DOES WEATHER IMPACT ON COMMUTERS’ TRAVEL DEMAND EMPIRICAL CASE STUDY OF BELGRADE Dragana Petrovic, Ivan Ivanovic, Vladimir Djoric Faculty of Traffic and Transport Engineering, University of Belgrade ABSTRACT Traffic impact on climate change is well-known, but there is also significant climate and weather impact on transport system. Impact of climate change on transport system can be defined as cumulative impact of weather conditions on transport system. Weather impact on transport system can be defined through weather impact on: infrastructure, transport supply, transport demand and road safety. Weather conditions are usually omitted from the research about transport demand or they are presented in a simplified manner. Weather impact on transport system should be explored at the local level. It is the only way to determine how transportation demand depends on different weather conditions. Characteristic weather appearance in the research area is showed in the paper, based on collected data about daily weather conditions during the previous five years (2009-2014). Under various weather conditions authors conducted pilot research. Based on that pilot research, weather conditions that most directly affect the transport demand were singled out. Weather impact on travel demand was not explored in Serbia and Balkan countries until now. Detailed research about weather impact on transport demand and travel behaviour was researched in Belgrade during 2014 and at the beginning of 2015. Authors conducted research on travel demand through online and paper household surveys, as well as through traffic monitoring. Employed Belgrade residents answered about their travel habits, travel patterns and changes that they would make because of different weather conditions through stated preference method. Employed residents were the target group, because the highest percentage of residents who have mode choice belongs to this group and also constitute a large part of the transport system users. Individual respondents picked one of given statements about weather impact on different travel purpose. Traffic monitoring was carried out at one of the most important Belgrade urban road. It was carried out in order to determine whether and how changes in transport demand, caused by different weather conditions, affect the hour traffic volume at one city road section, i.e. traffic volume distribution and variation. Research is still in progress and in this paper the first results will be presented. © AET 2015 and contributors

1

1. INTRODUCTION Traffic impact on climate change is well-known, but there is also significant climate and weather impact on transport system. Impact of climate change on transport system can be defined as cumulative impact of weather conditions on transport system. Weather impact on transport system can be defined through weather impact on: infrastructure, transport supply, transport demand and road safety. Weather conditions are usually omitted from the transport demand studies or they are presented in a simplified manner. (Koetse&Rietveld, 2009) The weather conditions impact on transport demand and travel behaviour is being explored in various regions of America, Australia, Europe and Asia. The largest number of researches is in Europe, but almost all previous studies were conducted in the Northern Europe (Aaheim & Hauge, 2005; Liu & Susilo, 2014; Berkum et al., 2006; Sabir, Koetse, & Rietveld, 2008; Böcker, Prillwitz, & Dijst, 2013; Böcker, Dijst, & Prillwitz, 2013). Weather conditions do not have the same influence on the transport demand and transport supply in different climate zones. Weather impact on transport system should be explored at the local level, because it is the only way to determine how transport demand depends on different weather conditions. (Eichhorst, 2009) There is great interest to mitigate the impact of climate change on global level. Many professional and scientific papers published in previous years confirm that this is up-to-date topic. Within different scientific disciplines there is increasing in the number of studies that consider the impact of weather conditions on the transport system. Most studies that analyse the relationship between traffic, climate change and weather conditions are dealing with the weather conditions impact on the vehicle speed, traffic accidents and cost of road infrastructure maintenance. Significantly smaller number of studies deal with the climate change impacts on transport demand. (Böcker, L., Dijst, M., & Prillwitz, J.,2013) The impact of weather conditions should be more carefully considered in transport demand studies. In past research, little attention was paid to the weather conditions impact on the individuals’ behaviour in everyday choices about activities (outdoor or indoor), destinations and transport modes. Most researches deal with the analysis of extreme weather conditions, while the daily weather conditions’ influences are less present themes. In terms of weather conditions impact on transport system, it is necessary to investigate the passengers’ behaviour both in normal and in adverse weather conditions. Analysis of the weather conditions impact on the usual passengers’ behaviour is important, especially bearing in mind the predicted climate change in the future.

© AET 2015 and contributors

2

Climate change and weather conditions impact on traffic system are not detail investigated in Serbia and Balkan countries until now. Previous studies (Böcker, L., Dijst, M., & Prillwitz, J.,2013) have shown that there is the potential impact of weather conditions on the trip generation, distribution (choice of destination) and modal split. From the above it can be concluded that in the transportation demand modelling it is necessary to take into account different passenger behaviour in transport demand in a different weather conditions. Study about the weather conditions impact on transport supply and demand in Belgrade was started during 2014 within two PhD thesis at the Faculty of Traffic and Transport Engineering. The aim of research is to collect data about the different weather conditions impact on transport supply and demand through survey and recording on the field. After conducted survey and field research the aim is to develop models for predicting relation between transport demand as well as transport supply with different weather conditions because for different weather conditions different trip generation, spatial and modal distribution may be expected, as well as different volume capacity ratio at street network. Planned comprehensive research required pilot survey and pilot field research. Results of pilot research are presented in this paper. Comprehensive research, which results will be topic of some future papers, is in progress. 2. BELGRADE WEATHER CONDITIONS OVERVIEW Belgrade has a moderate continental climate, with four seasons. Autumn is longer than spring, with longer sunny and warm periods. Winter is not so severe and January is the coldest month. Spring is short and rainy. The characteristic of Belgrade climate is also Kosava - the southeast-east wind, which brings clear and dry weather. It mostly blows in autumn and winter, in 2-3 days intervals. Authors present Belgrade climate characteristics for period from 2009 until 2013, five years period before pilot survey. The following graph shows the number of precipitation days. It includes all days in which precipitation was higher than zero.

© AET 2015 and contributors

3

Figure 1. Precipitation days for the analysed five-year period Average annual number of days with snow intensity greater of 5mm/m2 per day is 8.6, while the average number of rainy days with intensity greater of 5mm/m2 per day is 28,6. The rainiest months are May and June. The next graph shows other collected weather characteristics like freezing rain, fog, strong wind and snow on road per year.

Figure 2. Overview of other weather characteristics It can be seen (Figure 2) that number of windy days prevails among presented characteristics, on average 96 days per year. The average wind speed is 14km/h but certain wind gust can reach up to 130 km/h. There are also days where combination of very low temperatures and rain cause “freezing rain” phenomenon. This phenomenon is very important when it comes to traffic safety, but fortunately the number of such days is 5 days per year on average. When it comes to temperature, such as the authors provide, Belgrade has typically moderate continental climate. The following graph shows the average number of days with temperature higher that 300C, and below 00C. It is important © AET 2015 and contributors

4

to note that days during which the temperature at least once exceeds 30°C or falls below 0°C are shown.

Figure 3. Temperature characteristics 3. RESEARCH DESCRIPTION Detailed research about weather impact on transport demand and travel behaviour in Belgrade is in progress. Authors conducted pilot research on travel demand through online and paper household survey, as well as through traffic monitoring during 2014. Pilot research was conducted under various weather conditions. Online survey was conducted by means of open source online survey application LimeSurvey. Traffic monitoring was carried out using mobile cameras. Employed Belgrade residents answered about changes that they would make because of different weather conditions through stated preference method. Individual respondents picked one of given statements about weather impact on different travel purpose. Stated preference method has advantages and disadvantages but it was shown that individual’s stated preferences among hypothetical travel scenarios were a reasonably accurate guide to true underlying preferences (Wardman, D., 1988) The target group in survey was employed residents who answered about their travel habits and travel patterns. The survey included questions about socioeconomic characteristics of respondents, average number of trips per day, trip purposes and the most common mode of transport that respondents use for travel to work, for shopping, recreation and social/entertainment. Employed residents constitute a large part of the transport system users and during the day they have two types of trips, work-related and non-work-related. Most studies showed that there is higher weather condition impact on recreational compared with utilitarian trips. (Böcker et al. 2013). The survey was conducted on employed Belgrade residents to determine if weather conditions differently affected workrelated and non-work-related trips. Number of respondents who fully responded to survey was 285. Respondents were selected in the way that their residences © AET 2015 and contributors

5

are equally distributed in the Belgrade urban area. Gender ratio in the sample was 46% of women and 54% of men.

Figure 4. Respondents' residence in the Belgrade urban area Traffic monitoring in pilot research was conducted at four locations. The diagram shows the research locations.

Figure 5. Traffic monitoring locations

© AET 2015 and contributors

6

These four points were located on two main urban road, Boulevard Kralja Aleksandra in central zone of Belgrade and Jurija Gagarina in New Belgrade. Traffic flow parameters were recording during days with different weather conditions. Because the changes in travel time, speed, capacity and volume capacity ratio are more related to the transport system supply, data which may show whether there is connection between changes in transport demand and different weather conditions are presented in this paper. Traffic monitoring was carried out over two snowfall days, two rainy days and three days with clear weather, from 7PM until 10PM with the aim to cover the morning peak hour. 4. THE FIRST RESULTS OF RESEARCH In this paper results related to the transport demand change because of adverse weather conditions are shown. Based on pilot research, weather conditions that most directly affect the transport demand are singled out. Average number of trips in the sample during day with usual weather conditions was 3,5. It is important to emphasize that the average mobility in Belgrade is around 2.3 and mobility in this pilot survey conducted in Belgrade is higher because respondents were only employed Belgrade residents. Twenty percent of respondents did not make even one movement in addition to commute (going to work and back to home). Eighty percent of respondents who have movements with other purposes make on average two additional movements. In the age group below 30 years there are least of those who only commute, which is logical because they are young people who have activities related to recreation and social during the day. Respondents mostly go to work by motorized modes of transport, while 13% go by foot or bicycle. Use of public transport and passenger car in the sample is uniformly, 45% use passenger car and 42% public transport. Employed Belgrade residents go to planned shopping and social/entertainment mainly by motorized transport mode. For planned shopping the majority, precisely 74%, prefer passenger car while for social/entertainment trips 53% use passenger car and 33% public transport. Most recreation trips are non-motorized trips considering that 45% of employed Belgrade residents go to recreation by foot or bike, and 55% by some motorised mode of transportation.

© AET 2015 and contributors

7

Respondents answered whether they cancel planned shopping, social/entertainment or recreation trip due to some of offered adverse weather conditions (temperatures under 0 ° C, heavy rain, snow or sleet, snow on road, "freezing rain", fog, temperatures above 30 ° C, Strong wind or /storm). Survey showed that demand is most affected by heavy rain, freezing rain and storms. Most respondents would cancel shopping, social/entertainment and recreation trips in the case of freezing rain. Travel demand would be for 20% less if freezing rain falls during the afternoon when respondents go shopping and take social/entertainment and recreation trip. Rain and storm have slightly weaker influences on non-work demand than freezing rain. Survey showed that low and high temperatures have the least impact on the demand, since research shown that in the case of low temperatures the number of trips during the day would decrease by 9% and in the case of high temperatures by 8%. It should be emphasized that recreation trips are most affected by adverse weather conditions because most trips with purpose of recreation are non-motorized trips. Insignificant number of employed Belgrade residents can stay and work at home due to weather conditions. Basically it is expected that employees come to workplace in any weather conditions. Therefore employed Belgrade residents almost never cancelled their trip to work due to adverse weather conditions. In adverse weather conditions employed Belgrade residents make changes in mode choice and starting time in commute trips. Survey showed that more than half respondents reported that they always informed about weather conditions for the next day in order to adapt their mode choice and starting time. Based on traffic monitoring, the authors performed traffic counting in five minutes intervals on the aforementioned locations. Traffic counting results at four locations on main urban roads in different weather conditions are shown in the following graphs. The precipitation intensity during the monitoring days for this paper is not taken into account.

Figure 6. Boulevard Aleksandra Karadjordjevica – Location 1

© AET 2015 and contributors

8

Figure 7. Boulevard Aleksandra Karadjordjevica – Location 2

Figure 8. Jurija Gagarina street – Location 3

Figure 9. Jurija Gagarina street – Location 4

Curve that represents the average value of the results was formed for recording days with the same weather conditions. Based on results it can be concluded that greater transport demand is present at almost all four locations in days with precipitation. This greater transport demand can be explained in two ways: first is that transport users may decide to use the passenger car to a greater extent in the days in which precipitation is announced and second that they may decide to use city's major streets as safer and faster in adverse weather conditions. Graphs confirm that in morning peak period transport users decide to start their trips earlier in relation to the day with good weather conditions, in order to avoid delays and arrive at work on time. It results in shifting of peak hour which can be seen on Figures 6-9. © AET 2015 and contributors

9

5. SUMMARY AND CONCLUSIONS The pilot research showed that adverse weather conditions affect transport system users’ behaviour. On the Belgrade territory adverse weather conditions have greater impact on non-commute trips than on work-related trips. For noncommute trips changes in number of trips are observed especially for recreation trips. Usually respond to adverse weather condition is trip cancelation or in lower percentage trip postpone. Because employed residents rarely cancel trip to work under the influence of adverse weather conditions, changes for those trips are reflected in mode choice and starting time. Shift of morning peak period, which is determined through traffic monitoring on described locations, supports results about work-related trips obtained from survey.

6. REFERENCES Aaheim, H. A., & Hauge, K. E. (2005). Impacts of climate change on travel habits, A national assessment based on individual choices (pp. 1–38). Berkum, E. van, Weijermars, W., & Hagens, A. (2006). The impact of weather on urban travel demand in the Netherlands. In Proceedings of the EWGT2006 International Joint Conferences (pp. 245–252). Böcker, L., Dijst, M., Prillwitz, J. (2013) Impact of Everyday Weather on Individual Daily Travel Behaviours in Perspective : A Literature Review. Transport Reviews, 33(1), 71–91. Cools, M, Moons, E., & Wets, G. (2010). Assessing the impact of weather on traffic intensity. Weather, Climate, and Society, 1–14. Eichhorst, U. (2009) Adapting Urban Transport to Climate Change, Module 5f, Sustainable Transport: A Sourcebook for Policy-makers in Developing Cities D. G. für I. Z. (GIZ) Gmbh, ed. Division 44. Water, Energy, Transport, p.62. Koetse, M. J., & Rietveld, P. (2009). The impact of climate change and weather on transport: An overview of empirical findings. Transportation Research Part D: Transport and Environment, 14(3), 205–221. doi:10.1016/j.trd.2008.12.004

© AET 2015 and contributors

10

Koetse, M.J., Rietveld, P. (2009) The impact of climate change and weather on transport: Anoverview of empirical Transportation Research Part D, 14, 205–221.

findings.

Liu, C., & Susilo, Y. O. (2014). Investigating the impacts of weather variability on individual’s daily activity-travel patterns: a comparison between commuters and non-commuters in Sweden. In TRB 2014 (p. 25 p). Maze, T., Agarwai, M., & Burchett, G. (2006). Whether weather matters to traffic demand, traffic safety, and traffic operations and flow. Transportation Research Record: Journal of the Transportation Research Board, (August), 17. Wardman, D., (1988) A comparison of revealed preference and stated preference models of travel behavior, Journal of Transport Economics and Policy, 71-91.

© AET 2015 and contributors

11