scholarly journals Paleoflood investigations to improve peak-streamflow regional-regression equations for natural streamflow in eastern Colorado, 2015

2016 ◽  
Author(s):  
Michael S. Kohn ◽  
Michael R. Stevens ◽  
Tessa M. Harden ◽  
Jeanne E. Godaire ◽  
Ralph E. Klinger ◽  
...  
Keyword(s):  
Regression Equations ◽  
Regional Regression ◽  
Peak Streamflow

scholarly journals Topological and canonical kriging for design-flood prediction in ungauged catchments: an improvement over a traditional regional regression approach?

2012 ◽  
Vol 9 (10) ◽  
pp. 12193-12226 ◽  
Author(s):  
S. A. Archfield ◽  
A. Pugliese ◽  
A. Castellarin ◽  
J. O. Skøien ◽  
J. E. Kiang
Keyword(s):  
United States ◽  
Spatial Interpolation ◽  
The United States ◽  
Low Flow ◽  
Regression Equations ◽  
Ungauged Catchments ◽  
Interpolation Methods ◽  
Regional Regression ◽  
Flood Quantiles ◽  
Spatial Interpolation Methods

Abstract. In the United States, estimation of flood frequency quantiles at ungaged locations has been largely based on regional regression techniques that relate measurable catchment descriptors to flood quantiles. More recently, spatial interpolation techniques of point data have been shown to be effective for predicting streamflow statistics (i.e. flood flows and low-flow indices) in ungauged catchments. Literature reports successful applications of two techniques, Canonical kriging, CK, (or physiographical-space based interpolation, PSBI) and Topological kriging, TK, (or Top-kriging). CK performs the spatial interpolation of the streamflow statistic of interest in the two-dimensional space of catchment descriptors. TK predicts the streamflow statistic along river networks taking both the catchment area and nested nature of catchments into account. It is of interest to understand how these spatial interpolation methods compare with generalized-least squares (GLS) regression, one of the most common approaches to estimate flood quantiles at ungauged locations. By means of a leave-one-out cross validation procedure, the performance of CK and TK was compared to GLS regression equations developed for the prediction of 10-, 50-, 100- and 500-yr floods for 61 streamgauges in the Southeast United States. TK substantially outperforms GLS and CK for the study area, particularly for large catchments. The performance of TK over GLS highlights an important distinction between the treatment of spatial correlation when using regression-based versus spatial interpolation methods to estimate flood quantiles at ungauged locations. The analysis also shows that coupling TK with CK slightly improves the performance of TK; however, the improvement is marginal when compared to the improvement in performance over GLS.

Adjustment of Regional Regression Equations for Urban Storm-Runoff Quality Using At-Site Data

1996 ◽  
Vol 1523 (1) ◽  
pp. 141-146 ◽  
Author(s):  
C. Shane Barks
Keyword(s):  
Predictive Accuracy ◽  
Storm Runoff ◽  
Prediction Errors ◽  
Regression Equations ◽  
Local Regression ◽  
Water Quality Prediction ◽  
Regional Regression ◽  
Regional Prediction ◽  
Predicted Values ◽  
Urban Storm

Regional regression equations have been developed to estimate urban storm-runoff loads and mean concentrations using a national data base. Four statistical methods using at-site data to adjust the regional equation predictions were developed to provide better local estimates. The four adjustment procedures are a single-factor adjustment, a regression of the observed data against the predicted values, a regression of the observed values against the predicted values and additional local independent variables, and a weighted combination of a local regression with the regional prediction. Data collected at five representative storm-runoff sites during 22 storms in Little Rock, Arkansas, were used to verify, and, when appropriate, adjust the regional regression equation predictions. Comparison of observed values of storm-runoff loads and mean concentrations to the predicted values from the regional regression equations for nine constituents (chemical oxygen demand, suspended solids, total nitrogen as N, total ammonia plus organic nitrogen as N, total phosphorus as P, dissolved phosphorus as P, total recoverable copper, total recoverable lead, and total recoverable zinc) showed large prediction errors ranging from 63 percent to more than several thousand percent. Prediction errors for 6 of the 18 regional regression equations were less than 100 percent and could be considered reasonable for water-quality prediction equations. The regression adjustment procedure was used to adjust five of the regional equation predictions to improve the predictive accuracy. For seven of the regional equations the observed and the predicted values are not significantly correlated. Thus neither the unadjusted regional equations nor any of the adjustments were appropriate. The mean of the observed values was used as a simple estimator when the regional equation predictions and adjusted predictions were not appropriate.

Sign in / Sign up

Export Citation Format

Share Document