Random Functions and Hydrology

a₁ algorithm approximation assumed autocorrelation function autocovariance basin behavior Box and Jenkins broken-line process Chapter computed correlation function correlation structure corresponding covariance function defined density discharge discrete distribution drift equation example expected value Figure finite forecasting frequency Gaussian Hurst hydrologic implies input K₁ K₂ Kalman filter Kriging lag-one correlation linear system M₁ matrix mean area mean square error Mejía multivariate noise nonlinear normal number of stations observations obtained optimal Orinoco River p₁ parameter estimation partial autocorrelation prediction procedure rainfall random field random process random variables reservoir residuals River Rodríguez-Iturbe sampling seasonal semivariogram sequence shown in Fig simulation skewness solution spatial spectral spectrum standard deviation stationary stationary process statistics stochastic storm streamflow tion u₂ variance reduction variogram vector Wadi Halfa weights X₁ X₁(t Z₁ zero mean Σ Σ