Empirical or statistical methods have been introduced into meteorology and oceanography in four distinct stages: 1) linear regression (and correlation), 2) principal component analysis (PCA), 3) canonical correlation analysis, recently 4) neural network (NN) models.Despite the great popularity of NN models many fields, there are three obstacles to adapting method meteorology-oceanography, especially large-scale, low-frequency studies: (a) nonlinear instability with short data records, (b)...

Methods in multivariate statistical analysis are essential for working with large amounts of geophysical data, data from observational arrays, satellites, or numerical model output. In classical analysis, there is a hierarchy methods, starting linear regression at the base, followed by principal component (PCA) and finally canonical correlation (CCA). A time series method, singular spectrum (SSA), has been fruitful extension PCA technique. The common drawback these methods that only...

SUMMARY Forecasting the maize yield of China's Jilin province from 1962 to 2004, with climate conditions and fertilizer as predictors, was investigated using multiple linear regression (MLR) non-linear artificial neural network (ANN) models. Yield set be a function precipitation July August, in September amount used. Fertilizer emerged dominant predictor non-linearly related ANN model. Given difficulty acquiring data for maize, current study also tested previous year's place data. Forecast...

Nonlinear principal component analysis (NLPCA) can be performed by a neural network model which nonlinearly generalizes the classical (PCA) method.The presence of local minima in cost function renders NLPCA somewhat unstable, as optimizations started from different initial parameters often converge to minima.Regularization adding weight penalty terms is shown improve stability NLPCA.With linear approach, there dichotomy between PCA and rotated methods, it generally impossible have solution...

The authors constructed neural network models to forecast the sea surface temperature anomalies (SSTA) for three regions: Niño 4, 3.5, and 3, representing western-central, central, eastern-central parts of equatorial Pacific Ocean, respectively. inputs were extended empirical orthogonal functions (EEOF) level pressure (SLP) field that covered tropical Indian Oceans evolved a duration 1 yr. EEOFs greatly reduced size networks from those authors’ earlier papers using EOFs. 4 region appeared be...

The effects of viscosity and finite- differencing on free Kelvin waves in numerical models (which employ the Arakawa B- or C-grid difference schemes) are investigated using f-plane shallow-water equations with offshore finite-difference grids, (assuming alongshore geostrophy). Three nondimensional parameters arise: Δ [=(offshore grid spacing)/(Rossby radius)], ε characterizes lateral viscous effect α combined vertical effect. This study is more relevant to baroclinic which tend suffer poor...

Abstract Among the statistical methods used for seasonal climate prediction, canonical correlation analysis (CCA), a more sophisticated version of linear regression (LR) method, is well established. Recently, neural networks (NN) have been applied to prediction. Unlike CCA and LR, NN nonlinear which leads question whether nonlinearity brings any extra prediction skill. In this study, an objective comparison between three (CCA, NN) in predicting equatorial Pacific sea surface temperatures (in...

Two nonlinear regression methods, Bayesian neural network (BNN) and support vector (SVR), linear (LR), were used to forecast the tropical Pacific sea surface temperature (SST) anomalies at lead times ranging from 3 15 months, using level pressure (SLP) SST as predictors. Datasets for 1950–2005 1980–2005 studied, with latter period having warm water volume (WWV) above <mml:math...

Abstract The output from a coupled general circulation model (CGCM) is used to develop evidence showing that the tropical Pacific decadal oscillation can be driven by an interaction between El Niño–Southern Oscillation (ENSO) and slowly varying mean background climate state. analysis verifies changes in states are attributed largely ENSO statistics through nonlinear rectification. This seen because time evolutions of first principal component (PCA) mode decadal-varying SST thermocline depth...

Abstract. Estimates of surface snow water equivalent (SWE) in mixed alpine environments with seasonal melts are particularly difficult areas high vegetation density, topographic relief, and accumulations. These three confounding factors dominate much the province British Columbia (BC), Canada. An artificial neural network (ANN) was created using as predictors six gridded SWE products previously evaluated for BC. Relevant spatiotemporal covariates were also included predictors, observations...