the mathematical expressionsfor trend curves representingthese patterns and the methodsby which a trend curve can objectively be fitted to a givenseriesof historical traffic data (time series data) in order to producea projection into the future which is free from subjectivejudgement.involves the formulation of mathematicalrelationshipsbetween the traffic and the underlying factors. and to make these available to Contracting States. regional as well as global data. One technique.plannersof airports and route facilities and others actively engaged in practical forecasting work.econometricforecasting. In doing so. The work of the Organizationin the field of air traffic forecasting is governed by Assembly ResolutionAI6-22: A16-22: Air Traffic Forecastsand ForecastingMethods WHEREAS the Council. For such forecasts. where possible. initiatives to bring about an exchange of views among air transport forecasting bodies.This secondedition takes into account new material received by ICAO and experience gained during the intervening period. The Manual is addressed personnelof civil aviation to administrations. the Organization might take 4. Other techniquesare based on specific studies of individual sectorsof the air transport market or on studiesof plans and expectations of the parties engagedin the air traffic activity. Medium.and long-term traffic forecasting meansforecasting for periods of more than a few years. This is followed by a presentation techniquesfor traffic forecastingwhich are of based on studies of factors governing the traffic development. In so doing. PartlI of the Manual aims at illustrating the practical application of different forecasting techniquesin various (iii) THE ASSEMBLY: (1) REQUESTSthe Council to prepare and maintain. Techniquesand problems which are specially related to short-term forecasting are not included.the factors that need to be taken into accounttend to be different and more specific. This Manual representsa partial fulfilment of the requirement set out in the second resolving clause "to collect and develop material on current forecasting methods for dissemination to Contracting States from time to time as guidance in their own forecasting".airline planners.forecastingand practical guidanceon the application of these techniques. .long-term and medium-term forecasts of future trends and developments in civil aviation of both a general and a specific kind.and long-term air traffic . 5. as necessary. the Council should consult with other organizationsas appropriate. mustforeseefuture developments likely to require action by the Organization and must initiate such action in good time. A first edition of the Manual (Doc 8991)was published in 1972. including. in carrying out its continuing functions in the economicfield. WHEREAS Contracting States require global and regional forecasts of future civil aviation developments for various purposes.and WHEREAS the present activities of the Organization in the statistical field do not include assistance to Contracting States with regard to forecasting. It provides a survey of techniques currently used for medium. and hence different techniquesmay be needed. 2. Part I of the Manual first presents typical patterns along which the traffic may develop. and (2) REQUESTS the Council to make arrangements to collect and develop material on current forecastingmethods both for the purposes described in Clause (1) and for dissemination to Contracting States from time to time as guidance in their own forecasting. 3.FOREWORD I.

6. but eachmethod included has proved its practical applicability and may be applied again in appropriate cases.circumstances. It is recognized that the source material used in this Manual is not comprehensive and that in all probability there is additional material on approachesand methods that could be included with advantage. This is done by presenting numberof case a studiesof forecastsof airport and airline traffic and of the employment of forecasts in planning and investment evaluation. The fact that a forecastingmethod is summarizedin this documentdoes not imply that the ICAO Secretariatconsidersit superior to other methodswhich are not included. .

. . trend curves.Airport States) (United (Ivory States) Coast) 17 17 18 21 22 22 22 Canadian lATA airport aviation forecasting forecasting models. affecting the estimation. of independent functionalFORECASTING. international travel 24 24 25 26 Estimating the market share of carriers on a route.. air transport relationship. B -AIRLINE Passenger demand forecasting FORECASTS models models for for Australian Mexico. BY TREND PROJECTION 3 3 3 4 4 Specifications Statistical Graphical estimation. and traffic 7 8 10 10 12 C -MARKET AND INDUSTRY SURVEYS 13 13 Forecastingbased on market studies. Forecastsbased on opinions or planning D -PRESENTATION OF FORECASTS 14 14 PART II -CASE A -AIRPORT Forecasting STUDIES OF AIR TRAFFIC FORECASTING FORECASTS for Riyadh International Airport (Saudi Arabia) 16 Forecasting Forecasting for for Logan Abidjan Newark Western International Airport European International Airport (United Airports... (v) 16 1 . of extrapolation..TABLE OF CONTENTS Page INTRODUCTION PART I -METHODS A -FORECASTING OF FORECASTING TRAFFIC. B -ECONOMETRIC Forecasting Data Variables Statistical Specification collection. variables demand.

Page c APPLICATIONS Evaluation of a runway IN INVESTMENT extension at Yundum EVALUATION Airport (Gambia) 27 . at Townsville (Australia) Appendix A Appendix B Appendix C Appendix D SIMPLE LINEAR REGRESSION MULTIPLE LINEAR REGRESSION.28 29 29 Scenario Evaluation analysis of airport for airline development planning. ESTIMATES OF PASSENGERDEMAND ELASTICITIES INVESTMENT EY ALVA TION CRITERIA 31 35 37 41 .