1. Mark Koetse Department of Spatial Economics Vrije Universiteit Amsterdam Project A8 Consequences of climate change and weather conditions for the transport sector

2. 2 Researchers Olaf Jonkeren (ojonkeren@feweb.vu.nl) Muhammad Sabir (msabir@feweb.vu.nl) Mark Koetse (mkoetse@feweb.vu.nl) Jos van Ommeren (jommeren@feweb.vu.nl) Piet Rietveld (prietveld@feweb.vu.nl)

3. 3 Project A8: Three main lines of research A. Literature survey climate change and transport B. Climate change and road transport Modal choice and speed Traffic flow, congestion and traffic accidents C1. Climate change and inland navigation Economic loss due to changing water levels C2. Climate change and inland navigation Competitive position of the sector

4. 4 General research strategy 1. Analyse impact of weather on transport sector 2. KNMI scenarios show impact of climate change on weather conditions 3. Assess impact of climate change on transport sector

5. 5 A. Literature survey

6. 6 General areas of research Weather and road accidents Weather and bicycle use Weather, traffic flow, and traffic speed Water level fluctuation and inland navigation Extreme events: Storms and flooding Air transport delays and accidents

7. 7 Consequences for the transport sector Road transport: Safety Decrease in number of wet days decreases accident frequency, increase in rainfall at wet days and extreme rainfall increases accident frequency Ambiguity with respect to accident severity as well Increase in frequency and duration of dry spells decreases road safety when it starts raining

8. 8 Consequences for the transport sector Road transport: Congestion Increase in number of wet days and rainfall at wet days in winter increases congestion Decrease in number of wet days in summer decreases congestion Increase in rainfall at wet days in summer increases congestion

9. 9 Consequences for the transport sector Rail transport Higher temperatures reduce failures due to icing but increase failures due to high temperatures Changes in maximum wind speed are small Inland shipping Increase in frequency and duration of dry spells, implying higher prices and welfare losses Higher costs due to larger variation in water levels (increase in uncertainty)

10. 10 Consequences for the transport sector Air transport Relatively small impact as far as increases in maximum wind speed appear to be generally small Highly uncertain because it also depends on issues not in KNMI scenarios (e.g., visibility) Appears to depend to a great extent on changes in wind direction, which are uncertain and region specific

11. 11 B. Climate change and road transport

12. 12 Data sources Central question: How does weather influence the mode choice decisions of individuals? Data: OVG 1996 (AVV) More than 500,000 useable observations Contain 77 variables on personal characteristics and trip characteristics Survey covers the entire Netherlands Data: Weather Report 1996 (KNMI) Hourly weather data for every day in 1996 Recorded by 39 stations, covering all 458 municipalities of the Netherlands Contains information on temperature, wind, precipitation and sunlight

13. 13 Model specification Transport modes Walking (reference category) Bicycle Car Public transport (bus, tram, subways, train) Other (moped, motor, scooter, taxi, truck) Estimation method Multinomial logit model

14. 14 Main results Switch from bicycle to car under cold and extremely cold circumstances (4% to 8%) Switch from car to bicycle as temperatures increase above 10 °C (2% to 5%) For very high temperatures people switch back to car and public transport The effects of wind are small; the use of the bicycle decreases sharply at wind speeds higher than 6 Bft (around 4%) As precipitation increases, people reduce bicycle use and increase use of car and public transportation (around 3%)

15. 15 Current and future research Weather and morning commute: Speed, trip delays and substitution among different modes Weather, congestion and road safety Weather conditions Congestion Accident frequency and severity

16. 16 C1. Climate change and inland navigation

17. 17 Outline Analysis of economic loss in the inland waterway sector as a result of low water levels in the river Rhine? Research outline: Climate change Increase in frequency of low water levels in the river Rhine Higher price per ton for transportation by barge Costs for the economy

18. 18 Model specification Regression Dependent variables: price per ton, load factor, price per trip (all in logarithm) Explanatory variables Time trend Trip distance in logarithm Ship size Cargo type Navigation direction & backhaul Monthly dummies Water level (Kaub)

19. 19 Estimation results Average annual welfare loss period 1986-2004: € 28 million Welfare loss 2003: € 91 million

20. 20 C2. Climate change and inland navigation

21. 21 Model simulations To what extend will the competitive position (in terms of tons) of inland waterway transport in the Kaub related Rhine market deteriorate as a result of climate change? Model NODUS: A GIS based software model which provides a tool for detailed analysis of freight transportation over extensive multimodal networks Inputs for NODUS The infrastructure network (road, railroad, waterways) of Europe (supply) The transport flows between NUTS II regions in North West Europe (demand) Cost functions Output NODUS Estimates of the quantities that will be transported by each mode annually

22. 22 Climate scenarios Reference situation is a year that has no restrictions due to low water levels Base scenario: Normal water levels (average daily water levels 1986 – 1995) Alternative scenarios: Climate scenarios from KNMI Climate scenarioGlobal temperatures Change atmospheric circulation M +1 °C Weak M+ +1 °C Strong W +2 °C Weak W++2 °CStrong Table. KNMI climate scenarios

23. 23 Simulation results W+ scenario as an example; Difference base scenario and W+ scenario: 103 days cost increase of 23.4% - 17.8% = 5.6% 73 days cost increase of 23.4% Weighted average of 13.1% cost increase over 176 days No cost increases other transport modes Climate scenarioBaseMM+WW+ Average # of days with low water 10310013999176 Average % cost increase 17.8%17.9%19.9%17.6%23.4% Table. Number of low water days and change in transport costs per ton for KNMI scenarios

24. 24 Simulation results NSTR group Q0Delta Q % change Proportion rail Proportion road 02,052,664-94,484-4.6%0.630.37 12,894,153-53,792-1.9%0.090.91 25,597,912-53,024-0.9%0.570.43 320,161,424-187,682-0.9%0.270.73 41,876,955-37,588-2.0%0.660.34 51,295,421-42,175-3.3%0.250.75 69,608,355-159,866-1.7%0.050.95 71,981,214-44,096-2.2%0.960.04 85,250,013-100,511-1.9%0.320.68 93,768,604-139,357-3.7%0.870.13 Total (0-9)54,486,716-912,576-1.7%0.420.58 Table. Substitution from inland navigation to road and rail transport in W+ scenario

25. 25 Conclusions Literature survey Empirical findings often difficult to compare in a quantitative way Effects of weather (and climate change) are often ambiguous in sign Blind spots in knowledge, existing knowledge lacks robustness Mode Choice Most important pattern is substitution between car and bicycle due to changes in temperature, rain and wind Order of magnitude is between 2% and 10%, depending on the magnitude of the change in weather

26. 26 Conclusions Inland waterway transport Average annual economic loss due to low water levels in 1986-2004 period of € 21 million Substitution of freight of around 2% from inland waterways to road (share = 0.60) and rail (share = 0.40) transport

27. 27 Research output Koetse, Rietveld, 2007, Climate Change, Adverse Weather Conditions, and Transport: A Literature Survey, Department of Spatial Economics, Vrije Universiteit, Amsterdam. Sabir, Koetse, Rietveld, 2007, The Impact of Weather Conditions on Mode Choice Decisions: Empirical Evidence for the Netherlands, Department of Spatial Economics, Vrije Universiteit, Amsterdam. Jonkeren, Rietveld, Van Ommeren, 2006, Climate change and inland waterway transport: Welfare effects of low water levels on the river Rhine, Department of Spatial Economics, Vrije Universiteit, Amsterdam. Jonkeren, Jourquin, Rietveld, 2006, Modal split effects of climate change: a study to the effect of low water levels on the competitive position of inland waterway transport, Department of Spatial Economics, Vrije Universiteit, Amsterdam.