-

In probability distributions within hydrology, different methods are used regarding their application. The most prevailing has been central moment and with the use of computers, maximum likelihood method is used, too. The use of
L-moment has been adopted now only for a few years in the world and its application is in primary stages in our country. This research was carried out to make clear its use and make a comparison of it with previous common methods. In order to investigate the suitable probaility distributions for maximum, mean and minimum annual discharges, with the use of L-moment method, from nearly 70 existing hydrometric stations in the area, about 20 were selected for maximum and peak discharges and 25 hydrometric stations for mean and
minimum annual discharge. After selection of suitable stations and evaluation of
dishcarges from selected stations, 30 year statistic cycle was selected from
1968-69 water year until 1997-98 as a common statistic term. Missing data in the selected stations was accounted for by station regression method with the use of Minitab software program. The values for residual sum of squares were calculated by ordinary moment method for 5 used distributions in the study with
the use of HYF A software program. A comparison was made with the values of residual sum of squares obtained through L-moment method. With due attention to the lowest residual sum of squares for each distribution, the suitable distribution was chosen for each station by two methods of ordinary and linear moment. According to the results of this research regarding annually maximum and annually peak discharges, for most stations, Gumbel distribution, L¬momoent method, LN3 distribution, and ordinary moment methods were found fit. Regarding annually average and annually minimum discharge Gumbel distribution and L-moment methods were found to be suitable.