Date of Award

1993

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Civil and Environmental Engineering

First Advisor

Vijay P. Singh

Second Advisor

Donald D. Adrian

Abstract

A new procedure to compile and enhance hourly data has been proposed. A total of 92 rain gauges were used to generate 25 synthesized stations with long periods of record. Annual maximum series for various rainfall durations from the 25 synthesized stations were used for statistical analysis. Temporal and spatial characteristics of annual maximum precipitation are investigated and a new procedure for identification of climatologically homogeneous regions is proposed. This new procedure is based on statistical parameters of precipitation data and understanding of atmospheric processes. The mean annual precipitation, recorded extreme precipitations, geographical locations, and synoptic weather patterns were used to identify three climatologically homogeneous regions. The Generalized Extreme Value (GEV) was found to best represent the underlying distribution for the precipitation data by the L-moment ratios. The proposed regional rainfall frequency analysis uses the GEV as the base distribution with L-moments and indexed probability weighted moments (PWM) for parameter estimation. The proposed regional procedure performs significantly better than the existing procedures such as Technical Paper 40 and the newly developed Louisiana rainfall maps based on the descriptive performance indices. The relative root mean square error (RMSE) and relative bias (BIAS) values were reduced substantially when the proposed regional method was used. Methodologies developed in this study can be applied to other regional studies. The advantage of this regional technique lies in its ability to dampen the sampling variability encountered with single-site station records. The application of this method is also easily extended to obtain rainfall quantiles at ungauged sites.

Pages

201

DOI

10.31390/gradschool_disstheses.5586

Download

Share

COinS