Title:مدل سازی تعاملات توسعه شهری و فعالیت بدنی

این مجموعه وبینار در مورد «مدل‌های مبتنی بر عامل برای بهبود سلامت» تحقیقات فعلی را نشان می‌دهد که از روش‌های مدل‌سازی مبتنی بر عامل برای ارائه بینش‌های جدیدی در مقابله با چالش‌های فوری بهبود سلامت استفاده می‌کند. این قسمت بر چگونگی تأثیر توسعه شهری بر سطوح فعالیت بدنی و سفرهای فعال تمرکز دارد. این وبینار شامل دو سخنران است: دکتر لئاندرو گارسیا، دانشگاه کوئینز بلفاست هم افزایی بین تغییرات شهری در مقیاس بزرگ برای ارتقای فعالیت بدنی و اهداف توسعه پایدار سازمان ملل، دکتر هیزئو راین کوون، دانشگاه کالج لندن: شبیه‌سازی تعامل کاربری-حمل و نقل زمین مبتنی بر ABM: سالم‌تر توسعه شهری و رفتار سالم تر سفر

قسمتي از متن فيلم: Hello everybody welcome to the PHASE webinar  second one in the series in this one we’re   looking at modelling the interactions between  urban development and physical activity you’ve   got two really interesting speakers to talk to you  we have Dr Leandro Garcia from Queen’s University  

Belfast and we have Dr Heeseo Rain Kwon from  University College London we’re going to start   off first with Dr Leandro Garcia he’s a lecturer  in complexity science and public health at Queen’s   University Belfast Centre for Public Health he  has 10 years of experience in the development  

And application of systems thinking and complex  systems methods including agent-based modelling   to investigate and address public planetary health  challenges among his current roles he co-leads the   complexity planetary health research cluster at  Queen’s University Belfast he’s a member of The   WHO Collaboration Centre for research and training  on complex systems and network science for  

Non-communicable disease prevention and he also  co-leads the UKPRP development theme for systems   thinking and complex systems and is part of the  systems evaluation network organization committee   what Leandro is going to be talking  about today he’s going to present work   investigating the potential impacts of  large-scale recreational and transport  

Related physical activity promotion strategies  on sustainable development goal outcomes in   cities of high middle and low income countries  with that I’ll hand you over to the Leandro thank you very much Ricardo and hello  everyone thank you for joining us for   this webinar I’ll just share  my screen and then we start

Hopefully you can see my screen yes we can thanks very much okay so I will speak today about synergies between  large-scale urban changes for physical activity   promotion and the UN sustainable development  goals and this is our work that we conducted  

And then published in 2021 together with others  as actually this paper encompasses a range of   activities or outputs and the ABM agent-based  model is just one part of the whole paper   and the rationale for the paper is that the health  of humans and our planet hangs in balance as we  

All know and the UN SDGs lay out objectives for  saving the planet and enhancing quality of life   and while not quite a threat to planetary  existence as climate change or global poverty   physical activity still accounts  for five million deaths per year   so we try to investigate how large-scale  physical activity promotion strategies  

Might have synergies with the U.N  sustainable development agenda so   in these aspects of multi-solving for multiple  problems we face today in different countries   and for this paper we had four main aims the  first one was to describe the unresolved physical  

Inactivity pandemic the potential synergies with  the UN SDG agenda the next we try to identify   plausible links between large scale physical  activity promotion and as SDG achievement   the third objective which is related with the ABM  was to explore the possible impact of these at  

Scale physical activity promotional strategies  on both physical activity and the SDG related   outcomes across high middle and low-income country  city types and I’ll speak more about city types in   the next slides and lastly based on the outputs of  these three first aims we synthesize the results  

In sector specific recommendations  to guide policy research and action here in this slide I show very briefly what  we found in terms of linkages between the   physical activity promotion strategies and the  ۱۷ SDGs so we gathered information from three   different sources from the WHO Global Action  Plan for physical activity from the knowledge  

Of physical activity experts and from the  scientific literature to try to understand   what are the possible linkages between these  physical activity promotion strategies and all   the SDGs and we found that the strategies  that we investigated were the based on the  

Eight best investments for physical activity  promotion and what we found is that among   these eight strategies 15 out of the 17 SDGs  could be linked and then also have synergies   and try to move to solve the different issues  we observe across the globe so this is just to  

Set the scene where the agent-based model sits  within this paper so after we identified these   linkages we use these findings to inform  elements of our agent-based model, which I   will spend most of the time discussing today so the agent-based model has a purpose to   investigate the potential impacts of  large-scale recreational and transfer  

Related physical activity promotion  strategies on six SDG related outcomes   and these outcomes were first road  traffic deaths transportation mode share   convenient access to public transport  levels of finding particulate matter   access to public open spaces and levels of  CO2 emissions and beyond these six outcomes  

We also were interested in observing the  effects on recreational physical activity   and beyond the novelty of investigating  these six or seven outcomes together   we also aim to do that for different types of  cities so we designed three cities as abstract  

Representations of common cities in high middle  and low-income countries these cities were not   meant to resemble any particular real city but  were instead designed to represent cities which   were similar across a core set of characteristics  and what we did however was to use information  

From real world cities to inform the attributes  of each one of these abstract cities and we’ll   speak more about the icons you can see here in  this in this slide when we speak specifically   about the environment that we model within the  ABM so this was the first ABM that included two  

Physical activity behaviours each one shaped by  individual interpersonal and environmental factors   on three different cities each one offering  a unique environmental socioeconomic features these ABM was based on a conceptual model that  contained two main elements so on the top bubble  

You can see the factors and the relationships  that shape recreational physical activity practice   and the bottom bubble you can see the factors  and relationships affecting transportation   mode choices and you can see there’s a part of  these two bubbles that are overlapped which are  

Elements that affect both types of behaviours  this overarching structure was informed by the   socio-ecological model for physical activity  and elements of social practice and theories   and the decision model was learned by people’s  behaviour the people’s let me start again so the  

Decision models of running people’s behaviour  was informed by elements of discrete choice   theory opinion dynamics and multi-level theory  of behaviour and the general structure of both   bubbles was also informed by previous work  especially on developing leisure time physical   activity on how to use those conceptual models  to develop agent-based models of leisure time  

Physical activity and then we expanded that to  include the transportation behaviour as well   in terms of environments and agents so before I  go ahead I just very briefly say that the full   model description follows the ODD+D protocol and  it can be found in the article’s supplementary  

Material and the model itself the ABM itself  and all outputs generated from the model are   in an OSF page and the link to that page  is in the article which is free to access   so the environment was a 65 by 65 grid with  non-contiguous sites and each square patch  

Was equivalent to a quarter of square kilometre  and the total area of the environment was around   ۱۰۰۰ square kilometres and these dimensions  were based on the total area in the range of   what is observed in not to large size cities  around the world a meaningful path length  

Or area for calculating distance to locations  and also to adequately accommodate behavioural   interpersonal environmental factors are  relevant for this model for instance   the number of spaces for physical activity  practice within an agent’s perception rate within each one of these abstract  cities each one of these patches  

Were composed by different attributes were  represented different attributes so here’s   an example of a high income environment I’ll  just select the laser point here so you can   see there are four colours here the yellow colour  represents high income area the red colour middle  

Income area and here the centre run the low income  area and in the very middle the downtown it’s of   course very segregated of course we know cities  are not as segregated as these example shows but   it’s again an abstract representation so we are  basically saying that this is the largely a high  

Income area a mid-income area a low income area  and the downtown and each one of these patches had   an area level income as I showed you but also  a transport infrastructure score meaning how   good it is for car and for walking and cycling and  also representing the quality for public transport  

And each one of these patches could  be populated either by a space for   recreational physical activity either public or  private so these are these green patches here   workplaces which are the white patches here  other places for social interaction which   are represented here by the purple patches  public transportation is stops which are the  

Red flags and residential locations when none of  these other spaces was defined here is just an   example so of course for visualization purposes  the possible spaces or locations were much more   distributed and better distributed according to  the type of site of city but also the income area

And the environment then was informed by  information from real world cities and that   included income per capita income inequality  population density distribution of jobs outdoor   air pollution transport mode shares coverage  quality of road infrastructure walk infrastructure  

Etc and this information came from three specific  cities so for the high income city we use Phoenix   in the US for the mid income city Bogota in  Colombia and for the low income city from Accra   and the spatial characteristics were sourced  from multiple funds from for instance for the  

City boundaries we use the cities web  page and the humanitarian data exchange   for distribution environmental elements we  use SRI openshift Maps Google Places API Etc in terms of agents we modelled a hundred one  thousand adult persons with the following  

Attributes with a range of attributes so ranging  from residential location the income level the   ownership of car motorcycle bicycle and the size  of their personal network the preferred place to   do recreational activity their intention to  engage in recreational activity workplace  

Location and a number of other locations  they visit over the week and travel mode   choices over the week for each one of these  destinations so this is what the general   aspects or attributes of these ABM  in terms of environment and agents

In terms of scenarios we model six scenarios  based first on the business as usual case and   the next three models of scenario two and three  and four were selected among those we observed   from the list of all eight best  investments for physical activity  

That were very well linked with some  of the SDG outcomes or our indicators   so in the second scenario we aim to  reduce inequalities in public transport   in the third one we reduce inequality in public  recreational spaces and in the fourth one  

Inequalities in walking and cycling infrastructure  and we did that in the three scenarios we reduced   inequalities by systematically improving  what we observe in the best served area   and then bring every other space to the same  levels of services as the best served place  

Then we combine what we observe in scenarios 2, 3  and 4 for these three interventions and lastly we   did the same combination but added increasing  costs of car trips and these strategies were   implemented after one valid year in the model the  model was I will show you in the next slides run  

To be equivalent to 10 years okay so every time  stack was one real equivalent to one real world   week and the model was run to be equivalent to 10  years so in the general processes this ABM is that  

First let me show you again with the laser point  we update the intention to engage in recreation   of this activity considering aspects like past  behaviour of personal social network behaviour of   the community at large and now so that is modified  or moderated by the quality and accessibility of  

Spaces for recreational positivity that  then leads to the individual deciding to   engage or not in recreational activity and if  it does engage the weekly frequency of activity   and then every four time steps every month each  individual re-evaluate modes of transportation  

To every one of these places so to work to other  destinations including the recreational activity   space and to reassess that each person will then  consider their vehicle ownership travel distance   travel costs a person past travel behaviour the  behaviour of their personal social network of  

The community at large and the quality of the  transport infrastructure in their neighbourhood   and once that is considered they travel  to these different places and then every   four time steps every month they  re-evaluate the modes of transportation

I will show you very briefly some of the elements  about the selection of spaces or how we will   personalize every step or every element of that  flow chart so we have I’m not going through all   the equations here but just to show that for  each one of these main decision points people  

Had to select the best place to be physically  active based on their perception of the quality   of that space the accessibility to that space how  affordable that space is depending on the person’s   income so this is what I just said then we had the  recreational activity behaviour and again I’m not  

Going through this I’m happy for these slides to  be shared with you after in case you want to see   the details of the equations but just to say that  the physical activity behaviour was based on an   extension of the continuous opinions and discrete  actions model so that means that intention to be  

Physically active which is a continuous variable  would be updated every week every time step and   based on the intention each person would  then decide to do recreational physical   activity or not in that particular week and that  intention is based on the person’s past behaviour  

And the behaviour of their social network  and the behaviour that agent perceives of   their large environment the community at  large and each one of these factors have   a different weight that is assigned based  on a previous work published by us in 2017  

And the selection of modes of transport followed  two step or two-stage approach so first all modes   that person did not own were eliminated so if  a person would have a bicycle for instance so   cycling would be an option and then if two or more  mode remain then we did a utility calculation to  

Find the mode that had the highest utility for  a particular destination and again I leave these   slides to you so you can see after if you want so  it contains elements associated with the structure   or for that particular mode cost past travel  behaviour and travel time and the effects of  

The person’s network on community  behaviour or community past behaviour so some of very briefly then the results in many  lessons I won’t speak about I won’t show many   figures of tables of these paper because they’re  very complex very complicated tables and very long  

To be honest tables and figures but this is just  to show you the amount of results we had to assess   but I brought here some of the main lessons that  we got from this model so the first one is that  

Overall we saw more positive and significant  changes in low and middle country city types   so we saw much more significant decrease in car  trips road traffic that’s PM 2.5 and CO2 emissions   increases in public transport walk and cycling  trips compared with the high income city type  

And the sprawling high income city type we only  saw significant changes when the disincentive of   car use was also applied so only then we  could see decreases in car trips in road   traffic deaths and CO2 emissions and increases  in public transport walk and cycling trips  

Also we saw that some unintended consequences  could emerge so for instance when we use this   spatial inequalities in cycling walk  infrastructure in the low-income city   that will of course resulted in increases  in walking and cycling but at the same   time an increase in road traffic deaths  and a decrease in public transport trips  

We also saw that this now in five in six combined  interventions we saw super additive results and   mitigation of unintended consequences which  is a very significant important result   and also the benefits can take  at least five years to manifest  

So just to finalize here on my part of the  presentation so what some of the limitations   of our model was first that was a very simplified  spatial complex representation that we use in   our model people’s decision-making process and  weights might differ across countries of course  

And between people within the same city and we  didn’t account for this differences in our model   there might be other individual factors that  account in fact a recreation and travel behaviour   and of course these three cities don’t represent  the full spectrum of real cities even inside  

These three categories of common cities  in low-income countries in middle-income   countries and high-income countries some of the  strengths of our work was again first this was   the first model to include both physical activity  to all types of physical activity behaviour three  

Different cities with unique environmental and  socio-economic features six scenarios and we had   a range of seven outcomes I’m going to stop here  just to say that again I will share this slide   set also contains other technical aspects of  the paper if you want to see more about it too  

That’s all Ricardo thank you very much well no thank you that was  excellent so just to remind people   if you wanted to ask him questions  we’ve got a couple of minutes for   questions then please type them into the  Q & A box actually I have one how did you  

Get the social data for the cities specifically  like the social networks between them? yeah so the social network data didn’t come  specifically from any city because it wasn’t   available that easily we use information from what  have been identified through meta-analysis and  

Other systematic reviews about the average size of  a person’s social network and also some previous   work for instance dunbar number which isn’t a  contagious number but it says you know give some   indication of what is the amount of information  a person can handle that was specifically used to  

Indicate the amount of information coming from the  community at large that one agent would consider   so we use some of these information that came or  were tested through others over the literature right and one other question how did you decide  on which policy scenarios you wanted to test?

Yeah so we started using the list of eight  best investments for physical activity   it was our starting point and then we  selected three or the two ones that are   more directly connected with recreational and  transportation behaviour so that were changes  

In in the physical activity on the recreational  open spaces and parks for recreational physical   activity then public transport changes on  walk and cycle infrastructure so things are   largely associated with changes in the urban  form so this was the way we selected and we  

Aim to reduce inequalities as one thing  that was discussed and agreed internal as   a team but not could be anything else  could be not be a different strategy okay one final thing it’s a  geeky question it’s always a   geeky question for me what is the programming in? we programme in NetLogo

Right and that’s available  on your site for people? it is available on your OSF page of  the full model the link is available   in the paper but at the beginning  of the slide set I put my email   address you can contact me I’m happy to  direct to the page or even send myself

I probably will it looks a really interesting  model right I think we’re going to move on now   and the next talk is from Dr Heeseo Rain Kwon and  she’s currently a postdoctoral research fellow at   the Bartlett School of planning at the University  of London and she’s also visiting research fellow  

At the University of Reading she’s going to  be presenting ongoing research that builds an   agent-based model using spatial population  and health data from Greater Manchester   this model examines the feedback loops where  citizens switch from the main mode of transport to   non-car modes increasing physical activity leading  to better health and further encouraging mobility  

And more consumer demand for carless urban  environments leads to pro-health urban development   and land building and change that facilitates  further modes away from cars and therefore   creating a virtuous cycle well I’m really looking  forward to that and I will hand over to you Rain

Thank you very much Ricardo  okay let me share the screen right I can see that that’s super great Okay so brilliant yes my name is Rain thank  you for the introduction Ricardo so   many thanks for the opportunity to present my  research at this session so today my presentation  

Is on using ABM to conceptualize healthier  urban development and healthier travel behaviour   I don’t have the model results yet because we  are currently in the process of gathering the   interview and survey data so this presentation  will be more focused on the model design  

I will first give a quick explanation of my  approach to behavioural modelling and ABM   first to set the scene and then go dive into  this project so my broad research approach is   applying complexity theory in urban planning  to form policies for positive behavioural  

Change so I think a lot of the audience in this  webinar are probably very familiar with ABM but   yeah this involves modelling of individual  agent behaviours and their interactions in   feedback loops when system level behaviours  so agents interact system-wide patterns match  

To influence future interactions I am especially  interested in behavioural theories and Leandro has   mentioned quite a few of them so I was really  happy to see how many of them being used and I’m   very interested in linking behavioural theories  with types of behaviour data analysis methods and  

Rules or for behavioural modelling and I am very  much interested in applying that human behavioural   modelling on a spatial temporal model as an  attempt to pursue the three dimensions of human   environment dynamic modelling so it’s a spatial  temporal and human decision making complexities  

And I use agent-based modelling a lot because  it has particular strength in stimulating human   decision making dynamics and its interaction  with the environment and Netlogo is a widely   used platform for ABM as Leandro used it as well  and it conceptualizes four types of agents patches  

Turtles links and the observer and in today’s  presentation I’ll be talking about the patches   and turtles quite a lot and Sleuth urban growth  model is a widely known cellular autonoma model   in urban geography and in my previous research for  my PhD I implemented studies in Netlogo and added  

Residents on top as turtles so I’m just going to  quickly explain this because my current project   with PHASE built on my PhD model so in my PhD  model I looked at the feedback loop where more   residents switched from part to non-car this  would lead to less urbanization along the road  

Networks and then less urban sprawl would then  further encourage the remote switch to non-car   and I won’t go into details but this is the  results I got from my PhD research and the   key policy implications was about the importance  of having both top-down and bottom up approaches  

For a behavioural change and moving on to the  project that I’m currently doing with PHASE   so mode switch from car to non-car comes with  health benefits like physical activity and air   quality so quite naturally I repurposed this PhD  model with a strong focus on health and health  

Inequality for a project funded by Population  Health Agent-based Simulation Network PHASE   which is a network funded by UKPRP I’m using  data from Greater Manchester linking with the   ۲۰-minute neighbourhood concept being pursued  by the Greater Manchester combined Authority   GMCA under the strategy called streets for all  so I’m closely collaborating with the experts  

At University of Reading and Manchester for this  project just to locate the case study area Greater   Manchester is in north of England with population  around 2.8 million and there have been many   existing studies on linking physical activity  as a main basic factor or physical inactivity  

As a main risk factor of health issues and  the impact of built environments on physical   activity also the relationship between both  environment and travel behaviour as well as   how urban development decision making for  example like policy making investment in   development these affect the shaping of the urban  development shaping of the built environment  

However not many studies have looked at these  all together and that’s what we’re trying to   do in this research using complex systems theory  and agent-based modelling there are many elements   of healthy cities and this project focuses on land  use and transportation walking cycling access to  

Green area and essential functions and we talked  to some officers at the Transport For Greater   Manchester TFGM and identified a need for improved  policy understanding for promoting active mobility   and positive health outcomes in the context  of post-Covid market changes related to urban  

Land use change and the changes in the travel  demands with the increasing working from home   so this is a conceptual framework for this  model it examines the feedback loops where   the citizens switch from car to non-car modes this  increases physical activity as well as air quality  

Leading to better health and further encouraging  active mobility because people are more able   to engage in in active mobility and travel and  more consumer and user demand for carless urban   development I mean urban environment would then  lead into pro-health urban development in terms  

Of land and building use change which can be  captured in the 20-minute neighbourhood or 15   minute city metrics which then facilitates  further mode switch away from car so it’s   a impact up here and then another feedback  loop down here creating a virtuous cycle

So expert data we’re using census individual  micro data for the travel mode switch and I’m   currently performing interviews and surveys with  real estate as well as experts at transport for   Greater Manchester GMCA University of Manchester  to form the abandoned building exchange scenarios  

And in this model we are looking at residents  in employment only because the UK census asks   a question on main mode of transport only  for work and also this is in line with this   model’s focus on the impact of work from home  so this is what the model interface looks like  

On NetLogo this screen now shows simple land  use not all nature built and urban center it   is something just to illustrate what it looks  like in Greater Manchester these patches are   in 100 by 100 perfect cells so this model uses  GIS extension a lot and then uses spatial data  

So for example the building use maps are loaded by  building height for residential retail and office   multiple deprivation map by also lower super area  and a NHS disease prevalence state by GP price   points so this is the NHS quality and outcomes  framework I created a polygon maps for each the  

Eight disease types that we think that closely  relate to physical activity including obesity   coronary heart disease heart failure hypertension  COPD cancer diabetes depression and mental health   issues so I’m working very closely with an APC  candidate colleague called Matthew at Reading  

Matthew has a first degree from medical school  and produced a very comprehensive lit review on   the physical activity and health interdependencies  so building on this we are currently establishing   the roles in that model regarding the expected  impact of increased physical activity on the  

Improvements of different NCDs and so ideally in  this model we want to see how increased physical   activity can decrease the prevalence of some of  these NCDs and our hypothesis is that this would be seen differently in different parts of Greater  Manchester and in GIS data that we load on this  

Model directly links to the patch attributes  so this is like 100 by 100 meter grid on each   one of these so for example this patch here is in  rural number three yeah Greater Manchester has 10   boroughs and then has 33 meter office space  ۲۳% of the population here are deprived and  

Then it has a 2.6% of the population with obesity  and 14% with depression so we’ve also placed   ۲۱,۴۳۰ residents from the 2011 census data on the  residential patches in their boroughs so they’re   shown in blue and green dots which are their  main mode of transport green being non-car and  

Blue being car so within the boroughs we’ve  used the multiple deprivation map to locate   the residents matching with their individual  deprivation information in the census data because   of the privacy the exact residential  location of the people in the census  

Data are not given so we thought using the  deprivation index would be a good way to   assign the location for these turtles  and then each of these turtles have   distinctive socioeconomic properties based on  the individual level census data including age  

Deprivation education number of dependent children  in the household general health occupation Etc and   so on and each Turtle also has travel related  properties like distance to work number of cars   and main mode of travel to to work work from home  information place of work etc so healthier travel  

Behaviour will be measured in the metric of  motor car and as of 2011 greater Manchester   has 67% of motor car to measure healthier urban  development we’re using the 15-minute city metrics   of density proximity and diversity in line with  the Greater Manchester’s vision for 20-minute  

Neighbourhood included in its streets for all  policy so individual Turtles these residents   will have the related valuables calculated  within NetLogo like the distance to work green   area retail healthcare education entertainment  Etc so for example this specific resident here  

Doesn’t have a major cluster within two kilometre  this is zero but then has retail and office   clusters within two kilometre of its radius so  NetLogo is useful in calculating these things   very quickly also these individual turtles will  have distinct values for density and diversity  

In two kilometre radius from their residential  location so this will be population and building   density as well as diversity in terms of  building use and populating characteristics   so many of these individual resident attributes  will be used to calculate the likelihood of each  

Of the residents switching their main mode  of transport from car to non car linking with   behavioural theories most likely theories of  planned behaviour as interpersonal behaviour   and this is the rule for mode switch that  I used for my previous PhD model and I’m  

Currently building something similar to  suit the context of Greater Manchester   and this has been one of the questions that  we’ve been asking in our stakeholder interviews   in Great Manchester and the children variable I  think we will be putting a lot of weight in it

And this I’m probably not going to go  through this in detail but this is to   illustrate how the behavioural rules can  be written on NetLogo for this mode switch   behaviour these are the rules that I used  for my PhD model which I will apply quite  

Similarly in this model for Greater Manchester  and with the data loaded we see some interesting   spatial patterns already for example proximity  to nature retail office and health deprivation   which we’re examining along with the related  applications by Greater Manchester Combined  

Authority and I’ve also attempted to link the NHS  Health Data with the census individual residents   by randomly assigning the disease prevalence so  hypothetically this this Turtle has hypertension   and obesity and this is an attempt to link the  census data with the NHS disease prevalence  

Data because the NHS disease prevalence data  doesn’t have the individual information and   the census data has the individual information  but only has very brief health related questions   but I’ll be refining this methodology quite a  lot by considering some other census attributes  

Like health deprivation age etc and this would be  different for different disease types as well so   I’m currently getting lots of advice from our  health experts in the University of Manchester   so I have been experimenting with the 2011 census  data until now and the micro data for 2021 census  

Is supposed to be released soon and it’s hopefully  in the next two to three months that they said   over late spring early summer so I’m looking  forward to adding that in and making simulations   the aim is to run the model from 2021 until 2040  to test the likelihood of Greater Manchester  

Achieving its vision for 50% mode share of walking  cycling and public transport so in other words how   likely that the mode share of car currently 67%  based on 2011 data would go down to 50% by 2040.   so our aim is to also generate some top-down and  bottom-up policy implications for behavioural  

Change for both residents pro-health travel  decisions as well as real estate and planning   actors pro-health urban development  positions with regards to realizing   vision for a 20-minute neighbourhood in a  way that pursues health and health equity   and lately I’m classifying the working resident  Turtles into those who are more able to work from  

Home and those who need to commute which has a lot  of implications for socio-economic divide actually   and we are currently setting the policy scenarios  within the model based on the GMCA and TFGM’s   initiatives for active travel and also reviewing  policy we’ve reviewed latest documents to set  

Land building use change scenarios  and we’ve started our interviews to   receive feedback on this especially from  policy experts and real estate actors   this closely relates to trends like work from home  affecting retail and office buildings turning to   residential in the city centre especially with the  with the recent expansion of limited development  

Rights which doesn’t require planning permission  to switch their use from retail or office to   residential so this trend may encourage local  living and 20-minute neighbourhood lifestyle   but at the same time this would have different  effects on different people as a lower income  

Population tends to have jobs that might actually  require commuting and so I’m not going to go in   detail but this is a example of how we are asking  the experts in GM in real estate public policy  

And planning and health in terms of their feedback  on our scenarios for work from home as well as   land building use change and population location  so for example for scenario one we want to set a  

Commute five days a week scenario two three days a  week and one for one days a week or less and then   for the commute three days a week scenario we want  to set for example in classifying GM land patches  

Into three A Manchester and City centres B local  centres and C Suburban neighbourhoods and we want   to compare we want to make the difference between  non-deprived and deprived because they are very   interested in looking at the equity aspect so for  scenario two we want to say for example in A we  

Will decrease the office in retail by five percent  increase residential by five percent Etc and then   we want to make this change every year within  NetLogo so making a tick every year from 2023   until 2030 and then make the simulation run until  ۲۰۴۰ and then this would be also complemented by  

The policy scenarios related to the GM’s Streets  for All and 20 minute neighbourhood initiatives   so in summary it’ll be interesting to see  how the land and building use changes in the   built environment the work from home trend and  the 20 minute neighbourhood initiative affect  

And get affected by the residents and lead to  commute mode switch with a focus on how this all   relates to increased physical activity health  benefits and health inequalities in different   parts of Greater Manchester so this project  involves quite a lot of spatial analysis and  

The use of real data from Greater Manchester  thank you very much if I didn’t go over time no well thank you very much  that’s incredibly interesting so   people if anybody wants to ask questions  please put them up in the question answer  

Section and I’ll ask them but I have two  immediately off the back right I was really   interested in the production of your synthetic  population now I’ve got two questions for you how happy are you you’ve got a good representation  of what is Manchester in that synthetic population

So basically I’m using the CSV extension  within NetLogo to directly load the census   individual micro data so that represents around  two percent of the Greater Manchester population oh that’s interesting that’s right yes so this is I think it’s such a good  opportunity that NetLogo presents how we  

Can actually set attributes off of each  individual Turtles based on the real data is that available for a lot of  other cities that micro data? yes it is the UK census data that you can request  to receive in I think it’s Local Authority level  

Individual micro data which in the Excel or CSV  file you would have each row as a person so with   all of the questions what they answered so it’s  really very useful data and in terms of synthetic   population that’s something that I want to  actually answer is this 2011 population is  

Pre-Covid data so when we extrapolate to 2040 we  want to at first we want to use we want to have   a look at the micro data when 2021 data comes up  but then that would be quite tricky because how  

People answered during Covid times it is a unique  situation that they answered so my current thought   is probably continuing to use the 2011 data as  the pre-Covid and then post-Covid we want to   move the population based on our scenarios so that  higher income population who are able to work from  

Home who are currently living in city centre for  example or near the more urban areas or maybe make   five percent random five percent move out to more  outskirts because they can with the work from home  

And then that would make that choice happen every  year for the next 10 years or something like that   so and that would have a very interesting impact  on the equity aspect because the people who have  

That ability to make that move would be the  ones who are in higher income band as well as   who can work from home so that would  have interesting impact in terms of   how the 20 minutes or 15 minutes city  kind of neighbourhoods get formed because  

It would also have an impact on  different work from home scenarios having   more kind of affluent neighbourhoods their local  centres more flourishing with the increased demand   for local living as more people work from home in  that areas whereas the similar I guess investments  

Or demand or vibrancy wouldn’t be seen in  other parts that are more deprived so this   would be very interesting for the GMCA especially  the policy economics type of projection and how   it relates to the 20-minute neighbourhood physical  activity and also health and health inequality so  

Where this is it’s a pretty exploratory attempt  to look at these related factors holistically   applying complex systems theory and currently I’m  seeing basically a lot of different aspects that   require further investigation and this project  cannot carry on so I think it’ll probably have to  

Stop at some point and then making suggestions  for further investigation in this area Etc it’s an incredibly interesting  model I like the fact that it   probably be plug and play with different cities  but one of the things I know if Rich was here  

He’d be saying is becomes environmental  impact assessments sounds like it could   really be used nicely for that you could say  okay I’m going to change this part of the city   to something else and because you’ve got the  health I think that’d be really interesting

Right I have a question for both of you if you  don’t mind Leandro if that’s okay if you can   unmute so I’m going to ask so this question  is so both of you have talked a lot about   behavioural theory and I’m  kind of interested in how  

Easy you think or how much there is a link  between behavioural theory and translating   that into ABMs what’s been your experiences  with that? so if you want to go first I can go first my experience in general is that  these models are very linear at the beginning  

So there’s an exercise from the researcher to make  it transform that linear model into something more   dynamic of course the dynamicity is implied  but the diagram itself is very linear so I   think there’s this also the exercise so very  clearly saying well where the feedbacks are  

To make it more dynamic and more perceptual would  make them more adequate for an ABM model so this   has been my experience so far one thing that  I think it’s very useful for any model is to  

Start very early in the process put on the paper  what your mental model is about what the theory   of change or the behaviour change process is  and then start from there so of course we all  

Have lots of assumptions or informed by different  models so at a very early stage put that on paper   and then start adapting from there based on the  adequate models and theories for your per your ABM

Yeah I think I have a comment as well so I  looked at behaviour theories and how to use   it for behaviour modelling quite in detail for  my PhD research and it’s especially useful for   selecting variables that’s probably a very obvious  use of these behaviour theories and also it’s  

Important to link to see what kind of behavioural  determinants that each of these theories contain   because also in the context of the case study  area and that would involve quite a bit of unless   you’re very familiar yourself with the case study  area quite a bit of liaison with local authorities  

Or all the people there to see what kind of  factors and determinants are especially applicable   for that behaviour in that particular context like  for example just very quick example would be like   cycling behaviour in some parts of Northern  Europe could be well not northern Europe maybe  

It could be a lot of it would be connected to like  pro-environmental values that people have that’s   identified as one of the key variables whereas  in other parts of the world it’s very different   and how it relates to social class is also  very different and the relationship between  

Car use and social class also very different  in different parts of the world so that’s   something to very much keep in mind and  I also think behavioural theories are not   just useful for setting the variables but  also designing interventions or policies   to impact behavioural change like Notch  Theory Diffusion of Innovation Theory  

And when especially looking at  like a social network analysis and   Etc then what to do with the findings how to  then try to affect people to make that change well that’s great thank you I think we’ve had  two really really good speakers there thank you  

Very much indeed to Leandro and Rain and show your  appreciation with the virtual things on the side

ID: i1X0P3y_BLM
Time: 1682596327
Date: 2023-04-27 16:22:07
Duration: 00:59:26

منبع