Journal of Hydrologic Engineering

This paper identifies connections between the time variability of the volume of the Great Salt Lake (GSL), Utah, and selected atmospheric circulation indices. The indices considered are the Southern Oscillation Index (SOI), the Pacific/North America (PNA) climatic pattern, and the Central North Pacific (CNP) index. We focus on interannual time scales. Low-frequency (interannual or interdecadal) relationships between the Great Salt Lake (GSL) volume change and atmospheric circulation indices are of importance because of their significance for the understanding and prediction of the GSL volume. We use Singular Spectral Analysis (SSA), Multichannel Singular Spectral Analysis (MSSA), and Multitaper Spectral Analysis [or Multitaper Method (MTM)], to identify persistent or nearly periodic patterns in time in each series. MSSA examines the joint modes of variability across the time series, while SSA decomposes each series into its component time patterns. MTM is used for identification of peaks and frequency band structure.

A new approach for estimating floodplain limits using continuous-simulation modeling is proposed. A hydraulic model with flood-routing capabilities is coupled with a continuous-simulation hydrologic model to compute river stages. Simulations are made using historical precipitation data to produce long, continuous flow records. Simulations are also made using precipitation data for extreme storms to gain additional information on overbank flooding. A new statistical approach is then directly applied to the simulated peak stages to estimate peak stage exceedence probabilities. To illustrate the benefits of this approach, floodplain limits for an Illinois stream are estimated with the new method and compared to estimates based on a design storm approach and conventional frequency analysis applied to the peak stages.

L-moment diagrams are constructed for annual minimum, average, and maximum streamflows at more than 1,455 river basins in the United States. Goodness-of-fit comparisons reveal that the generalized extreme value (GEV), three-parameter lognormal (LN3) and the log Pearson Type III (LP3) distributions provide good approximations to the distribution of annual maximum flood flows. These results are consistent with other L-moment studies. A World Meteorological Organization survey of 54 agencies in 28 countries reveals that the LN3 distribution is not a standard in any country, GEV is a standard in one country, and LP3 is a standard in seven countries. The time is ripe for agencies and countries to reevaluate their standards with respect to the choice of a suitable model for flood frequency analysis. L-moment diagrams also reveal that among numerous alternatives, the Pearson Type III (P3) distribution provides the best fit to both annual minimum and annual average streamflows.

Hydrologic modeling of the responses of two study watersheds to climate change is presented. The watersheds are the Upper Campbell and the Illecillewaet watersheds located in British Columbia. The first is a maritime watershed located on the eastern slopes of the Vancouver Island mountains; the second is located in the Selkirk Mountains in Eastern British Columbia. The Canadian Climate Centre General Circulation model has been used for the prediction of potential effects of climate change on meteorological parameters. In addition to the changes in the amounts of precipitation and temperature usually assumed in hydrologic climate change studies, other meteorological and climatic parameters also considered are the effect of climate on the spatial distribution of precipitation with elevation, as well as on cloud cover, glaciers, vegetation distribution, vegetation biomass production, and plant physiology. The results showed that the mean annual temperature in the two watersheds could increase by more than 3°C and the annual basinwide precipitation could increase by 7.5% in the Upper Campbell watershed and by about 17% in the Illecillewaet watershed. The higher temperatures changed some snowfall to rainfall and the extra precipitation was mainly in the form of rain. The increase of the CO ₂ concentration caused stomata closure that reduced evapotranspiration. This effect was compensated by increased biomass in the Upper Campbell watershed, but not in the Illecillewaet watershed. These changes produced higher flows in winter and smaller flows in summer. The largest change in the hydrograph shape was in the Illecillewaet watershed where the mean annual maximum daily flow decreased by about 13% and its frequency was reduced. On the other hand, the mean annual runoff increased by 21%. In contrast, although the shape of the simulated annual hydrograph of the Upper Campbell watershed was not affected, magnitude and frequency of the annual maximum precipitation increased. Also, the mean annual runoff in the Upper Campbell watershed increased by 7.5%. These results indicate that different management procedures may be needed to minimize the effects of climate change on the water resources of the two climatically different watersheds and the regions that they represent.

A new method of parameter estimation for any probability distribution used in flood frequency analysis is proposed as a modification to the conventional probability-weighted moments (PWMs) method. The PWMs can be estimated by computing either the nonexceedence or the exceedence probability of each element in the sample series using the distribution itself instead of a plotting position formula. The formulation and solution of the suggested method is given for the general extreme-values distribution. Experiments on 1,000 short-length synthetic series indicate that the proposed method predicts the 1,000-yr return-period events better than both the maximum-likelihood and the conventional PWMs methods.