Predicting runoff in ungauged UK catchments

Abstract

[1] A new approach to regionalization of conceptual rainfall-runoff models is presented on the basis of ensemble modeling and model averaging. It is argued that in principle, this approach represents an improvement on the established procedure of regressing parameter values against numeric catchment descriptors. Using daily data from 127 catchments in the United Kingdom, alternative schemes for defining prior and posterior likelihoods of candidate models are tested in terms of accuracy of ungauged catchment predictions. A probability distributed model structure is used, and alternative parameter sets are identified using data from each of a number of gauged catchments. Using the models of the 10 gauged catchments most similar to the ungauged catchment provides generally the best results and performs significantly better than the regression method, especially for predicting low flows. The ensemble of candidate models provides an indication of uncertainty in ungauged catchment predictions, although this is not a robust estimate of possible flow ranges, and frequently fails to encompass flow peaks. Options for developing the new method to resolve these problems are discussed.

Conditioning rainfall-runoff model parameters for ungauged catchments and land management impacts analysis

“…A probability distributed soil moisture (PDM) model together with two parallel linear routing stores (Fig. 4) is selected, as this is perceived to be widely applicable in the UK (Calver et al, 2005;Lamb and Kay, 2004;Lee et al, 2006). This model has five parameters (units are listed in Table 4): c max is the maximum soil water storage capacity within the element, b is a shape parameter defining the storage capacity distribution, α is a parameter defining the proportion of quick runoff, and k f and k s are routing store residence times.…”

“…The variance of BFI HOST is specified for each class, representing spatial and temporal variabilities within classes. Various other researchers have found that BFI HOST is the catchment characteristic of principal importance in the UK and alone contains significant information about rainfall-runoff model parameters (Lamb and Kay, 2004;Lee et al, 2006;Young, 2006). In some cases, other catchment properties or HOST outputs have been found to be more important.…”

“…The core of most regionalisation approaches is an estimation of either the dependence of the model parameter on physical catchment characteristics (e.g. catchment area, steepness, soil permeability) (Lamb and Kay, 2004;Lee et al, 2006;McIntyre et al, 2005;Young, 2006); or the dependence of response indices (e.g. mean annual discharge, daily discharge standard deviation) using physical characteristics (Bardossy, 2007;Yadav et al, 2007), so that the indices can then be used to restrict the parameter space.…”

Bayesian conditioning of a rainfall-runoff model for predicting flows in ungauged catchments and under land use changes

“…HOST cannot be presumed to hold [e.g., Wagener and McIntyre, 2005;Lee et al, 2006]. The initial soil storage (at the start of May 1980) was assumed equal to C max , and the subsequent month was neglected when assessing performance to reduce sensitivity to this assumption.…”

Short-Term Forecasting of Water Yield from Forested Catchments after Bushfire: A Case Study from Southeast Australia

“…The most commonly applied data-driven methods in short-term estimating and forecasting of catchment water yield are the parametric regressions such as the multivariate regressions [15][16][17], nonparametric regressions such as the K-nearest neighbor method [18][19][20], symbolic regressions such as genetic programming [21][22][23], and artificial intelligence based methods such as neural networks [24][25][26]. Hydrological processes contain non-linearities that are commonly modeled with data-driven techniques as an alternative to linear regression methods.…”