HOMOGENEITY TESTS BASED UPON GUMBEL DISTRIBUTION AND A CRITICAL-APPRAISAL OF DALRYMPLE TEST

Homogeneity tests are an important component of many regional flood fr equency analysis methods, particularly index flood methods. The index method was first suggested by Dalrymple, who also proposed a homogenei ty test. Dalrymple's test has been widely used in hydrologic practice for more then 30 years. This paper analyzes the relative performance o f Dalrymple's test, a normalized quantile test based upon L-moment par ameter estimation (X-10 test), and a method of moment Cv test (MoM-Cv test). Dalrymple's original test is shown to be in error and a correct ed version is developed. A Monte Carlo study compares the power of the se three tests, after applying correction factors to the test statisti cs to achieve a 5% Type I error. The L-moment X-10 test was always mor e powerful then Dalrymple's test or the MoM-Cv test. Moreover, the X-1 0 test needs much less correction to achieve the specified Type I erro r. An analytical analysis of the power of the test is consistent with the Monte Carlo results. The superior performance of the X-10 test is due to its L-moment estimation procedure, which yields a smaller sampl ing variance and a sampling distribution closer to normality than clas sical product moments. The relationship of the X-10 test with other L- moment based tests proposed by other workers is discussed. The origina l version of Dalrymple's test should not be used.