FREQUENCY AND REGRESSION ANALYSIS OF HYDROLOGIC DATA

R.J. Oosterbaan

On web site www.waterlog.info

Copy of part of chapter 6 in: H.P.Ritzema (Ed.), Drainage Principles and Applications,
ILRI Publication 16, second revised edition, 1994, Wageningen, The Netherlands

Download or view the chapter from the FAQ's page

 FREQUENCY AND REGRESSION ANALYSIS PART I: FREQUENCY ANALYSIS   This article discusses the role of frequency analysis of agricultural data to appreciate the occurrence of climatic phenomena, like rainfall, and soil properties. It deals with different techniques of frequency analysis. In addition it introduces the principles of probability distribution fitting with emphasis on the normal, Gumbel, and exponential distributions. TABLE OF CONTENTS       6.1 Introduction   6.2 Frequency Analysis          6.2.1 Introduction        6.2.2 Frequency Analysis by Intervals        6.2.3 Frequency Analysis by Ranking of Data        6.2.4 Recurrence Predictions and Return Periods                                   6.2.5 Confidence Analysis   6.3 Frequency-Duration Analysis          6.3.1 Introduction        6.3.2 Duration Analysis        6.3.3 Depth-Duration-Frequency Relations   6.4 Theoretical Frequency Distributions          6.4.1 Introduction        6.4.2 Principles of Distribution Fitting        6.4.3 The Normal Distribution        6.4.4 The Gumbel Distribution        6.4.5 The Exponential Distribution        6.4.6 A Comparison of the Distributions   6.7 References Category selection:       Software & models     Articles & manuals     Reports case studies       FAQ's & papers    Home page

 Analysis of data in famers' fields encompasses the regression of crop production on important growth factors of the soil. The figure shows a probability distribution fitted to rainfall data as the climatic factors are important for crop growth. While making assessments about the state of the agriculture it may be usefull to apply models to the relations between water table and hydraulic conductivity. Also factors like soil salinity, irrigation and drain discharge can be modelled.
 The normal probabilty distribution plays an important role in frequency analysis. Freqency analysis can be done by intrvals or by rakning the data. The aim of frequency distributions is to provide recurrence predictions and return periods. The Gumbel and exponential probability distributions are useful for evaluating extreme values.