Search

A comparison of six procedures for estimating the linear trend parameter β in annual maximum 1 day river flows at five sites in southern Brazil showed marked differences between, on the one hand, estimates obtained by incorporating trend into the generalized extreme value (GEV) location parameter with all parameters estimated by maximum likelihood (ML) and on the other hand, estimates found by least squares, trend removal prior to fitting the GEV by ML, boot-strap sampling, and Theil-Sen estimation. ML estimates of trend were considerably smaller than those given by all other procedures. The same was true where trend had been incorporated into the Gumbel location parameter. Where 95% confidence intervals were calculated for the 'true' trend β by different procedures, some confidence intervals bracketed zero (indicating that the trend was not 'significant' at the 5% level), but there was no consistency between results from different procedures; Theil-Sen confidence intervals always bracketed zero, confidence intervals given by detrending never did. It is concluded that not only do different estimation procedures give different measures of trend uncertainty, as reported elsewhere, but the estimated trends themselves may differ, and the paper suggests an explanation why this may occur. Some philosophical issues relating to estimation of trend in climatological and hydrological extremes are discussed, and it is concluded that selection of a method to estimate trend must depend on context. [ABSTRACT FROM AUTHOR]

The paper compares forecasts of mean monthly water levels up to six months ahead at Ladário, on the Upper Paraguay River, Brazil, estimated from two long-range dependence models. In one of them, the marked seasonal cycle was removed and a fractionally differenced model was fitted to the transformed series. In the other, a seasonal fractionally differenced model was fitted to water levels without transformation. Forecasts from both models for periods up to six months ahead were compared with forecasts given by simpler "short-range dependence" Box-Jenkins models, one fitted to the transformed series, the other a seasonal autoregressive moving average (ARMA) model. Estimates of parameters in the four models (two "long-range dependence", two "short-range dependence") were updated at six-monthly intervals over a 20 year period, and forecasts were compared using root mean square errors (rmse) between water-level forecasts and observed levels. As judged by rmse, performances of the two long-range dependence models, and of the ARMA (1,1) short-range dependence model, were very similar; all three out-performed the seasonal short-range dependence ARMA model. There was evidence that all models performed better during recession periods, than on the hydrograph rising limb. [ABSTRACT FROM AUTHOR]

The article presents the study which explores the relationship between suspended sediment concentration (SSC) and turbidity for a small rural catchment in Brazil. It evaluates two calibration methods by comparing the SSC estimates obtained from the calibrated turbidity readings with direct measurements using the soil taxonomy system of the U.S. Department of Agriculture. The first method is derived by relating the turbidity reading to simultaneous measurements of concentration obtained from SSC samples collected from the vicinity of the turbidity probe during flood events, while the second is based on the readings obtained from the turbidity meter when the probe immersed in samples using soils collected from the catchment. The study shows the efficiency of the first method over the second.