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 II: REGRESSION ANALYSIS
 
 This article discusses the role of regression analysis of agricultural data to appreciate how dependent variables, like crop production, vary with different magnitudes of influential factors, like soil salinity or depth of water table. Various techniques of regression analysis are introduced including the regression in segments to detect up to which point a growth factor affects the crop yield and to define the range where there is no influence.


TABLE OF CONTENTS
 
 
6.1 Introduction
 
6.5 Regression Analysis
 
       6.5.1 Introduction
       6.5.2 The Ratio Method
       6.5.3 Regression of y upon x
       6.5.4 Linear Two-way Regression
       6.5.5 Segmented Linear Regression
 
6.6 Screening of Time Series
 
       6.6.1 Time Stability versus Time Trend                                        
       6.6.2 Periodicity of Time Series
       6.6.3 Extrapolation of Time Series
       6.6.4 Missing and Incorrect Data
 
6.7 References
 		Category
selection:
 
 
Software
& models
 
 
Articles
& manuals
 
 
Reports
case studies
 
 
 
FAQ's
& papers
 
 
Home
page 		 


segmented regression Regression analysis serves the purpose to detect relations between dependent and independent variables.

A regression equation gives a regression coefficient, or slope, and a constant.
Segmented regression is done to divide the domain into two parts with different properties.
The figure shows the relation between crop yield and soil salinity. In the first part, the salt has no influence on crop production, but in the second part the salt content becomes so high that the yield declines under a certain slope.