Applicability of a Nu-Support Vector Regression Model for the Completion of Missing Data in Hydrological Time Series

Abstract

This paper analyzes the potential of a nu-support vector regression (nu-SVR) model for the reconstruction of missing data of hydrological time series from a sensor network. Sensor networks are currently experiencing rapid growth of applications in experimental research and monitoring and provide an opportunity to study the dynamics of hydrological processes in previously ungauged or remote areas. Due to physical vulnerability or limited maintenance, networks are prone to data outages, which can devaluate the unique data sources. This paper analyzes the potential of a nu-SVR model to simulate water levels in a network of sensors in four nested experimental catchments in a mid-latitude montane environment. The model was applied to a range of typical runoff situations, including a single event storm, multi-peak flood event, snowmelt, rain on snow and a low flow period. The simulations based on daily values proved the high efficiency of the nu-SVR modeling approach to simulate the hydrological processes in a network of monitoring stations. The model proved its ability to reliably reconstruct and simulate typical runoff situations, including complex events, such as rain on snow or flooding from recurrent regional rain. The worst model performance was observed at low flow periods and for single peak flows, especially in the high-altitude catchments.

Double Tensor-Decomposition for SCADA Data Completion in Water Networks

“…As we can see, it collects data corresponding to a whole number of weeks. The third operation consists of constructing a simplified version of the tensor, named χ 288×7×n w (1) in Figure 1 by using a tensor decomposition. In Figure 1a it was represented the Tucker decomposition in Figure 1b the CP.…”

“…In practice, when processing this amount of information, the problem of incomplete or missing data has to be addressed. The management of data from water networks [1] or from hydrological resources [2][3][4] is no exception. The problem of data loss is especially challenging when it occurs in long bursts of consecutive values.…”

High-Resolution Discharge Forecasting for Snowmelt and Rainfall Mixed Events

“…Such events should be of major consideration especially in urban catchments, where they can promote flooding [1]. The hydrological processes related snowmelt and rainfall mixed events, mostly concerning the rain-on-snow phenomena, were simulated or analysed recently in a number of studies [2][3][4][5]. However, snowmelt and rainfall mixed events received relatively little attention in high-resolution discharge forecasting, especially in study areas where snow processes tend to be disregarded.…”

Application of a Hybrid Interpolation Method Based on Support Vector Machine in the Precipitation Spatial Interpolation of Basins

“…Later, with the introduction of insensitive loss functions, SVM also showed good learning performance in regression estimation of nonlinear systems. In recent years, SVM has been widely applied to the fields of hydrology [4][5][6][7], meteorology [8,9] and water environment [10], etc., as well as research on spatial interpolation. In spatial interpolation with SVM, by learning of known samples, non-linear relationships between data and properties are approached to realize the forecast of unknown samples.…”