Preliminary Analysis of Hydrologic Data

Abstract

In Chap. 1, hydrologic processes and related variables are introduced as being inherently random, as a result of the complex interaction of multiple interdependent factors that can possibly act upon them. To start dealing with this randomness, the first task hydrologists usually face is to gather data related to the hydrologic variable of interest, to organize them in a sample, and to describe and explore their variability in some meaningful way. Such a preliminary analysis of sample data, looking for evidence of their empirical patterns of variability, is an application of descriptive methods of statistics to hydrology and is the focus of this chapter. The next stages of Statistical Hydrology, of setting up and fitting mathematical models that synthesize the variability pattern of the population, from which the sample data were drawn, are applications of probability theory and inferential statistics to hydrology, and covered in later chapters. Preliminary analysis of a hydrologic data set includes a collection of methods and techniques which can be separated into three groups: (1) graphical representation; (2) numerical summaries and descriptive statistics; and (3) exploratory methods. As a complement to the preliminary analysis of a sample of a single hydrologic variable, a brief discussion on the possible association between two hydrologic variables is provided in Sect. 2.4 of this chapter.