Modified K-NN Model for Stochastic Streamflow Simulation

Abstract

This paper presents a lag-1 modified K-nearest neighbor ͑K-NN͒ approach for stochastic streamflow simulation. The simulation at any time t given the value at the time t − 1 involves two steps: ͑1͒ obtaining the conditional mean from a local polynomial fitted to the historical values of time t and t − 1, and ͑2͒ then resampling ͑i.e., bootstrapping͒ a residual at one of the historical observations and adding it to the conditional mean. The residuals are resampled using a probability metric that gives more weight to the nearest neighbor and less to the farthest. The "residual resampling" step is the modification to the traditional K-NN time-series bootstrap approach, which enables the generation of values not seen in the historical record. This model is applied to monthly streamflow at the Lees Ferry stream gauge on the Colorado River and is compared to both a parametric periodic autoregressive and a nonparametric index sequential method for streamflow generation, each widely used in practice. The modified K-NN approach is found to exhibit better performance in terms of capturing the features present in the data.

Combining regional moist static energy and ENSO for forecasting of early and late season Indian monsoon rainfall and its extremes

“…Many local estimation procedures have been theorized and tested including splines, kernel-based estimation [e.g., Bowman and Azzalini, 1997], and local polynomials [Loader, 1999]. Local polynomials are easier to understand and implement and have been widely used in a variety of applications such as streamflow and salinity modeling [Prairie et al, 2006], streamflow forecasting [Bracken et al, 2010], water quality modeling [Towler et al, 2009[Towler et al, , 2010, and others.…”

First physical record of Allen's lappet-browed bat (Idionycteris phyllotis) in Colorado

“…So-called "weather generators" (see Gangopadhyay et al 2005) use historical weather data that are re-sampled according to the conditions projected by the climate model. The same resampling technique can be applied to historical streamflow data to provide future hydrologic sequences that are consistent with both the historical variability (Prairie et al 2006) and the climate model average projections.…”

Changes in summer monsoon rainfall in the Upper Chao Phraya River Basin, Thailand

“…seasonal rainfall (Prairie et al 2006, Podestá et al 2009, Eum et al 2010). The modified k-nn model locally captures the relationships between dependent (y) and independent variables (x 1 , x 2 , x 3 ,…,x m ) at a given point by a number of its neighbors.…”