung-Il Eum ◽  
Paulin Coulibaly ◽  
Joshua Hartmann
Keyword(s):  
Statistical Method ◽  
Climate Models ◽  
River Flow ◽  
Climate Model ◽  
Peak Flow ◽  
Hydrologic Model ◽  
Climate Change Scenarios ◽  
Peak Flows ◽  
Athabasca River ◽  
Projected Changes

Flows originating from alpine dominated cold region watersheds typically experience extended winter low flows followed by spring snowmelt and summer rainfall driven high flows. In a warmer climate, there will be a temperature-induced shift in precipitation from snowfall towards rain along with changes in precipitation intensity and snowmelt timing, resulting in alterations in the frequency and magnitude of peak flow events. This study examines the potential future changes in the frequency and severity of peak flow events in the Athabasca River watershed in Alberta, Canada. The analysis is based on simulated flow data by the variable infiltration capacity (VIC) hydrologic model driven by statistically downscaled climate change scenarios from the latest coupled model inter-comparison project (CMIP5). The hydrological model projections show an overall increase in mean annual streamflow in the watershed and a corresponding shift in the freshet timing to an earlier period. The river flow is projected to experience increases during the winter and spring seasons and decreases during the summer and early fall seasons, with an overall projected increase in peak flow, especially for low frequency events. Both stationary and non-stationary methods of peak flow analysis, performed at multiple points along the Athabasca River, show that projected changes in the 100-year peak flow event for the high emissions scenario by the 2080s range between 4% and 33% depending on the driving climate models and the statistical method of analysis. A closer examination of the results also reveals that the sensitivity of projected changes in peak flows to the statistical method of frequency analysis is relatively small compared to that resulting from inter-climate model variability.

scholarly journals Integrating partitioned evapotranspiration data into hydrologic models: vegetation parameterization and uncertainty quantification of simulated plant water use

2022 ◽  
Author(s):  
Adam Schreiner-McGraw ◽  
Hoori Ajami ◽  
Ray Anderson ◽  
Dong Wang
Keyword(s):  
Groundwater Recharge ◽  
Water Use ◽  
Precision Agriculture ◽  
Land Surface ◽  
Performance Metrics ◽  
Model Performance ◽  
Hydrologic Model ◽  
Plant Water ◽  
Hydrologic Models ◽  
Plant Water Use

Accurate simulation of plant water use across agricultural ecosystems is essential for various applications, including precision agriculture, quantifying groundwater recharge, and optimizing irrigation rates. Previous approaches to integrating plant water use data into hydrologic models have relied on evapotranspiration (ET) observations. Recently, the flux variance similarity approach has been developed to partition ET to transpiration (T) and evaporation, providing an opportunity to use T data to parameterize models. To explore the value of T/ET data in improving hydrologic model performance, we examined multiple approaches to incorporate these observations for vegetation parameterization. We used ET observations from 5 eddy covariance towers located in the San Joaquin Valley, California, to parameterize orchard crops in an integrated land surface – groundwater model. We find that a simple approach of selecting the best parameter sets based on ET and T performance metrics works best at these study sites. Selecting parameters based on performance relative to observed ET creates an uncertainty of 27% relative to the observed value. When parameters are selected using both T and ET data, this uncertainty drops to 24%. Similarly, the uncertainty in potential groundwater recharge drops from 63% to 58% when parameters are selected with ET or T and ET data, respectively. Additionally, using crop type parameters results in similar levels of simulated ET as using site-specific parameters. Different irrigation schemes create high amounts of uncertainty and highlight the need for accurate estimates of irrigation when performing water budget studies.

Sign in / Sign up

Export Citation Format

Share Document