ABSTRACT

Water as an essential commodity plays a major role in national development, the availability of water through pipe borne to individual, communities, and the country at large will help in achieving the New Millennium Development Goal (MDG7), hence the constant study, analysis and forecast for  supply of water using past water supplied. The supply of pipe borne water from 2009 to 2015 was examined with the aim of developing a working model to forecast future supply of pipe borne water. All these were achieved using the Box-Jenkins methodology. The Box-Jenkins methodology is a statistical tool that is adequate for the forecasting of time series data. However, before the Box-Jenkins methodology, time plot was plotted to know the components that are present in the supply of pipe borne water in Ilorin, from the graph plotted it was noticed that irregularity and upward trend were the components present with the absent of seasonal and cyclical variations in the series. The series was not stationary but was made stationary after the first difference, Augmented Dickey Fuller Unit Root test was used to confirm the stationarity of the series. After stationarity was established, the model of the supply of pipe borne water was determined using the correlogram and it was suspected that the model was an Autoregressive Integrated Moving Average model. The order of the models and parameter of the models were confirmed from the information criterion computation and it was discovered with ARIMA(0,1,1), ARIMA(0,1,0) and ARIMA(1,1,0). The adequacy of the models was also checked and they were found to be adequate from a series of test like analysis of residual using qq-plot. The best model was picked using the AIC, BIC and HQIC, ARIMA(0,1,0) had the relatively lowest AIC, BIC and HQIC. Therefore, we concluded that our best model is ARIMA(0,1,0). Using the model developed, the supply of pipe borne water for a period of 24 months was forecasted and it was discovered that it follows a slight upward movement. The study concludes that though the forecast proves to be moving upward with time, continuous measures should be carried out so as to check for any decrease in the supply of water in the state.

TABLE OF CONTENTS   

                                                                                                                       PAGE

Title Page i           

Attestation         ii

Certification         iii

Dedication iv

Acknowledgement           v

Abstract             viii

CHAPTER ONE: INTRODUCTION 

1.1 Background to the Study 1               

1.2 Study Area 1          

1.3 Definition of Water 4                                                                                                     

1.4 Pipe Borne Water             5                                                                                                   

1.5 Pipe Borne Water in Kwara 7                                                                                          

1.6 Water Related Diseases   8                                                                                                         

1.7 Challenges of Pipe Borne Water  

1.8 Effect of Pipe Borne Water

1.9 Aim and Objectives 8                                                                                                                    

1.10 Source of Data

1.11 Justification

1.12 Organization of Study 9                                                                                              

CHAPTER TWO: LITERATURE REVIEW 

2.1 Conceptual Framework 10                                                               

2.2 Options for Water Problem Solution 11

CHAPTER THREE:  METHODOLOGY                                                                 

3.0 Time Series as a Stochastic Process 11                                     3.1 Introduction 12                                           

      Types of Time Series ` 12                                                          

      Importance of Time Series 21                                                    

3.4 Components of Time Series 22                                                                                      

3.5 Box-Jenkins Methodology 23                                                                                                                

3.5.1 Advantages of Box-Jenkins Methodology 23                                                                                       

3.5.2 Disadvantages of Box-Jenkins Methodology 24                                                        

3.5.3 Summary of Box-Jenkins Methodology 25                                         

3.6 Identification Procedure 25                                   

3.7 Autocorrelation Functions 25                                                 

3.8 The Partial Autocorrelation Function 26                                                                                   

3.9 Stationary and Non-Stationary Process 26        

3.10 Test for Stationary Series 27                                                                                  

3.11 Removal of Trend and Ensuring Stationarity 28                                              

3.12 Identifying a Tentetive Model   29     

2.19 Stationary and Non Stationary Processes   30                                         

2.20 Test for Stationary Series     30                                                                    

2.21 Removal of Trend and Ensuring Stationarity 31                                      

2.22 Identifying a Tentative Model     32                                                      

2.23 Autoregressive, AR (p) Process             32                                               

2.24 Moving Average, MA (q) Process     33                                                   

2.25 Autoregressive Moving Average ARMA (p, q) Process         34

2.26 Model and Model Order Determination       34                                      

2.27 Autoregressive Models   34                                                                              

2.28 Moving Average Model 35                                                                                     

2.29 Autoregressive Moving Average Model   35                                               

2.30 Information Criteria   35                                                                                            

2.31 Akaike Information Criterion (AIC) 36                                            

2.32 Bayesian Information Criterion (BIC) 36

2.33 Corrected Akaike Information Criterion (AICC)   37                                                     

2.34 Diagnostic Checking   37                                                                                            

2.35 Forecasting                     38                                                                                             

CHAPTER THREE: ANALYSIS AND INTERPRETATIONS OF 

RESULT                                               39

CHAPTER FOUR: CONCLUSIONS AND RECOMMENDATIONS            51

References             53