Ajuste linear para preenchimento de falhas em dados de cotas do rio Madeira, em Humaitá-AM

Abstract

Hydrological studies are constantly improving, aiming to establish more and more real-world phenomena. In this sense, statistical hydrology comes in the field of water resources linked to studies and hydrological engineering projects that require theories for a probabilistic solution of a given phenomenon. One such study is the hydrological series, which include collections at regular intervals over several years of record. In this type of analysis the most important feature is the dependence of neighborhood values where the order of the data becomes crucial. Thus, the objective of this work was to perform a statistical approach to the historical series of Madeira River quota data from the ANA station in HTA. Fluviometric data were used from the ANA station hydrological series of Madeira river basins from 1968 to 2018. A correlation was made between the PVH station and linear HTA for the filling of faults in the series in the years 1990 and 2013 with the use of simple linear regression. Then, through the analysis of the complete series, the cotagram was generated and in the interval from 1968 to 1978 and the trend of the series in that time interval where it was positive by the analysis of the signal (+) in the equation of the line was verified. The correlation between the PVH and HTA season showed R2 = 0.97 for the year 1989 used to fill the year 1999 R2 = 0.99 for the year 2012 used to fill the 2013 failures in HTA. Thus, it was verified that the fluviometric station of HTA receives the corresponding to 98% of the total of water of the Madeira river that passes through PVH. And plotting the graphs of quotas every 10 years from 1972 on the PVH and HTA stations, it was possible to observe that they almost mirrored a difference in the value of the quotas that are deduced to be due to the RN or TDH.