Meteorological characteristics of Turkey airports

Air transportation is a prime mover for all sectors of a country and especially for national economy. Air transportation which is especially backbone of tourism sector has great importance as a result of transporting more than 60 percent of worldwide tourist number. Growth trend in air transportation is rapidly in progress throughout the world. Since 2006 passenger traffic has been increased without slowing down. Risks regarding to weather conditions have substantial negative effect on aircraft safety. According to recent research, weather conditions are responsible of 20 percent to 30 percent of worldwide air accidents. At Europe, 22 percent of air traffic delays are due to adverse weather conditions. There are many meteorological risks for aviation. In this context, at this study firstly meteorological factors such as icing, wind, thunderstorm, precipitation, cloud ceiling and visibility which aggrieve passengers via posing risk to aviation and damaging companies in aviation sector are dealed together with their effects on aircrafts, diagnosis techniques and prediction with mitigation methods, then METAR, SPECI and TAF data obtained from Turkish Meteorological Service which includes 5 year period between 2010 and 2014 and data which includes delayed or diverted flights due to meteorological reasons between 2010 and 2014 which are obtained from a private corporation were examined in order to reveal the distribution of meteorological factors regionally and with respect to airport. In that study, between 2010 and 2014 were especially chosen due to reveal how meteorological factors affected negatively to passenger traffic of domestic flights which was increased from circa 50 million to 100 million and doubled with a conspicuously increase between 2010 and 2015. 46 airports which were active between 2010 and 2014 in Turkey were examined within the study. While data of cancelled or diverted flights due to meteorological reasons in 5 year period between 01/01/2010 and 01/01/2015 that were obtained from a private corporation were examining respectively, TAF forecasts which included data of departure and arrival airports and which were published at least 3 hours before departure time of the flight and METAR and SPECI data close to departure time with updated TAF forecasts for cancelled flight data; TAF forecasts which included data of departure and arrival airports and which were published at least 3 hours before departure time of the flight and METAR and SPECI data which were valid between expected flight time with updated TAF forecasts for diverted flight data were dealed together in the light of meteorological criterias such as surface wind, surface visibility, actual weather conditions, cloud information and air temperature with dew point temperature groups which were available in data and aerodrome approach procedures and approach minimum values that private airline corporation used for domestic airports. 97 meteorological parameters that were detected to have influence on cancelled or diverted flight data were standardized month by month according to airports via ratio between individual effect of meteorological parameter on cancelled or diverted number of flights and total effect of all the meteorological parameters on cancelled or diverted number of flights to determine their distribution in terms of airport by using Cluster Analysis method. By using SPSS Statistics software program which is provided by IBM, successively Centroid, Ward, Single Linkage (Nearest Neighbor), Complete Linkage (Furthest Neighbor), Average Linkage (Between-Groups Linkage) and Average Linkage (Within-Groups Linkage) hierarchical cluster analysis methods were applied to determine similarity relationships between airports with respect to 97 meteorological parameters. Dendogram graphs, Agglomerative coefficients and scatterplots Airport vs Cluster numbers were examined together. Considering the fact that, in each step of agglomerative algorithm the two most similar clusters come together to create a new cluster and that process repeats until all the objects are combined in one cluster whereas increasingstep number decreases similarity, optimum number of clusters were chosen 6 and 8 in Ward's Method, 6 and 9 in Average Linkage method, 6 and 8 in Complete Linkage method, respectively. Finally, One-way ANOVA statistic was applied in order to detect suitability of method and in 95% confidency level in each method how the number of clusters were clustered according to number of significant parameters that taken into account were shown. After applying each method; Ward's Method, Average Linkage (Within-Groups Linkage) and Complete Linkage (Furthest Neighbor) were found more useful beside the others. Because of evaluating more parameters Complete Linkage (Furthest Neighbor) method and Ward Method were seen to drive forward according to significancy analysis. Furthermore, considering obtained Dendograms and cluster diagrams Ward Method and Complete Linkage (Furthest Neighbor) method were seen to show more realistic results such as geographical regions clustering. In analysis for 7 geographic regions of Turkey between 01/01/2010 and 01/01/2015 period, in 6 (Aegean, Mediterrenean, Central Anatolia, Black Sea, Eastern Anatolia and Southeastern Anatolia) of 7 regions excluding Marmara region, it was shown that fog was the main meteorological factor that caused cancelled flights and diverted flights. It was seen that snow phenomena which carried weight with 5 (Marmara, Central Anatolia, Black Sea, Eastern Anatolia and Southeastern Anatolia) of 7 regions was another important meteorological risk parameter after fog. It was found that fog phenomena which was most remarkable meteorological risk factor was more influent on city basis especially in Şanlıurfa, Hatay, Konya, Gaziantep, Batman, Diyarbakır and Erzurum airports. In city basis, it can be said that main factors like continental climate, proximity to water resources, localizing in seaside are effective in fog occurrences. In addition, lake is an important factor that has risk of reducing visibility with its humidity content in brume situations. Furthermore, thunderstorm with CB activity can be associated with coastal structure and instability conditions. In this study, it was seen that in order to mitigate the risks of the most critical risk factor fog and the other significant meteorological parameters in regional basis, some projects should have been developed. Especially some of kind of projects like establishment of CAT3 Instrumental Landing System to airports where fog occurs frequently such as Hatay, Konya, Gaziantep, Batman and Diyarbakır that provides landing in very low visibility values as available in Istanbul Atatürk International Airport and Ankara Esenboğa International Airport, projects like draining areas that can be moisture sources such as pond and swamp, can give significant benefits. Similarly, developing projects or making investments for retrofitting or improving aeronautical approach procedures in aerodromes where aeronautical approach minimum requires higher value like Kahramanmaraş, Van and Erzincan, for the purpose of avoiding risks of brume and etc. that affects flights negatively can minimize risk factors.