Features of the application of the method of quantile regression in problems of hydroecology

Vladislav Artemenko, Мaster of Ecology,Ukrainian Hydrometeorological Institute, State Service on Emergencies of Ukraine and National Academy of Science  of Ukraine, Hydrochemical Research, Scientific Enployee, https://orcid.org/0000-0003-0536-5415

Volodymyr Petrovych, National Transport University. Senior Researcher, Professor of the Transportation Construction and Property Management Department, https://orcid.org/0000-0003-0422-2535

 

Abstract: Such perspective method of the analysis of the natural time series is considered in article as method of quantile regression. They are discussed imperfection existing methods of the analysis of the natural processes (least square method and the method of linear robust regression). It is shown that transition towards quantile regression allows greatly to raise efficiency of the investigations of the natural time series. Main gоal of the work consists in practical applications of the quantile regression method for decision of the different problemsof the hydroecology.

Hydrochemical time series were considered in article. By means of quantile regressin method was investigated behavior of the dissolved oxygen in water depending on temperature of river water and depending on discharge of river water (for quantiles of order 0.05; 0,50 und 0,95). When performing the investigations more perfect method of quantile regression was designed such as method piecewisf quantile regression (with uce polynomial degree 1; 2; 3).

Numenical experiments when use the natural time series have shown greater advantage of the designed method of piecewise quantile regression in contrast with classical method of quantile regression.

 

Article language: Ukrainian

 

Referenses:

  1. Artemenko V.A. Sezonna dynamika biohennykh rechovyn krupnykh vodnykh ob’yektiv. V.A. Artemenko, V.V. Avtomobilʹni dorohy i dorozhnye budivnytstvo, vyp. 104. K.: Vyd-vo Natsion. transp. un-tu. 2018.  С. 31-43.
  2. Khardle V. Prikladnaya neparametricheskaya regressiya. : Mir. 1993. 349 с.
  3. Koenker R. Quantile Regression. Cambridge University Press. 2005. 366 p.
  4. Hao L. Quantile Regression. L. Hao, D. Naiman. Sage Publications Inc. 2007. 137 p.

 

Open Access: http://publications.ntu.edu.ua/avtodorogi_i_stroitelstvo/110/90.pdf

 

Online publication date: 25.02.2021

 

Print date: 01.02.2021

 

Search