Jahir Uddin Chowdhury

Goodness‐of‐fit tests for regional generalized extreme value flood distributions

This paper develops critical values and formulas for computing several goodness-of-fit tests for the generalized extreme value (GEV) distribution. These tests can check if data available for a site are consistent with a regional GEV distribution, except for scale, or if the data are consistent with a GEV distribution with a regional value of the shape parameter κ. Three tests employ unbiased probability-weighted moment (PWM) estimators of the L moment coefficient of variation (L-CV), and coefficient of skewness (L-CS) using formulas for their variances in small samples. In a Monte Carlo power study the L-CV test was often more powerful than the Kolmogorov-Smirnov test at detecting L-CV inconsistencies. A test based upon L-CS generally has equal or greater power than the probability plot correlation test at detecting L-CS differences; both are poor at detecting thin-tailed alternatives. Finally, a new chi-square test based upon sample estimates of both the L-CV and L-CS, and their anticipated cross correlation, was much better than other tests at detecting departures from the assumed L-CV, L-CS, or both, particularly when the regional distribution was highly skewed.

A comparison of four distributions used in flood frequency analysis in Bangladesh

Abstract This paper discusses how lognormal, Gumbel and log-Pearson type 3 distributions were eliminated as candidates for modelling at-site annual maximum discharge series in Bangladesh. Among candidate distributions, the generalized extreme value distribution best represented the statistical characteristics of observed data. Comparisons of distributions were based upon the root mean square deviation in fit, the probability plot correlation coefficient and L-moment ratio diagrams. The last method assesses whether a candidate distribution is capable of simulating the site-to-site variation in statistical characteristics that is present in observed flood samples.

Models for Cyclone Shelter Planning in Bangladesh

ABSTRACT The coastal areas of Bangladesh are prone to disaster due to storm surge floods generated by tropical cyclones which form in the Bay of Bengal. A master plan, prepared at the request of the Government of Bangladesh, estimated that 2,500 new cyclone shelters are needed to provide adequate protection for the projected population in the year 2002. Since this will require an investment of 300 million USD (12 billion Bangladeshi Taka) which may not be possible in the near future, a means of identifying priority areas is needed. To this end, a mathematical programming model is developed. The model relies on input from hydrologic simulation models which determine the spatial distribution of flood depth, as well as from surveys and an infrastructure inventory which provide an estimate of population vulnerability. Using the mathematical programming model, a multi-objective analysis is conducted to explore some trade-offs involved in allocating shelters among a number of planning units. Results indicate th...

Comparison of Five Distributions for Suitability in Flood Frequency Analysis in Bangladesh

A study has been conducted to compare five distributions used in flood frequency analysis in Bangladesh. The distributions are Log-normal, Gumbel, Generalized Extreme Value, Pearson Type 3 and Log-pearson Type 3. Comparisons have been based upon at-site root mean square error in fit, at-site probability plot correlation coefficient, regional L-moment ratio diagrammme, regional distribution of L-coefficient of skewness and regional distribution of largest normalized discharges. The study indicates that the Generalized Extreme Value distribution represents the statistical behaviour of observed annual maximum discharges best. Pearson Type 3 and Gumbel distributions are not suitable for frequency analysis of discharge data in Bangladesh.