Climate Change and Health Seminar Series: "Climate change, air pollution, and public health: Bridging science to policy"

Transcript

• 00:00Today is my great pleasure and
• 00:04to introduce to this speaker,
• 00:06doctor Susan Annenberg.
• 00:08Susan is associate professor and chair
• 00:11of the Department of Environmental
• 00:14and Occupational Health in the George
• 00:18Washington University and she's also
• 00:20the current director of the DW PAN and
• 00:24Health Initiative Institute and Doctor.
• 00:27August Research focus on the
• 00:29health implications of air
• 00:30pollution and climate change,
• 00:32from both local to global skills.
• 00:35And we talk a lot about
• 00:37like policy implications.
• 00:38And Doctor Annenberg is really
• 00:40the true pioneer of making the
• 00:43science of this policy relevant.
• 00:46So she serves on the US EPA Science
• 00:49Advisory Board and the Clean Air
• 00:52Act Advisory Committee and The Who
• 00:55Global air pollution and Health
• 00:58Technical Advisory Group and the
• 01:00National Academy of Sciences Committee
• 01:02to advise the US Global change.
• 01:05Research programs.
• 01:06She also serves as currently the
• 01:09President of the Jail House section
• 01:11of the American Geophysical Union.
• 01:14So without first deal,
• 01:15let's welcome those energy.
• 01:19And for being here today,
• 01:21I really appreciate you taking the time
• 01:23out of your days to to to be here.
• 01:25So I'm Susan Annenberg from
• 01:27George Washington University,
• 01:28and I will be talking today
• 01:31about linking climate change,
• 01:32air pollution and human health and
• 01:34bridging science to the policy,
• 01:35which is really what I'm
• 01:37very passionate about doing.
• 01:44Shape that, people.
• 01:47OK, so before I start,
• 01:50let me just say that a lot of the work
• 01:52that I'm going to show today is really
• 01:53standing on the shoulders of giants.
• 01:55This is work that would not be
• 01:57possible without the people who have
• 01:59spent many years that, you know,
• 02:01detecting associations between air
• 02:03pollutants and health outcomes,
• 02:05developing air pollution exposure
• 02:06datasets that are open and publicly
• 02:09available for others to use.
• 02:10And I appreciate the efforts
• 02:12of many people in this room and
• 02:14contributing to that science.
• 02:15And this really makes the the bridging.
• 02:18From science to the policy possible
• 02:20by creating these datasets and
• 02:23associations that others can use.
• 02:25But based on the information that
• 02:27we have from Epidemia epidemiology
• 02:29and exposure science,
• 02:31we know that air pollution continues
• 02:33to be a leading health risk
• 02:35factor in nearly all countries.
• 02:37Is currently considered to be
• 02:39the 4th leading risk factor
• 02:40affecting global mortality.
• 02:42That's not the 4th leading
• 02:44environmental risk factor.
• 02:45That's the 4th leading overall risk factor.
• 02:48And and really indicates that air pollution
• 02:50needs to be central on the global health
• 02:53agenda for improving people's health.
• 02:55And if you look at the diseases
• 02:57that air pollution impacts,
• 02:58it is not a small fraction of these diseases
• 03:01that air pollution is responsible for.
• 03:03I mean this is a, you know,
• 03:0640% of chronic obstructive pulmonary disease,
• 03:0820% of diabetes,
• 03:0920% of ischemic heart disease and you can
• 03:11read the rest of the the percentages there.
• 03:14So this is a preventable risk
• 03:16factor that it is responsible for.
• 03:19Millions of premature deaths
• 03:20globally and a very large,
• 03:22very substantial fraction of the incidence
• 03:26of these diseases around the world.
• 03:29And we also know that climate
• 03:31change worsens air pollution.
• 03:33So climate change is contributing
• 03:35to worsening ozone,
• 03:36increased wildfire smoke,
• 03:38increased dust,
• 03:39worsened allergy conditions,
• 03:41and even potentially impacting
• 03:43airborne infectious diseases,
• 03:45both the spread and the severity
• 03:47of airborne infectious diseases.
• 03:49So air pollution and climate
• 03:51change are highly interlinked.
• 03:53This is just one of the ways
• 03:54that they're interlinked,
• 03:55and we're going to talk about
• 03:56some of the others.
• 03:57But climate change is now worsening.
• 03:59Air pollution making it harder
• 04:01for us to protect the air,
• 04:02making it healthy for people
• 04:05to breathe and and.
• 04:06One of the ways that we one of the most
• 04:09prominent effects of climate change on
• 04:11air pollution is now wildfire smoke.
• 04:14So I just want to look at some
• 04:16of the most recent work that I
• 04:17was a part of looking at PM 2.5,
• 04:20you know,
• 04:21very fine particle concentrations
• 04:22across the United States,
• 04:24the Eastern US and then the Western
• 04:26US and we see across the the last
• 04:29couple decades for the United States,
• 04:31PM 2.5 concentrations have
• 04:33been declining substantially.
• 04:34That's a huge public health win and.
• 04:37Is the result of many years of
• 04:40effective regulations under the
• 04:42Clean Air Act in the United States.
• 04:44So PM 2.5 concentrations have been
• 04:46declining substantially and even more
• 04:48substantially in the eastern US,
• 04:50where we have very strong anthropogenic
• 04:52emissions that have been controlled
• 04:53from our power plants and our vehicles
• 04:55over the last couple of decades.
• 04:57And we've seen this very dramatic decrease.
• 04:59Again, 2.5 crowded Eastern
• 05:01US in the western US,
• 05:03we have a different story here with a lot
• 05:05of interannual variability in those PM 2.
• 05:0725 concentrations in the last 5-10 years,
• 05:10and that's driven by wildfire smoke.
• 05:13If you draw a line through this
• 05:15very large interannual variability,
• 05:16you see that PM 2.5 concentrations are
• 05:18actually increasing in the Western US,
• 05:20despite the very effective
• 05:21regulations that we have on power
• 05:24plants and industry and vehicles.
• 05:25And that different disparate
• 05:27picture between the western US and
• 05:30the eastern US is driving what we
• 05:32see here for the US on average,
• 05:34that we actually see that the PM 2.5
• 05:37concentrations are beginning to flatten out.
• 05:39They're not declining to the same
• 05:41degree as they have been for
• 05:42the past couple of decades.
• 05:43We're actually seeing that they're starting
• 05:45to stagnate in the coming decades.
• 05:47We might actually start to see that
• 05:49they're starting to rise again.
• 05:50And this makes it more difficult for
• 05:52us to attain our national ambient
• 05:54air quality standards for PM 2.5
• 05:56because of this climate induced
• 05:59change and wildfire smoke keeping
• 06:02PM 2.5 concentrations high.
• 06:03I had the honor of working with the
• 06:06US Environmental Protection Agency on
• 06:08their climate change impacts and risk
• 06:11analysis project, their Sierra project.
• 06:13I used to actually work at the
• 06:15EPA from 2010 to 2014,
• 06:17and when I was there,
• 06:18we were starting this project to quantify
• 06:20the different damages of climate
• 06:22change on life in the United States,
• 06:24and that includes air pollution,
• 06:26but it also includes a lot of other
• 06:27things like labor and extreme temperature
• 06:29mortality and coastal property
• 06:31and roads and back at that time,
• 06:33the only.
• 06:34Estimate of how climate change impacted
• 06:37air pollution and therefore damages
• 06:39through human health in the US was via ozone.
• 06:43So temperature worsens ozone and that
• 06:47contributes to premature mortality.
• 06:50And you can see that that back in about 2014,
• 06:542015,
• 06:54air quality was the 4th largest
• 06:57damage of climate change in the
• 06:59United States once valued.
• 07:01But we really recognized, you know,
• 07:03we also think that climate change
• 07:05is influencing.
• 07:05Yeah,
• 07:062.5 and PM 2.5 has a very strong
• 07:08relationship with premature mortality.
• 07:10So if we were able to quantify the
• 07:12impacts of climate change on PM 2.5,
• 07:14in addition to ozone,
• 07:15we likely would get an A large,
• 07:18potentially a larger number.
• 07:19We would likely get a different number.
• 07:21Back in 2015,
• 07:23climate models were still very uncertain
• 07:25about where the precipitation happens,
• 07:28what's going to happen to PM
• 07:292.5 in different locations.
• 07:31And that still remains a a big uncertainty.
• 07:33But we do know that climate
• 07:35change is driving.
• 07:36Quote UN quote natural sources of PM 2.5,
• 07:39which are no longer,
• 07:40I think can no longer be considered
• 07:43fully natural anymore because
• 07:45climate change is impacting them.
• 07:47So dust exposure in the southwest US,
• 07:50Wildfire PM 2.5, which we just talked about.
• 07:53So I partnered with the EPA and a
• 07:55number of other scientists and we
• 07:57quantified the potential damages of
• 07:59climate change on dust exposure and
• 08:02therefore premature mortality in the US
• 08:04and same with wildfire smoke exposure.
• 08:07And we valued that.
• 08:08And we came up with about $47 billion a
• 08:11year from climate induced contributions
• 08:14to dust exposure and its effects on
• 08:17premature mortality and about $25
• 08:19billion a year for wildfire smoke.
• 08:21And if you add those together
• 08:23with the ozone impact that we had
• 08:25previously quantified,
• 08:26we see that air pollution is one
• 08:28of the largest damages of climate
• 08:31change in the United States.
• 08:33And this is an estimate that I
• 08:35think is likely to grow, I think.
• 08:37We underestimated this impact due to the
• 08:39methods that were available at the time,
• 08:41so I think this number is likely
• 08:43to get larger.
• 08:44Another reason it's underestimating
• 08:46the damages of climate change
• 08:48on air pollution is because.
• 08:51We can't just add together the impacts
• 08:53of heat on mortality and the impacts
• 08:56of air pollution on mortality.
• 08:58These actually have synergistic effects.
• 09:00So the total impact of increased
• 09:02heat and increased air pollution
• 09:04is more than the sum of its parts.
• 09:07In the previous slide I just showed you
• 09:09we were only capturing the impact of
• 09:10each of these risk factors individually,
• 09:12not considering the others.
• 09:14But because we know that there
• 09:16are these synergistic effects,
• 09:17we're likely missing some of these
• 09:19damages of both heat exposure.
• 09:22And air pollution.
• 09:23And as more research comes out looking
• 09:25at the pollen impacts as well,
• 09:26I think that could be potentially
• 09:28a factor to consider here too.
• 09:31So I talked about how there's
• 09:33different links between climate
• 09:35change and air pollution.
• 09:37We talked about this one,
• 09:38how climate change can impact air pollution.
• 09:40Air pollution can also impact climate change.
• 09:42We have short lived climate pollutants,
• 09:44for example black carbon and
• 09:46methane that contributes to poor
• 09:48air quality and warm the climate.
• 09:50This arrow here is.
• 09:52Sorry should go from climate change
• 09:54to public health.
• 09:55Not that the other association between
• 09:57climate change and air pollution
• 09:59that I want to talk about is how
• 10:01they share the same emission sources.
• 10:03Anytime we burn anything,
• 10:05primarily fossil fuels but also biofuels,
• 10:07we're releasing both airplanes
• 10:09and greenhouse gases.
• 10:10So if we want to address climate
• 10:12change and air pollution,
• 10:13we should be reducing the amount of
• 10:16fuel that is burned and therefore
• 10:19addressing those emission sources.
• 10:22What we've done so far in the
• 10:24United States by to.
• 10:25Bring down our PM 2.5 levels.
• 10:27We've tried to break this arrow
• 10:29between emission sources to air pollution.
• 10:31So we put catalytic
• 10:32converters on our vehicles.
• 10:33We put diesel particulate
• 10:34filters on our trucks,
• 10:35scrubbers on our power plants,
• 10:36and these have been very effective
• 10:38at reducing the amount of pollution
• 10:40from these emission sources.
• 10:42But they've done nothing to this era here.
• 10:44We're still continuing to make
• 10:46greenhouse gases largely unabated,
• 10:47and that climate change is contributing
• 10:49to the air pollution problem.
• 10:51So if we want to again mitigate both
• 10:54air pollution and climate change.
• 10:56We need to be burning less stuff,
• 10:57primarily fossil fuels,
• 10:59but also biofuels.
• 11:00I have focused a lot of my work,
• 11:02especially the most recent years,
• 11:03on the urban context,
• 11:04and the reason for that is because
• 11:06a lot of cities around the world
• 11:08are experiencing poor air quality.
• 11:09This is just a map of nitrogen
• 11:11dioxide concentrations in the US,
• 11:13but a lot of cities around the world are
• 11:16experiencing much greater levels of of
• 11:18pollution than we do in cities in the US,
• 11:21especially in cities in Africa and Asia,
• 11:24which are rapidly growing.
• 11:26These are experiencing
• 11:27rising air pollution levels.
• 11:30They're also, cities are also
• 11:31experiencing CO2 emissions growth.
• 11:33Right now,
• 11:33cities are responsible for about 3/4
• 11:35of global greenhouse gas emissions,
• 11:37and that's projected to rise as
• 11:40the world continues to urbanize.
• 11:42We also have very strong
• 11:44health inequality effects.
• 11:45So this is a map of Washington,
• 11:47DC, where I live.
• 11:48And the green colors here show the
• 11:50pediatric asthma emergency department
• 11:52visit rate for 10,000 people.
• 11:54And the red dots show life expectancy.
• 11:57We have about a 20 year life expectancy
• 11:59differential between neighborhoods
• 12:00in the southeast quadrant of the city
• 12:03right here versus neighborhoods in
• 12:05the northwest quadrant of the city.
• 12:0720 year life expectancy differential
• 12:09between people that live about
• 12:112 miles away from each other.
• 12:12We also have very dramatic differences in
• 12:16pediatric asthma Ed visit rate as well.
• 12:19So this is just, you know,
• 12:21DC is not unique.
• 12:22We have problems for sure,
• 12:23but we're not unique.
• 12:24Most of the cities across the country
• 12:26are experiencing problems like this and
• 12:28then we have growth growing populations.
• 12:30So right now about half the world's
• 12:32population lives in urban areas.
• 12:34That's expected to grow to about 2/3 by 2050.
• 12:37And nearly all of that increase is
• 12:40anticipated to happen in African
• 12:42and Asian cities, where, again,
• 12:44pollution levels are also continuing to rise.
• 12:47So there's a lot of problems happening
• 12:50simultaneously in the urban context,
• 12:52and if we were to address the
• 12:54way that our cities burn fuel,
• 12:57we likely would be able to get
• 12:59at multiple of these problems.
• 13:02What we what we can't see,
• 13:04we can't fix.
• 13:05We have to be able to see the
• 13:07pollution in order to fix it.
• 13:09Right now this is where the monitoring
• 13:11happens for air pollution around
• 13:12the world you can see most of the
• 13:15monitors are in the US and Europe,
• 13:17and increasingly in China and in India.
• 13:20But much of the world is left uncovered.
• 13:23And even in places that look like
• 13:25they're densely covered by monitors,
• 13:27like take Washington DC,
• 13:29we only have 5 monitors,
• 13:31looks like 4, but two.
• 13:33We only have 5 monitors for the
• 13:35entire city of Washington DC,
• 13:37so how are we supposed to
• 13:39capture the inequality
• 13:40and pollution levels if we if these are
• 13:44this is our only source of information?
• 13:47Luckily, we have a new source of information
• 13:50which is Earth observing satellites.
• 13:52So NASA, the European Space Agency
• 13:54and other space agencies around the
• 13:56world have been launching satellites
• 13:58and they are constantly taking
• 14:00pictures about miseric composition.
• 14:02And we can tease out that information
• 14:05and understand what are people exposed
• 14:07to in places that have no monitors.
• 14:09This is a map of what nitrogen dioxide
• 14:12look looks like from the Tropo ME
• 14:14sensor on the Sentinel 5P satellite
• 14:15from the European Space Agency.
• 14:17Um, this map was created by Dan
• 14:19Goldberg and you can see where N 2 is
• 14:22the highest and the fact that we have
• 14:24the full geospatial coverage here.
• 14:25So we can get beyond the monitors,
• 14:27we can get beyond the monitors
• 14:29and see what people are exposed
• 14:31to all around the world.
• 14:32So what does satellite data look like?
• 14:34Well, this is a daily snapshot of
• 14:36satellite and O2 Tropo mean No2 nitrogen
• 14:39dioxide that Dan Goldberg created.
• 14:41It's, this is now available on our website.
• 14:45You can download for every day.
• 14:46It's automatically putting up this image
• 14:48of No2 concentrations over the US and
• 14:51over different regions of the US and you
• 14:53can see there's a lot of white areas,
• 14:56right?
• 14:56These are where clouds are.
• 14:57So the satellites can't see through clouds.
• 14:59We're still limited in that way
• 15:01and there's also a lot of noise.
• 15:03This is just one snapshot per day.
• 15:05The TROPONE sensor is polar orbiting.
• 15:08That means it goes around the
• 15:10earth and it takes an image of the
• 15:13atmospheric composition at about 1:30
• 15:14PM everywhere on Earth local time.
• 15:16So just the one snapshot per day,
• 15:19and this is what it produces.
• 15:20Pretty noisy.
• 15:21But when we start to average
• 15:23over longer time periods,
• 15:24we have a lot more data and
• 15:26it starts to look more smooth.
• 15:28So this is a season of data of N2
• 15:32concentrations over the US and then
• 15:35the comparison for 2021 to 2019.
• 15:37And again you can get this on our website.
• 15:44So what we can do with the full
• 15:46geographical coverage of satellite
• 15:48data and increasingly high spatial
• 15:49resolution as well is that we can
• 15:51start to tease out what is happening
• 15:53in all urban areas globally.
• 15:56And there's about 13,000
• 15:57urban areas globally.
• 15:59So we can use that continuous spatial
• 16:02map that we get from satellite data
• 16:05and integrate and aggregate that
• 16:07up to the urban areas from Veronica
• 16:10Sutherlands and Ross Mohegan and.
• 16:12Danny Balashov,
• 16:13who have all worked with me,
• 16:15have done this for PM 2.5,
• 16:17for N2 and for ozone.
• 16:19So we now have available on a Nice
• 16:22website as well interactive website
• 16:24the the levels of these three
• 16:26pollutants as well as their trends
• 16:28overtime and their contributions to
• 16:29the burden of disease in those cities
• 16:32for all 13,000 cities globally.
• 16:33We've given the data to the health
• 16:35Effects Institute who runs the
• 16:37state of Global Air project and
• 16:38they have published this report,
• 16:40air quality and health in cities
• 16:42for the first time.
• 16:43Making the data more available for
• 16:45cities around the world to use.
• 16:47And, you know,
• 16:47I think it's important to note
• 16:49that in most of these,
• 16:50probably the vast majority
• 16:51of these 13,000 cities,
• 16:53there is no air quality monitoring.
• 16:55So this is the first time that
• 16:57there's really any estimate of the
• 16:59pollution levels in those cities.
• 17:01They're likely to be very uncertain,
• 17:03probably wrong in a lot of different ways,
• 17:05but at least it's a first,
• 17:06you know, first guess at what,
• 17:09an educated guess at what pollution levels
• 17:11are driven by the observations from these.
• 17:14The lights.
• 17:16The other thing we can do
• 17:17with the satellite data,
• 17:18with the continuous coverage,
• 17:19the continuous geospatial
• 17:20coverage from the satellite data,
• 17:22is that we can get at what is
• 17:24happening within individual cities.
• 17:26And again,
• 17:27we know that cities are experiencing
• 17:29health inequality issues.
• 17:30There's a long history of science
• 17:32telling us that air pollution
• 17:34levels are inequitably distributed
• 17:36within cities as well.
• 17:38But again,
• 17:38we can't get that just from
• 17:40the four or five monitors that
• 17:41we have in individual city.
• 17:42So we need to use the,
• 17:44you know,
• 17:45we need to use approaches for
• 17:47estimating pollution levels between
• 17:48those monitors to understand
• 17:50inequality and air pollution levels.
• 17:52So this is a study led by Maria Castillo,
• 17:55who's now an urban planning student at MIT.
• 17:58And we partnered with the
• 18:00DC local government,
• 18:01the DC Department of Energy and Environment
• 18:03and the Office of HealthEquity,
• 18:05who had they had.
• 18:07Settlement funds from the
• 18:09Volkswagen Diesel gate scandal.
• 18:11Anyone remember in 2015 there
• 18:14was a big revolution that.
• 18:17Volkswagen vehicles were equipped
• 18:18with these defeat devices,
• 18:20pieces of software that would turn
• 18:22the emission control equipment on
• 18:24when the vehicle was undergoing
• 18:26regulatory testing of emissions and
• 18:28then off when they were being driven
• 18:30around in real world driving conditions.
• 18:32And that was leading to substantially
• 18:34higher orders of magnitude higher
• 18:36NOx emissions in real world
• 18:38driving conditions than during
• 18:40certification testing.
• 18:41So there's a big lawsuit,
• 18:42people went to jail and now cities
• 18:45have access to settlement funds.
• 18:47They can use to direct resources
• 18:48to improve air quality.
• 18:49So the DC government had settlement
• 18:51funds and they came to us and they said,
• 18:53can you help us understand how air
• 18:55pollution is contributing to the health
• 18:57inequality problem in the city so
• 18:58that we might be able to direct these
• 19:00resources to places that are overburdened?
• 19:03So we estimated PM 2.5 attributable
• 19:06mortality using one of those continuous
• 19:08data sets of PM 2.5 in this case from
• 19:10the Washu group led by Randall Martin.
• 19:13And we estimated PM 2.5 mortality rates and.
• 19:17We saw that the highest PM 2.5
• 19:19mortality rates occurred in the
• 19:20eastern half of the city,
• 19:21and lower PM 2.5 mortality rates
• 19:23in the western half of the city.
• 19:25And this lined up almost exactly
• 19:26with the map of racial segregation,
• 19:28segregation in the city,
• 19:29so the eastern half of the cities
• 19:31primarily black and the western
• 19:33half of the cities primarily white.
• 19:35I know what that noise is.
• 19:39This research was received some
• 19:41interest from NASA and they created
• 19:43this really nice looking map and
• 19:46they made it the image of the day
• 19:49on the NASA Earth Observatory,
• 19:51which was really cool.
• 19:52And because so many people follow
• 19:54the NASA Earth Observatory, if you
• 19:56don't you should on Instagram or you know,
• 19:58whatever your social media choices.
• 20:00They posted it there and it got picked up
• 20:03by another influential Instagram accounts.
• 20:06Washingtonian probs.
• 20:07Which has hundreds of thousands of followers.
• 20:10So this was a way that our,
• 20:12you know study which was published in an
• 20:16esoteric journal Geo Health was picked
• 20:18up and brought to people who wouldn't
• 20:21normally read papers of Geo Health.
• 20:23And I know they tell you,
• 20:24you never read the social media
• 20:25comments about you know your work.
• 20:27But you know in a lapse of judgment
• 20:29one day I decided to read those
• 20:31comments and anyone want to take a
• 20:32guess at the most frequent comment
• 20:34of the Washingtonian probs account
• 20:36got when they when they posted?
• 20:39Yes. Thank you for your work.
• 20:43That would be nice.
• 20:46The most frequent comment was done.
• 20:50So you know, I think people know this,
• 20:52people know that air pollution
• 20:53is inequitably distributed.
• 20:54But again, if you don't
• 20:56show it with data and maps,
• 20:57then it's difficult to address.
• 21:00In this case, again, we work directly
• 21:01with the DC local government.
• 21:03So it was a way that they were able
• 21:05to help us design the study to answer
• 21:07the question that they had and then
• 21:09they can use the results to, you know,
• 21:11to determine how they're using those,
• 21:13those settlement funds.
• 21:14Fun Facts on Friday I just
• 21:17recorded a video at NASA studio,
• 21:19NASA Goddard Space Flight Center.
• 21:21They are going to now have it now and
• 21:24then that the lobby of NASA headquarters,
• 21:26a giant screen with me talking
• 21:28about this study.
• 21:29And they did not tell me my face was going
• 21:32to be up there at the size of my course.
• 21:34So I'm not excited about that.
• 21:35But I'm excited that they are,
• 21:38that they're highlighting this important
• 21:39work because I really think it does
• 21:41show the value of satellite data and
• 21:43what it can tell us in terms of.
• 21:45Real world's problems that
• 21:47we're experiencing in cities.
• 21:49So that was PM 2.5 and CO2 is a pollutant
• 21:52that a lot of us don't think that much about.
• 21:55We often think about PM 2.5.
• 21:57That's the largest contributor to the
• 21:59burden of disease from air pollution,
• 22:01followed by ozone.
• 22:03And O2 is a precursor to both
• 22:06PM 2.5 and ozone.
• 22:07So if we want to address those pollutants,
• 22:09we have to know where the
• 22:11No2 is and and control it.
• 22:13It's also a high resolution tracer for urban
• 22:16traffic in particular it's associated itself.
• 22:20With asthma development,
• 22:21that's just not that's not
• 22:22just asthma exacerbation,
• 22:24but new development of asthma among children.
• 22:27And very conveniently it is highly correlated
• 22:30satellite and O2 is highly correlated
• 22:33with ground level O2 from monitors.
• 22:35So this,
• 22:36for example,
• 22:37is a scatter plot created by my
• 22:39colleague Dan Goldberg and Gage Kerr
• 22:41who showed that trouble me No2 columns.
• 22:44That's the amount of N2 in the
• 22:45column of air between the satellite
• 22:47and the surface of the Earth.
• 22:49Is highly correlated to N2 at the
• 22:51ground level monitor monitored by our
• 22:53AQS monitor monitors our air quality
• 22:56system monitors and so this makes it
• 22:58a very convenient pollutant to study.
• 23:00Whereas for PM 2.5,
• 23:02the satellites are monitoring
• 23:03at different quantity,
• 23:04aerosol optical depth and then we need
• 23:05to do a bunch of science to convert
• 23:07that to ground level PM 2.5 here,
• 23:09even if we just took the Tropo
• 23:11Vienna 2 columns,
• 23:12we have a pretty good sense for
• 23:14where where the ground level 2 is.
• 23:17So around the time that the pandemic hit,
• 23:20we had just hired Dr.
• 23:22Gage Kerr as a postdoc and we
• 23:26were wondering whether or not we
• 23:28could use these troponin data.
• 23:31So troponin data started that
• 23:33the records started in 2018,
• 23:35so it was very new.
• 23:37And you know, when the
• 23:39pandemic hit in spring 2020,
• 23:41Dan Goldberg had been going through
• 23:43these energy readings and looking at
• 23:44different urban areas and seeing how the
• 23:46trends differed in different cities.
• 23:48And we wondered, you know,
• 23:49could we use this data set to explore
• 23:52how No2 changed during the pandemic?
• 23:55There are a lot of people working on air
• 23:57quality changes during the pandemic.
• 23:58Of course,
• 23:59there's a whole community of people.
• 24:01We actually got on the phone
• 24:03once a month talking about air
• 24:04quality changes during COVID.
• 24:05But we wanted to take this a step
• 24:07further and really leverage the
• 24:09value of the satellite data with
• 24:11that complete geospatial coverage.
• 24:12And one of the, you know,
• 24:14values of that satellite data is the
• 24:16fact that we can look within cities,
• 24:17different subpopulations.
• 24:18Living within cities.
• 24:20So we had no idea whether we could use
• 24:22this data set to explore disparities.
• 24:24And I know two concentrations,
• 24:26but we we thought,
• 24:27let's just give it a shot, see what happens.
• 24:30Probably we won't see anything.
• 24:32Well,
• 24:32it turned out we did see
• 24:33something and it was really,
• 24:35really striking to me.
• 24:36So prior to the pandemic in 2019,
• 24:39the least white census tracts
• 24:41across the United States had no
• 24:43two concentrations that were
• 24:45about double the concentrations
• 24:46and the most white census tract.
• 24:49Again, that's prior to the pandemic.
• 24:52During the lockdowns in 2020,
• 24:54both the orange dots and the
• 24:56blue dots shifted left,
• 24:57and that indicates that No2 dropped
• 24:59for both the least white census tracts
• 25:01and the most white census tracts.
• 25:03Just good thing,
• 25:04you know,
• 25:04we had about 50% fewer passenger
• 25:06vehicles on the road.
• 25:08It's a good thing that we can
• 25:10observe and O2 just by itself.
• 25:11That was useful to know that we
• 25:14could use this tromi data set
• 25:16to observe that drop in and O2
• 25:18during this natural experiment.
• 25:21But.
• 25:21One thing that we found that
• 25:23was really concerning was that
• 25:25during the 2020 lockdowns,
• 25:27then O2 concentrations in the
• 25:28least white Census tracts were
• 25:30still about 50% higher than the
• 25:32concentrations and the most white
• 25:34census tracts prior to the pandemic.
• 25:37And this indicates that the
• 25:39disparities in antipollution were
• 25:40so large prior to the pandemic that
• 25:43even about a 50% drop in passenger
• 25:45vehicle traffic was not enough
• 25:47to eliminate those disparities.
• 25:49And that held,
• 25:50that pattern held for nearly all
• 25:52major cities across the US and also
• 25:55held for educational attainment
• 25:56and for income.
• 25:57But really,
• 25:58that only tells us about exposure.
• 26:01We're really just concentrations,
• 26:02not even exposure.
• 26:03It doesn't tell us about the
• 26:05susceptibility of the population.
• 26:07That is breathing those concentrations.
• 26:09So Gage took this a step further
• 26:12and looked at both PNC .5 and N2,
• 26:15and not just the concentrations,
• 26:16but the health outcomes that are
• 26:19associated with those concentrations.
• 26:20So he's comparing PM 2.5 attributable
• 26:23mortality per 100,000 people and
• 26:26that NATO attributable pediatric
• 26:28asthma incidence rate as well.
• 26:30And let's just look at
• 26:32PM 2.5 first. We see that PM 2.5
• 26:34concentrations are dropping over time for
• 26:36both the most white and the least white
• 26:38census tracks across the United States.
• 26:40This is very similar to the graph I
• 26:42showed you at the beginning, showing
• 26:43that PM concentrations are going down,
• 26:45starting to stagnate a little bit
• 26:47due to those Western US fires.
• 26:49But the disparities persist,
• 26:50as many others have found
• 26:52in the literature as well,
• 26:54that PM 2.5 concentrations and
• 26:56associated disease burdens are higher
• 26:59for the least weight census tracts,
• 27:01and then they are for the
• 27:03most white census tracts.
• 27:04And the relative disparity,
• 27:06the relative ratio between blue
• 27:07dots and the orange dots here,
• 27:09is actually rising over time.
• 27:11So the relative disparity is getting
• 27:13worse even though the levels are coming
• 27:15down for both populations subgroups.
• 27:17For No2,
• 27:18on the right hand side here we see
• 27:20that no two and its associated impact
• 27:23on asthma incidents among children
• 27:25is also decreasing over time,
• 27:27again to very successful
• 27:28regulations under Clean Air Act.
• 27:30But the disparity is much,
• 27:32much larger than it is for PM 2.5.
• 27:34In fact,
• 27:35the relative disparity is about 7 1/2,
• 27:37meaning that the most the least
• 27:39white census tracts have values
• 27:41that are about 7 1/2 times larger
• 27:43than the most white census tracts,
• 27:46whereas that value is only 1.3.
• 27:48For PM 2.5 not to diminish 1.3,
• 27:51that's still 30% larger PM
• 27:53mortality impacts for the least
• 27:55white Census tracts compared to
• 27:56the most white census tracts,
• 27:58but no two exhibits far greater
• 28:01disparity than PM 2.5 does.
• 28:04Now,
• 28:04all all of this that I've just
• 28:07showed you is based on one expose 1
• 28:10concentration data set per analysis.
• 28:12And there's a lot of people working on a
• 28:14lot of different concentration data sets,
• 28:15both PMC .5 and No2,
• 28:17and we don't know which one is the best.
• 28:19People are using different methods,
• 28:21they're using different approaches,
• 28:22different data inputs.
• 28:23And so we wanted to know how much
• 28:25is of the result that we that I just
• 28:28showed is actually driven by features
• 28:29of the one data set that we used as
• 28:32opposed to other datasets where we
• 28:34find this across multiple datasets.
• 28:36So gauge is now comparing No2 disparities
• 28:41for four population subgroups using
• 28:43the EPA air quality system regulatory
• 28:46monitors on the left hand side here.
• 28:48For the 10 most populous cities in the US,
• 28:51the numbers on the right show the
• 28:53number of monitors in those cities.
• 28:55And we see a pattern that's
• 28:56kind of all over the place,
• 28:58in fact no pattern.
• 28:59So this these air quality system
• 29:01monitors are not able basically
• 29:04to capture the disparities that
• 29:06we think exist and that a lot of
• 29:08other studies have found to exist.
• 29:10When we use a land use regression
• 29:12model for nitrogen dioxide,
• 29:14which uses statistical approaches
• 29:16to approximate No2 concentrations
• 29:18at pretty high resolution across
• 29:21the entire continental US,
• 29:23we see a stronger pattern pop out here.
• 29:28So for every major city we
• 29:30have the the lowest
• 29:31No2 concentrations in the non
• 29:33Hispanic white population and higher
• 29:36concentrations among the Hispanic,
• 29:38Asian and black populations.
• 29:40The ordering. Differs by by city,
• 29:42but it's very similar to what we find
• 29:45using just the troponin No2 columns.
• 29:48So this is the land use regression model.
• 29:50Approximates surface level
• 29:52and O2 concentrations.
• 29:54The Tropo me data is No2 columns that
• 29:58are more directly from the satellite
• 30:01and we see a very similar pattern here.
• 30:03We see that for both the non Hispanic white
• 30:06population has the lowest No2 concentrations.
• 30:08For some cities we see that.
• 30:10Ordering of the population
• 30:12subgroups is very similar,
• 30:13so in Philadelphia the
• 30:15ordering is very similar.
• 30:16In other cities we see differences,
• 30:18but nevertheless there's much closer
• 30:21consistency between the land use
• 30:22regression data set and the troponin data
• 30:24set compared with the monitor data set.
• 30:26It's really not surprising.
• 30:28I mean the monitor data set was not
• 30:30intended to be used for this purpose
• 30:32and we're really was intended to
• 30:34monitor regional average pollution
• 30:36and not neighborhood scale pollution
• 30:38that differs within cities.
• 30:40So that was for No2.
• 30:42That would really LED us to wonder,
• 30:43OK,
• 30:44well the data set that you use for N2 has
• 30:47a big impact on the estimated disparities.
• 30:50What about for PM 2.5,
• 30:52which is a prudent that doesn't vary
• 30:55as much spatially as an O2 does,
• 30:56and the two has a very
• 30:58short atmospheric lifetime,
• 30:58it stays pretty close to the mission source.
• 31:01PM 2.5 has a lot more emission sources.
• 31:05A lot of it is secondarily formed
• 31:07in the atmosphere.
• 31:08It lives longer in the atmosphere,
• 31:09so it spreads out and sort
• 31:11of smooth spatially.
• 31:12So we but there's a lot
• 31:14of attention on PM 2.5,
• 31:15right,
• 31:16the Justice 40 initiative of
• 31:17this current administration.
• 31:19This is a new initiative that
• 31:20is aimed at 40% of the benefits.
• 31:23Of federal investments going
• 31:25to disadvantaged communities,
• 31:26the data set they're using to do that,
• 31:28to identify communities as
• 31:30disadvantaged as a 12 kilometer
• 31:33spatial resolution for PM 2.5.
• 31:35That's this CMAC model monitor
• 31:38fusion data set.
• 31:39That's the one that's used in EJ screen.
• 31:41It's used in a lot of EPA regulatory
• 31:43support documents and now it's used
• 31:45in the climate and economic justice
• 31:47screening tool suggest that is
• 31:48used for the Justice 40 initiative.
• 31:49So we wondered,
• 31:50if we used a different high resolution
• 31:53data set that's now available
• 31:55from the scientific community,
• 31:57would that lead to differences in
• 31:59which communities are flagged as
• 32:00disadvantaged in the Justice 40 initiative?
• 32:03So we're now comparing.
• 32:06The CMAC Model monitor fusion data
• 32:08set at 12 kilometer spatial resolution
• 32:10with the the data set I talked
• 32:13about earlier from the Washu team,
• 32:15the bins unclear at all data set that
• 32:19fuses satellites with a geophysical model.
• 32:22And then there's a new data set led
• 32:24by Haresh mini that's available at 50
• 32:27meter resolution within cities and
• 32:291 kilometer resolution outside of cities.
• 32:32And you can see just looking
• 32:33at the spatial resolution,
• 32:34the spatial distribution
• 32:35in Los Angeles at the top,
• 32:37Chicago in the middle and Phoenix
• 32:39on the bottom. These datasets,
• 32:40they look somewhat similar in terms
• 32:42of their being a BLOB over the city.
• 32:45But once you start to look a little bit
• 32:47closer, they really differ in terms of
• 32:50which neighborhoods are popping out
• 32:51at having the highest concentrations.
• 32:53So this is still a work in progress,
• 32:56but this is led by Doctor Tess Carter,
• 32:58who just recently finished her PhD at MIT.
• 33:02And I just want to point your
• 33:03attention to the top few rows here,
• 33:05which show all census tracts,
• 33:07urban tracts and rural tracks across the US.
• 33:11On the left hand side here is that
• 33:13comparing the most non Hispanic
• 33:16white populations to the least
• 33:18non Hispanic white populations and
• 33:20then on the right hand side is most
• 33:23Hispanic versus least Hispanic.
• 33:25And we see for each of these
• 33:27three datasets the CMAC Fusion,
• 33:29the vans angular .01 is that
• 33:32spatial resolution and then a mini,
• 33:35all three of these data sets are
• 33:37very consistent in what they show
• 33:40for at those geographies.
• 33:41And it's very similar to for each region.
• 33:44The absolute magnitude of the values
• 33:46of the PM 2.5 concentrations differ,
• 33:50but the disparities,
• 33:51the patterns and disparity are similar.
• 33:54This is on a regional average basis.
• 33:56So what this tells us,
• 33:58I think I'm still processing this,
• 34:00is that on a regional average basis,
• 34:03this EMAC data set not so bad for
• 34:05estimating those disparities.
• 34:07And you can imagine why that might be.
• 34:09For PM 2.5,
• 34:10we have two things happening simultaneously.
• 34:12We have.
• 34:14We have regional PM 2.5 concentrations.
• 34:18PM is sort of higher in California
• 34:20and the southwest US than it is in
• 34:22other parts of the US and we have
• 34:24that happening at the same time as
• 34:26regional sorting of populations.
• 34:28There's a very large Hispanic population,
• 34:30for example,
• 34:31in California and the southwest
• 34:33breathing those high PPM concentrations
• 34:35in that same region.
• 34:36So that's sort of regional nature of
• 34:39both population sorting as well as pollution.
• 34:41That's one effect.
• 34:42The second effect is what's
• 34:44happening in urban areas.
• 34:46PM 2.5 has some intra
• 34:49urban spatial variability,
• 34:51or so the literature tells us.
• 34:53And that,
• 34:54you know,
• 34:55driven by anthropogenic sources
• 34:57within cities could be contributing
• 34:59to differences in neighborhood scale
• 35:02pollution levels within cities.
• 35:05So this maybe is actually not that
• 35:07surprising that this lines up
• 35:09pretty well regardless of the data
• 35:11set because the spatial resolution
• 35:12of data set doesn't matter that
• 35:14much for that regional effect,
• 35:16that first effect I was describing.
• 35:18But for the intra urban effects,
• 35:21the 12 core meter data set is not
• 35:23going to be able to capture those
• 35:26that intra urban variability.
• 35:28So what do we see within cities?
• 35:30We see something different so.
• 35:32In the top 10 most populated
• 35:35cities across the US.
• 35:37One thing is consistent and that
• 35:39the non Hispanic white population
• 35:41has the lowest PM 2.5 concentration
• 35:43in all three of these datasets.
• 35:46So we see a lot of the
• 35:48dark blue color left of 1.
• 35:49One is the average the the the mean PM 2.5
• 35:55concentration for the entire population.
• 35:58The non Hispanic white population has
• 36:00lower than average concentrations for
• 36:03every one of these major cities in
• 36:05all of the datasets but the ordering.
• 36:07Of the other population subgroups
• 36:09really varies quite a bit depending on
• 36:12the data set and that again is driven
• 36:14by the spatial distribution of the
• 36:17concentrations in the input datasets.
• 36:18I want to point out a couple other things.
• 36:21The CMAC Fusion data set,
• 36:22that 12 kilometer data set that's
• 36:25being used by the Justice 40 initiative
• 36:27team right now that has the least
• 36:30variability between population subgroups.
• 36:32And again not surprising,
• 36:34this is 12 kilometer datasets not
• 36:36capturing that heterogeneity.
• 36:37But we definitely see that play out
• 36:39or we have the the narrowest range
• 36:42here for I just picked up Philadelphia
• 36:46for Chicago and for New York.
• 36:48But then there's really interesting
• 36:50things that happen.
• 36:50So New York, Chicago,
• 36:52and Phoenix all show pretty
• 36:53different effects here,
• 36:54where in New York we have the same
• 36:59ranking of population subgroups in
• 37:01terms of their PM 2.5 concentration for
• 37:03both of the two high resolution datasets,
• 37:06but not in the CMAC Fusion data set.
• 37:09In Chicago,
• 37:10we hardly get much variation at
• 37:12all in any of the three datasets.
• 37:14And then in Phoenix,
• 37:15all three of the data data sets,
• 37:17including CMAC, the CMAC Fusion data set,
• 37:20do have similar disparities
• 37:23across these population subgroups.
• 37:25So we're still trying to dig into
• 37:26each of these cities and understand
• 37:28why they're showing these different
• 37:30different patterns.
• 37:31I'm really excited about the future because.
• 37:34The satellite data we have available
• 37:35right now,
• 37:36these this polar orbiting satellite data,
• 37:37that's a major improvement over what we had,
• 37:40what we had before,
• 37:41which is no satellite data,
• 37:43but we are now launching geostationary
• 37:46satellites which are going to hover
• 37:48over the US as the earth spins.
• 37:49It'll always be taking measurements
• 37:51over the US so tempo is launching
• 37:54in April and that will be a
• 37:56geostationary satellite that's
• 37:58measuring atmospheric composition.
• 38:00Really excited about that.
• 38:02And then Noah is working.
• 38:04On Geo EXO,
• 38:06which is an operational satellite that is
• 38:09intended to launch in the early twenty 30s.
• 38:11And there's so many stages of
• 38:14explaining why this is important.
• 38:16So they asked us to help them explain
• 38:18why this is important for air quality
• 38:20management and for public health.
• 38:21So we've been really happy to be
• 38:24working with them and showing them
• 38:27the value of satellite data for for
• 38:30managing air quality and for public health.
• 38:33And this is work led by Doctor Kate Odell,
• 38:36who is quantifying the number of
• 38:37four air quality alert days across
• 38:39the US that you would get if you
• 38:42had a geostationary.
• 38:43Satellite which is taking measurements
• 38:44across all hours of the daylight versus
• 38:47if you only had that one snapshot
• 38:49from a polar orbiting satellite
• 38:51at 1:30 PM and she's showing that
• 38:53the number of air quality alert days
• 38:55is much much higher for the Geo case,
• 38:58that's the Geo stationary
• 38:59case versus the Leo case.
• 39:01Leo stands for low Earth orbit,
• 39:02which is the polar orbiting satellites.
• 39:05And we wanted to look at the disparities
• 39:08in the populations that are receiving
• 39:11receiving these air quality alerts
• 39:13if we had the geostationary data
• 39:16versus the polar orbiting data.
• 39:18And she finds that actually you know
• 39:21while the magnitude differs overall,
• 39:23the pattern of who, what you know the,
• 39:27the population sub categories
• 39:29experiencing these poor air quality
• 39:31alert days is actually pretty
• 39:33similar depending regardless of the.
• 39:36Geostationary or the polar
• 39:38orbiting satellite?
• 39:39Really quickly,
• 39:40I want to go back to the framing
• 39:42of climate change, because again,
• 39:44air pollution and climate change
• 39:45come from the same sources.
• 39:48Anytime we burn fossil fuels
• 39:49and we burn biofuels,
• 39:51or releasing both air
• 39:52pollutants and greenhouse gases,
• 39:54we want to solve a lot of the
• 39:55problems that I just talked about.
• 39:57We could be burning less fuel and also
• 39:59be gaining by reducing CO2 emissions.
• 40:02So I have been able to partner
• 40:03for the last few years with C-40
• 40:06cities as well as a variety of
• 40:08other partners who had been.
• 40:09Planning,
• 40:10the largest worldwide effort for cities to
• 40:13undertake urban climate action planning.
• 40:15And these are cities that have
• 40:18committed to very deep decarbonization
• 40:20and creating ambitious plans
• 40:22for reducing greenhouse gases.
• 40:24And we help them understand not
• 40:26just their greenhouse gas reduction,
• 40:28which they're already very good at,
• 40:30but now understand also the reduction
• 40:33of PM 2.5 that they would get from
• 40:36taking those ambitious actions
• 40:38to reduce greenhouse gases.
• 40:40This is the framework that we
• 40:41did this within,
• 40:42and we implemented this in six
• 40:44pilot cities around the world.
• 40:45And I just want to show two of the
• 40:47examples of these are actually
• 40:49graphs that are now in these cities
• 40:51climate action plans for the first
• 40:53time integrating air quality into
• 40:55their climate action planning.
• 40:57So Buenos Aires saw their PM 2.5
• 41:01concentrations go down from about
• 41:0312 micrograms per meter cubed
• 41:05in 2050 to around 8,
• 41:08which was under the World
• 41:09Health Organization.
• 41:10Headline at the time we did this analysis,
• 41:12but it's now over the W 1 because
• 41:14that would has been reduced.
• 41:16And then Johannesburg took a bit
• 41:18of a different approach here where
• 41:20they looked at each type of action
• 41:22they could implement and they they
• 41:24looked at the percent of total
• 41:27PPM concentration reduction from
• 41:28that action versus the percent
• 41:30of total CO2 emission reductions.
• 41:32And the one that achieved the
• 41:35greatest dual benefit was a mode
• 41:38shift from on road vehicles.
• 41:40We're now helping them understand
• 41:41CO2 emissions a little bit more.
• 41:43So right now,
• 41:45each city is developing its own
• 41:47urban inventory of CO2 emissions,
• 41:51and that has advantages,
• 41:54strengths and weaknesses.
• 41:56The scientific community is very
• 41:58hard at work developing gridded
• 42:00CO2 emission data sets as well
• 42:03based on satellite observations
• 42:05of light at night and
• 42:06other data sources.
• 42:07And so we're looking at whether or
• 42:09not the self reported inventories
• 42:11from the cities match what we think
• 42:13might be happening in the scientific
• 42:15community using these gridded datasets.
• 42:17And this is work led by Doctor Doyon
• 42:20on where we he's comparing the GPC
• 42:23inventory that's the self reported
• 42:25inventory versus a very widely used.
• 42:27Um, globally gridded emissions inventory
• 42:30called Edgar and he sees that the there,
• 42:33sorry, in this other one is,
• 42:34is ODC, as well as the different
• 42:37gridded CO2 emissions data set.
• 42:39They want it pretty well.
• 42:41This is actually better than I might
• 42:43have expected prior to this project,
• 42:44but he sees a lot more scatter outside
• 42:47of the US and Europe and a lot more
• 42:50consistency in US and European cities.
• 42:53So just to conclude that climate
• 42:55change is worsening air pollution,
• 42:58which is already a leading factor
• 43:01for global health around the world.
• 43:03We have now access to data that we
• 43:06that's completely unprecedented,
• 43:07these novel geospatial datasets they're
• 43:11increasingly capable of providing.
• 43:13Information about pollution
• 43:14levels everywhere in the world
• 43:16with full geospatial coverage,
• 43:18high temporal frequency and in some cases
• 43:21now building long temporal trends too.
• 43:24Some of these,
• 43:25some of these satellites have been
• 43:26flying for years and that's enabled
• 43:28us to do a lot of different things.
• 43:30I just talked today about air pollution
• 43:32levels globally and at 13,000 cities,
• 43:34as well as intra urban disparities.
• 43:36But people are using these satellite
• 43:38data sets and all kinds of unique and
• 43:40very useful ways like spotting wildfire,
• 43:42smoke and dust storms.
• 43:44Thanks.
• 43:45And you know,
• 43:46I really think that this
• 43:47improved information,
• 43:48if we integrate this into our
• 43:51environmental management techniques,
• 43:52including policy development,
• 43:54we can achieve multiple societal
• 43:59improvements simultaneously.
• 44:00I've been really,
• 44:01really fortunate to be in a position
• 44:04now where I can be training the next
• 44:06generation to be using data sets like this,
• 44:09and there's new ways of doing environmental
• 44:12health that are now possible.
• 44:14So bringing that in,
• 44:16bringing in systems approaches
• 44:18and an equity and justice lens in
• 44:22addition to engaging multidisciplinary
• 44:24teams and diverse partners,
• 44:26I talked about some of the partners
• 44:28I've worked with including C-40
• 44:30cities and the DC government.
• 44:31That's just sort of scratching the
• 44:34surface that if you work directly
• 44:37with these action oriented partners
• 44:39from the beginning of a project,
• 44:41you can actually design a project
• 44:43to achieve the needs that they have.
• 44:45To improve life for people.
• 44:47And, you know,
• 44:48leveraging novel geospatial datasets
• 44:49is not something that I was,
• 44:51you know,
• 44:52well,
• 44:52actually I was trained to use novel
• 44:54geospatial datasets that were novel
• 44:55at the time that I did my training,
• 44:57which is before satellites.
• 44:59But you know,
• 45:00a lot of people in the field
• 45:02didn't have that,
• 45:03don't yet have that training and
• 45:04something that we can bring into
• 45:06public health more frequently.
• 45:08There's a lot of communities of
• 45:10practice to plug into as well.
• 45:12We've developed the climate and
• 45:14Health Institute at GW.
• 45:15We just are now completing a NASA
• 45:17supported team called satellite data
• 45:19for environmental justice that brought
• 45:21together a lot of people that were
• 45:23using satellite data for this purpose
• 45:25and a plug to shameless plug to get
• 45:28involved in the AGU health community,
• 45:32which you know includes a lot of
• 45:34people who are using these big
• 45:36geospatial datasets to answer
• 45:38environmental health problems.
• 45:40Very excited that Doctor Chen is
• 45:42part of that community as well.
• 45:43So that's it for me.
• 45:45Just wanted to acknowledge a lot
• 45:47of support and again reiterate
• 45:49that without open data sets none
• 45:51of this would have been possible.
• 45:53So thank you to the data set developers.
• 45:55Thank you.
• 46:01I think it takes around
• 46:035 to 10 minutes for Q&A,
• 46:05so if you do have a question please.
• 46:08Sure. Thank you so much for stopping.
• 46:12It's really interesting.
• 46:12I, I know that one of the major
• 46:15concerns amongst environmental justice
• 46:17communities with datasets such as EJ
• 46:20screen is that they're not specific
• 46:22enough that they don't get down
• 46:24to that really granular level of.
• 46:27Look, fenceline impacts umm.
• 46:29And I'm curious how,
• 46:32when working with large datasets
• 46:35from satellites such as troponin,
• 46:38which only takes about once
• 46:39a day measurement,
• 46:40you can also bring in those qualitative
• 46:43data points from environmental justice
• 46:46communities on the ground to our
• 46:49experiencing air pollution impacts.
• 46:51I love that question because it really
• 46:53shows the value not just in this kind
• 46:56of quantitative data work that I do,
• 46:58but in the lived experience as well.
• 47:00And we've we've run into this
• 47:02multiple projects and I just couldn't
• 47:04agree more with that because.
• 47:06As I showed the we still have disagreement
• 47:08between several of the high resolution
• 47:10datasets that we're looking at.
• 47:12I mean they are better I think than
• 47:14the course resolution data set.
• 47:16But if you're,
• 47:16let's say you're looking at a map of
• 47:19Houston and you've got our land use
• 47:21regression data set of N2 and then
• 47:23the tricomi data set of two and you
• 47:25live in an area which is high in
• 47:27one data set and not in the other,
• 47:29what then?
• 47:29And you know where that is the
• 47:31reality we are in right now,
• 47:33we're in this messy space of data sets.
• 47:36Not matching at that granular scale
• 47:38and I just think it shows the
• 47:41limitation of what we can do with a,
• 47:44you know,
• 47:45one-size-fits-all approach
• 47:45you consistent across the US.
• 47:47We need to bring in people's lived
• 47:50experience and understanding of the
• 47:53local sources affecting their community
• 47:55for this datasets to be improved.
• 47:58How we do that, I think let's like.
• 48:01Get creative, right?
• 48:02I mean, we could bring in story
• 48:04maps of people's life experiences,
• 48:06you know, there's a lot of ways it's,
• 48:08it's not even just about,
• 48:09you know, community monitoring,
• 48:11which can be quite helpful.
• 48:13And, you know,
• 48:14we're rapidly expanding that
• 48:15in the US right now,
• 48:16but, you know.
• 48:17You have people going out and
• 48:20writing about their experiences,
• 48:21taking videos of their experience.
• 48:23So I think that's sort of community
• 48:25contributed,
• 48:26qualitative approach has a lot of value.
• 48:32Yes, just make sure there's any
• 48:34students have a question for us.
• 48:36Is there a hand over there?
• 48:38Umm, so my question or this, first of all,
• 48:41thank you for the fabulous presentation.
• 48:42I greatly enjoyed it.
• 48:44My question, slash comment is
• 48:46about environmental disparity.
• 48:47So I you know, a lot of times we see
• 48:49more and more and more beautiful,
• 48:51beautiful, more and more detailed maps.
• 48:53However, if we could press a button today
• 48:56that made exposure equal across the world,
• 48:59first of all, we press it.
• 49:00Second of all,
• 49:01environmental disparities would
• 49:02still exist because people
• 49:04respond differently to health.
• 49:06So my comment to you my question.
• 49:08Is. What are your thoughts on this?
• 49:10Because I have had.
• 49:12Like when I talk with communities,
• 49:1599 to 100% of them talk about
• 49:17exposure without talking about the
• 49:19fact that and it is a fact that we
• 49:22know that people respond differently.
• 49:24And to what degree do you think
• 49:27environmental health disparities should be?
• 49:29Are there may be some environmental
• 49:32disparities not incorporated into the
• 49:34world's most perfect exposure map?
• 49:37The way they agreed to you that we
• 49:38have very focused on pollution levels,
• 49:40and the same pollution level
• 49:42can cause dramatic different,
• 49:43dramatically different impacts
• 49:45for different populations.
• 49:46I showed that map of Washington,
• 49:48DC and the high PM 2.5 mortality rate
• 49:50on the eastern half and the low PM 2.5
• 49:53mortality rate on the western half.
• 49:55That actually comes from a pretty consistent
• 49:59PM 2.5 concentration for the entire city,
• 50:02but vastly different mortality rates.
• 50:06The, you know,
• 50:07Southeast Quadrant has had no hospital.
• 50:09GW Building went out.
• 50:10I'm very happy that that that's happening.
• 50:12But no hospital,
• 50:13so no access to healthcare,
• 50:14no easy access to healthcare.
• 50:17This is the same,
• 50:17you know,
• 50:18in cities all around the country
• 50:20and around the world that there's,
• 50:22you know,
• 50:22social determinants of health are a major,
• 50:24major.
• 50:25Doctor Diamond exposure and I
• 50:27think in terms of addressing it,
• 50:30I mean we have like I said,
• 50:32the there's this time and economic
• 50:34justice screening tools being used
• 50:35now for the Justice 40 initiative.
• 50:37We have EJ screen to show where
• 50:40these disadvantaged communities
• 50:42are in a nationwide basis.
• 50:43Some are not accounting for those
• 50:47that increase susceptibility,
• 50:48increase mortality rates,
• 50:50higher mortality rates,
• 50:51higher health outcome rates.
• 50:52The Cbest tool right now is.
• 50:55Includes poverty and one
• 50:57additional indicator.
• 50:58So that could be PM 2.5 and EJ.
• 51:02Screen has an index.
• 51:03I think if we were to use more of like that
• 51:06index approach that brings in poverty,
• 51:08brings in health and some of these
• 51:11other social determinants of health in
• 51:13addition to the pollution exposure,
• 51:15we can start to identify not just
• 51:17who is experiencing bad pollution,
• 51:19but who is most impacted by that bad.
• 51:23Thank you.
• 51:25Any idea what's causing the differences
• 51:28in disparities between cities?
• 51:30I'm originally from Chicago.
• 51:31The expressways run through
• 51:32black and brown neighborhoods,
• 51:34which is true everywhere.
• 51:36But disparities there,
• 51:38both knocks and five were fairly
• 51:40modest compared to the other cities.
• 51:44That's it's such a great question.
• 51:46And we now have a big project with
• 51:48an environmental Defense fund to
• 51:49dig into Chicago specifically to
• 51:51understand that because Chicago
• 51:52does have a whole lot of trucking
• 51:54that is coming through the city.
• 51:56And as you say it is associated
• 52:00geographically with with with
• 52:02black and Hispanic populations.
• 52:03There is no some other major roads
• 52:05that are more in wealthier whiter
• 52:08neighborhoods like Lakeshore
• 52:09Drive going going north.
• 52:11So when you take like an urban average.
• 52:14It also very much depends on learning.
• 52:16It very much depends on how
• 52:17you define what the city is.
• 52:19Are you looking just at Chicago,
• 52:20the entire county, entire MSA?
• 52:23And actually we've seen that the
• 52:26disparities flipped depending
• 52:27on how you define their opinion.
• 52:30More details coming soon at Chicago,
• 52:32so that's an interesting one.
• 52:35Thank you so much for the fascinating part.
• 52:37I just have a question about the time trends.
• 52:40You showed that by racial,
• 52:43ethnic of their exposure and O 2:00 PM,
• 52:48but how, how does that change over time?
• 52:52Is there any like convergence
• 52:53across those groups?
• 52:56They have to share share the slides
• 52:58but the the project that I showed
• 53:00that had the PM on the left hand side
• 53:02and the No2 on the right hand side
• 53:03that showed PM mortality rates and
• 53:05then No2 attributable asthma rates.
• 53:07Those do show trends overtime
• 53:09and the concentrations for both
• 53:11PM and NS are going down for all
• 53:14population subgroups really great.
• 53:16But the relative disparities
• 53:18are increasing for both parents
• 53:20because of the like the the changes
• 53:22in that that overall magnitude.
• 53:24So the. That's this one.
• 53:26Thank you.
• 53:27So the overtime,
• 53:29the PM concentrations have come
• 53:31down approximately the same amount
• 53:33for all population subgroups
• 53:35and that leads to an increased.
• 53:38Ratio between the population subgroups and
• 53:39then for N2 this doesn't really look like it,
• 53:42but these orange dots are going down as well.
• 53:46Much greater energy reductions
• 53:47for the least white communities,
• 53:50but still we see rising ratios of
• 53:54disparity relative disparities.
• 53:57Thank
• 53:57you. The the reason I ask this is
• 54:00from the population migration and
• 54:03point of view is very mixed picture.
• 54:05The data shows that it's more
• 54:08segregation across cities unless so
• 54:11within cities in many parts of America.
• 54:14So that's interesting.
• 54:17We only looked at the the temporal
• 54:18trends and the pollution levels,
• 54:19not where where people are living,
• 54:21so that would be an interesting
• 54:23question to look into.
• 54:25Uh, we we do have a comment online,
• 54:27but I think it's more like
• 54:29suggestion we can look at and
• 54:31thank you all for coming because
• 54:32we have a class right office.
• 54:34So we have to end today.
• 54:35Thank you all and thanks.