Uncertainty Analysis of Intensity-Duration-Frequency Curves of Babylon City.

A substantial subject for the hydrological cycle, engineers must be able to determine the amount of rainfall in order to design structures and dams that deal with collection, transport and storage runoff. The study provides information on the amount of rainfall (measured in millimeters) that occurred annually from 1991 to 2021 in a single meteorological station in Iraq, specifically Babylon. The distributions of observed frequency are described, and the study attempts to fit three theoretical distributions (Gamma, Log Normal, and Normal) to the data. The distributions and tests chosen for the study play crucial roles in statistical analysis. The Normal distribution, a fundamental pattern in statistics, helps identify common trends in data and is a key part of quality control processes. It is often taught early in statistical education due to its significance in understanding natural variations and environmental factors. The Gamma distribution is widely used in statistics to model time durations for tasks or events. The Log-Normal distribution describes a variable whose logarithm follows a normal distribution. Chi-square test is a statistical method to compare expected and observed outcomes across different categories, commonly employed in survey analysis. The Kolmogorov-Smirnov test compares distributions of two independent samples, useful for evaluating how well a theoretical distribution fits actual data. Researchers use this test under similar conditions as the Chi-square test. The Kolmogorov-Smirnov and Chi-Square and Anderson-Darling indices where, the calculated value of nonparametric test (statistic value) is extract and compared with the tabular value (critical value) based on the significant (α) and degrees of freedom, if the calculated value is less than the tabular value, we accept the null hypothesis (good fit); otherwise, we do not accept the null hypothesis (poor fit). They are used to compare the theoretical distributions to the observed data. The study then focuses on the Intensity-Duration-Frequency (IDF) curves for extreme rainfall values, with durations of 15, 30, and 60 minutes. The results reveal that rainfall intensity decreases as the duration of the storm increases. Additionally, rainfall of a specific duration shows higher intensity if the return period is greater. Gumbel's extreme value distribution, Normal distribution, and Log Normal distribution are used to fit rainfall data with 5, 10, 15, and 50-year return periods. The Excel Software showed that the lognormal, normal, and Gamma probability distributions were the best fit for the data group for all durations. The software estimated the intensities of precipitation for return periods of 5, 10, 15, and 50 years, and the Log Normal distribution presented as the good agreement. [ABSTRACT FROM AUTHOR]