modification of reconnaissance drought index (rdi) based on the best method of evapotranspiration estimation and probability distribution function

reconnaissance drought index (rdi) is based on fitting a log-normal distribution to the ratio of precipitation to evapotranspiration (eto) values in selected periods. in this index value of eto were calculated based on mean temperature by thorenth-waite (th) method. th method, may underestimated eto values comparing to the actual in arid and semi arid regions. the log-normal distribution may not be fitted to the ratio of precipitation to eto values of some regions. in order to investigate the effects of these two limitations on drought situations' changes, meteorological parameters have been used during 50 years period at 8 synoptic stations of iran. in the first step, the values of rdi(th) for any stations during the mentioned time were calculated. then, eto values were calculated from best fitted empirical equation in any situation of lack of parameters. subsequently rdi(select) index were established. the kolmogorov–smirnov (ks) test is used to assess the goodness of fitting most appropriate distribution function to the ratio of precipitation to eto values. then, according to equi-probability transformation the values of rdi(th) were modified to *rdi(th). the occurrence of different classes of drought according to rdi(select) and/or *rdi(th) comparing to rdi(th) showed the elimination of any mentioned limitations may leads to changing the amount of occurrence of any drought classes in rdi(th). hence, the rdi(th) modified to *rdi(select) by estimating eto values from selected method and applying appropriate distribution function to the ratio of precipitation to eto values.