Estimativa da precipitação máxima provável para o estado do Paraná

Abstract

The maximum rainfall study has a great importance for society, once it is possible to find Maximum Probable Precipitation (PMP) values that are used in the designing process of hydraulic works, such as reservoirs, dams, hydroelectric plants, etc. PMP can be estimated in several different ways, and because of its practicality, statistical methods are widely used around the world. One of the procedures recommended by the World Meteorological Organization (WMO) is the statistical one proposed by Hershfield, which along with statistical parameters, uses a frequency factor (𝐾𝑚) while determining the PMP and considers that all samples have a single outlier, which, from a statistical point of view, cannot be considered true, since each data set has its own particularities. Therefore, some researchers suggest a modification in the traditional process: the use of an outlier identifier test (IQR). The present work selected 65 rainfall stations in the State of Paraná with at least 50 years record and after applying the IQR test, 80% of the stations presented an outlier amount other than 1. Furthermore, the results obtained by the two methods, the traditional and the modified, were compared and the differences reached 57%. The 𝐾𝑚 was also determined in two ways, one by the Hershfield graph and the other based on the envelope curve generated for the State of Paraná, the comparison between the frequency factors showed that the 𝐾𝑚 generated by the regional enveloping was at least 50% lower in all stations studied. In addition, the PMP values were spatially analyzed, through maps, highlighting that PMPs tend to follow the pattern observed in the distribution of the state's climate classification and, in general, higher PMPs appear in the eastern and western extremities and lower ones appear more frequently in the northern and southern regions of the state. All things considered, it is restated that the modified method presented safer results, consistent with the Parana’s reality, and it can, with no doubt, be used in the design of hydraulic structures.