Mo she gu hjhh hhjk - Transport Studies Unit
Transcription
Mo she gu hjhh hhjk - Transport Studies Unit
Workshop Programme 8:45 – 9:00 Tea and Coffee (served in seminar room 9) 9:00 – 10:00 Overview and progress to date (Moshe Givoni and Yoram Shiftan) 10:00 – 12:00 Workshop discussion and exercises 12:00 – 13:00 Lunch and discussion (served in seminar room 9) 1 Britain-Israel Research and Academic Exchange Partnership (BIRAX) project: The use of state-of-the-art models by policy makers to address global energy and environment challenges: The case of transport policy Knowledge Utilisation and the use of models in policymaking 2nd Workshop, 27 October 2011, St. Anne's College, Oxford, UK 2 Project overview Research team: University of Oxford: Moshe Givoni and Eda Beyazit Technion: Yoram Shiftan, Robert Ishaq and Omer Tzur Duration: 12 months Budget: £30,000 Funder: The British-Council 3 Objectives Formal: To investigate the contribution of quantitative analysis to the formulation of policies to address energy and environment challenges, focusing on the transport sector, and to explore ways to increase such contribution. Practical: To understand the nature of and reasons for the gap in (transport) policy making between the „knowledge‟ state-of-the-art modelling techniques can provide policy makers with and the „knowledge‟ on which they base their decisions on. 4 Context and assumptions “For political reasons, it might be that policy makers ignore the modelling results, or try to only use the modelling process to support pre-decided policies („solutions looking for problems‟ as coined by Kingdon, 1984)”. Project report • • • 5 While an important and unavoidable part of policy making, it is outside the scope of this research We focus on „objective‟ use of models and analysis to inform policy making Does not mean „value-free‟, „fully rational‟ use of models Methodology Emphasis on travel demand modelling The most important information needed to assess various policy actions and their impacts General approach: The (general) policy process State-of-the-art model development The models used in the policy process The role of model results in the decision making UK Israel 6 Policy process √ √ (given) Models √ √ (what can we do?) Model development Three broad categories of travel demand modelling developments: •Sketch Planning Model •Four step model •Activity based model More on this in the next presentation... 7 Use of models in the policy process Main issues: Are the right tools used? It is important to use the appropriate model for particular problems making sure that the models meet the needs of the planning process. Furthermore, models need to be designed to help the users of the models, and the users need to be able to adjust the model to answer specific questions (Jonsson et al., 2011, our emphasis). Is the right knowledge produced? Are „models‟ or modellers ready for changes on the context side and in the policy agendas? Are transport modellers well equipped to detect and respond to those changes? (Gudmundsson, 2011, our emphasis). 8 Use of models in the policy process (cont.) Is the knowledge communicated effectively? Communication between modellers and practitioners is crucial in order for transport planners to understand how the new approaches can better address their planning needs (Davidson et al., 2007) Lack of communication can create barriers for planning and decision making in transport. From a practical standpoint, relationships seem to thrive best in environments where there is more communication, flexibility, and focus on meeting business needs of the transportation agency (Schofer, 2007, our emphasis) There is a need for models to support the learning process of different interest groups as well as the need for those groups to understand the basic philosophy and assumptions of models. There is a need for developing a common language between transport modellers and “people that are not insiders in the domain of modelling” (Brömmelstroet and Bertollini, 2011, our emphasis). 9 Evidence from the UK Focus on one county and on the LTP3 process •Longer planning time horizon (suggested by the county) •Objectives set locally and in consultation (but broadly rely on and follow DfT objectives) The use of models: The DfT strongly advises local authorities to consider different methods for making their decisions and suggests that it is sensible to consider whether a model is necessary or not for local transport plans as the creation of a transport model is costly and time consuming. However, the DfT presents an extensive section on WebTAG about general principles of transport modelling, choice of modelling approach, transport-land use interaction models, data sources and the level of detail of the models. 10 Evidence from the UK (cont.) Models: SATURN - Traffic Model Emme 2 - Traffic Software Accession - Accessibility model for public transport VISSIM - Microsimulation Model for small scale projects Intra-SIM - a transport policy simulation tool for decision makers “The general view I think would be the models give too much detail... We [need] to take the results of that and simplify it, simplify and simplify just coming down to [an understandable level] ... that gives you the information at a sort of level that the councillors can deal within their constraints of time”. (Policy analyst) “„X‟ model goes only so far in detail. It doesn‟t go to the very fine detail. It has just the key routes in „Y city‟ models. (...) [Models] helps us to understand the strategic impacts [of a development]”. (Planner) 11 Evidence from Israel Focus on „mega-projects‟ of mass transit systems in Tel- Aviv, Jerusalem and Haifa Policy process, context Very centralised decision making. Done mainly at the Ministry of Transportation and Road Safety (MOT) in coordination with the Ministry of Finance (MOF) Local authorities have some say (participate in infrastructure budget usually up to 30%, in public transportation about 15%) MOT operates through various sub-companies who are either own by the MOT or co-owned by the MOT and a local authority. Lack of local Transportation Planning Authorities complicate the decision process and put a heavy load on the MOT 12 Evidence from Israel (cont.) Models: • The Tel-Aviv metropolitan area - completed and starting to use a new activity based model which represents the state of the art in travel demand models today • The Jerusalem metropolitan area - in the process of developing such a model. • The Haifa and Beer Sheva metropolitan areas - regional models, which are traditional trip based, but of good quality Use of models The companies will either develop the models (own manpower) or bid and contract the model development to private consultants. Senior people in these companies act as a very important actor in the decision making process between the modellers (often work the company) and the policy maker level of the MOT and MOF. They are the „professional officials‟ (sometime the general managers of the companies). 13 Evidence from Israel (cont.) “The transportation models are very significant and have great effect on the planning process… I think their effect is excessive… There are many problems in models and they need further development… The problems [in the models] may be ignored due to „addiction to numbers‟ in the [decision making] process”. (Policy analyst) “The models are "black box" for some of the officials involved in decision making… [Because of] their complexity, and the need of professional knowledge in their operation, they are not accessible to external intervention of policy makers and other unprofessional officials” (Policy analyst) “Most decision makers who use a model are not familiar with the model‟s properties such as the model structure, model assumptions, explanatory variables etc.…” (Policy maker) 14 The models‟ paradox • (In Israel) The professional level believes that advanced models are really needed. • It is not clear how much this motivation is supported by the policy makers, and whether they also believed so or they felt they need the best state of the art models to support their decisions The key people in the supply-chain of models‟ knowledge see the models as a black box, and one that „surly‟ know the answers => Development in state of the art models seems to mainly serve to increase the size of the black box, resulting in increasing the gap between state of the art and state of the practice Tow options to move forward: • Develop „simple‟ state-of-the-art models • Communicate the knowledge produce by (and the limitations of) current state-of-the-art models 15 The Role of Travel Demand Models in Policy Making Yoram Shiftan Stages in Externalities‟ Evaluation Land use Socio economic Demographic Transportation system Travel Demand Model Traffic Data: Speeds and Volumes Travel time Accessibility emissions noise accidents Pollution Concentrations exposure Casualties/damage Monetary value Brief History of Travel Demand Models generation distribution assignment Mode choice Activities Need of Behavioral Models Aggregate problems Individual Behavior Policy Representing Activity/Travel Behavior Tours Schedule Space Trips Space H Space H H W W W S H H H D H Time S H H: Home W: Work H H H D Time Time S: Shop W S D: Dinner out S D D Complex Responses to Policies (e.g., peak-period toll) Potential Responses to Toll Pre-Toll Schedule (a) Change (b) Change Mode & Pattern Time & Pattern Space (Home) Space (c) Work at Home Space Space Car Car Bus Work Work Work Car Shop Time Car Shop Time = Peak Period Shop Time Time Shop The use of Activity Based Models for Policy Analysis • Secondary effects - adjustment to the activity pattern that have to be made in response to the primary effect • Can better capture “induced demand” • Can provide better input requires for externalities evaluation • Can provide detailed travel data by various variable for equity issues and other purposes D ay an Sa M pi en a c Kor r he es m h Be Ba gi n tY Ka f T H am et a'a N ov m a em l Yo b er se ft Ba al rI la Ba n Ja l fou bo r t R in sk ot h y H ssh a' at i ld z M ma ac u hr t oz H et ab as h Is t ak ov Er li c Be h n Z Sa vi la m El e ife l Al et le n b C ar y li b ac Sh a u Yeh h l H ud am it Ar e le lo ch so Ab rof ba f H il e l Be Bi a n li k Ah Gu a r rio on n ov Sh i tz en sh ka ac r h be a m i li n s D on an kn er O Pe rl o ta ch Pi n v Ti sk kv er a C BS M os he Tel-Aviv Red Light Rail Line Demand Estimates 2030 passengers per morning peak hour northbound Neta Model New Activity Based Model 10000 9000 NTA Northbound 8000 #371276 TAM Northbound 7000 6000 5000 4000 3000 2000 1000 0 Red Line Demand Estimates 2020 and 2030 Morning Peak Hour Boarding סה"כ עולים לקו האדום 50,000 45,900 47300 45,000 40,000 35,600 35,000 30,000 23,300 25,000 24,800 20,000 15,000 12,900 10,600 10,000 5,000 - מודל ת"א גרסה 2.2מודל ת"א גרסה 2.1 שנת 2030 מודל נת"ע זכיין שנת 2020 Role of Travel Demands • Compare alternative investments alternatives • Justify a large project/investment • Details analysis of projects (route alignment station location) • Policy implementation • There is no one suit that fit them all Questions about the Models • How accurate they are? The role of uncertainty. • Are they biased? • The role of various assumptions, and input into the model. • What projects/policies are they capable to analyze? • Is their output sufficient for the evaluation of externalities? • How well they deal with induced demand? • How all these factors affect project ranking? Advantages of Advanced Models • Provide better understanding of travel behavior • Provide more and better information (enable better equity analysis) • But is it really more accurate??? • Provide policy makers the “best” tools/support • The “ideal” trade-off between complexity and simplicity is context specific Questions for Discussions • Are we using the “right” models? Should we invest less or more in model development? • How much do the models really affect policy makers? and at what level of decisions? • The role of assumptions/input vs. the model itself. • How and what output/results should be presented to policy makers? • The use of models by policy makers Behavioral Realism and Computational Complexity Behavioral Realism Computational complexity Benefits from Behavioural Computational Simplicity ? Behavioral Realism Realism and Total Model Benefits Computational Simplicity THE USE OF ACTIVITY BASED MODEL FOR PROJECT APPRAISAL Yoram Shiftan Stages in Externalities‟ Evaluation Land use Socio economic Demographic Transportation system Travel Demand Model Traffic Data: Speeds and Volumes Travel time Accessibility emissions noise accidents Pollution Concentrations exposure Monetary value Casualties/damage The Added Value of ABM Output • Can get detailed travel data: – By socio-economic – By auto ownership – By type of tours – By type of trips (cold/hot starts) • Ability to deal with Induced demand Induced Demand -The Assumption of Fix Demand Supply Projects • Highway Project – travel time savings per vehicle will be less than estimated – vehicle kilometers of travel will be more than estimated – emissions and other externalities will be higher than estimated – Benefits for new riders are ignored • Transit Project – fewer passengers will enjoy the improved service and accessibility. – revenue will be lower than estimated. The Assumption of Fix Demand on Users’ Benefits D S1 a e S2 b h g f c Bias from the Assumption of Fix Demand Fixed Demand Bias by V/C and Demand Elasticity 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 0.25 0.50 0.75 1.00 1.25 V/C E=(-0.25) E=(-0.5) E=(-0.75) 1.50 Air Quality • Important variables for air quality analysis – VMT – Travel by Mode and by Occupancy Rates for Auto Modes – Fraction of Cold/Hot Starts – Speed/Acceleration/Driving Profile. – Travel by Time of Day and Time/Location of Starts – Travel by Vehicle Class and Model Equity • Ability to analyze data by various categories: – Income level – Auto ownership – Residential lcoation The use of Activity Based Models for Policy Analysis • Secondary effects - adjustment to the activity pattern that have to be made in response to the primary effect • Can better capture “induced demand” • Ideally can provide all the input requires for policy analysis and externalities evaluation • Equity issues Nohal Perat • A single projects vs. a system wide project • Highway projects vs. public transport projects • For a single project use the default OD trip table provided by the model team (assume no induce demand) • For system wide projects need to run the full model without and with the project. System Wide Projects • Major public transport projects • Large highway project (over 250 Million NIS) • The induced demand would be evaluated by the Rule of the Half Need to do a Test Case • • • • Netania BRT Small highway project Large highway project Existing highway project פרויקט BRTנתניה – מאפיינים נבחרים – הפרש בין מצב "עם פרויקט" למצב "ללא פרויקט" אמצעי אוטובוס ת"צ רבת קיבולת רכבת רכב פרטי מאפיין סה"כ ק"מ-נוסע סה"כ שעות נוסע סה"כ עליות סה"כ ק"מ-נוסע סה"כ שעות נוסע סה"כ עליות סה"כ ק"מ-נוסע סה"כ שעות נוסע סה"כ עליות סה"כ ק"מ-רכב סה"כ שעות-רכב סה"כ נוסעים שלב א' 24,962 784 1,273 15,151 769 3,007 7,348 109 224 -5,558 -411 -319 שלב ב' 14,877 126 210 26,227 1,394 5,292 1,120 24 100 -18,051 -416 -1350 48 Second and Third Phase • Study the use of the variance from the various runs (minimum and maximum) • This should not be instead of sensitivity analysis currently recommended in Nohal Perat. • Improve air quality output • Use for equity analysis Britain-Israel Research and Academic Exchange Partnership (BIRAX) project: The use of state-of-the-art models by policy makers to address global energy and environment challenges: The case of transport policy Discussion 2nd Workshop, 27 October 2011, St. Anne's College Oxford, UK 50 Consider the last time you were involved with (witness to) use of models in planning / policy making. How will you rate the models used? 100 Model benefits (for decision making) 1 Behavioural realism Computational Simplicity 1 100 51 100 1 Assuming “model benefits” were less than 100. What will be your recommendation to increase „Model benefits‟? 100 Model benefits (for decision making) 1 Behavioural realism Computational Simplicity 1 100 52 100 1 Which line of action you feel should get more attention? •Develop „simple‟ state-of-the-art models •Communicate the knowledge produce by (and the limitations of) current state-of-theart models 53 STAGES Q I: The weakest link Q 2: The weight of each stage Input Data and Assumptions Model Structure Model Output Modellers‟ Presentation Policy Implications Decision Q I: Which stage is the weakest link in the supply chain of information/knowledge for decision making? Q 2: What is the weight of each stage in producing information? 54 What could be the next steps in the on the advancing the useful use of models in the policy making / planning process? What forms of research design can usefully involve both the policy and the modelling „camps‟? 55 Thank you! 56