Reductions in air pollution due to COVID-19 in China have probably saved 20x the number of lives than have so far been lost to the virus. Does not mean pandemics are good, but rather that our economies absent pandemics are bad for health (Thread 1/n)
Yesterday was the 3rd worst wildfire pollution event in US history, we calculate, just behind 2 days in 2020 that hit the west coast. Don't have today's data yet, but it could be the worst. Pollution levels are lower in this ongoing event but many more people are exposed.
June 7th was the worst wildfire day on record in the US since 2006, by far. June 6th was the 4th worst. Just a massive, awful event, with highly populated areas getting hit with unprecedented levels of pollution. Great data work by
@StanfordECHOLab
@minghao_qiu
Jessica Li.
Not sure folks have internalized what a ~$200 social cost of carbon (per new EPA numbers) means. We generate ~40bn tons CO2/yr, so that is $7.6T in damages! 8% of global GDP! OR: 1 gallon gas == ~$1.50 in external costs. These numbers are huge.
Like other academics I catch up on journal referee reports on planes. But: CO2 levels on plane I was just on >1600 ppm, which experimental studies have shown can reduce cognitive function by 50%! Maybe better to just watch movies, which is what I want to do anyway…
Real banger of a new paper on global temp/gdp relationships by Adrian Bilal and Diego Kanzig. SCC of >$1000 under pretty conservative interpretation of their results! Link here:
Does this mean pandemics are good for health? No. Instead it means that the way our economies operate absent pandemics has massive hidden health costs, and it takes a pandemic to help see that. (/n)
Fascinating new paper exploring how historical air pollution shaped the structure of cities. Wind blows W-> E in northern temperate latitudes. In cities with coal, E side of cities more exposed, thus poorer. Pattern persists even after coal gone
Quick 🧵 on historic wildfire smoke event ongoing throughout eastern Canada and much of eastern US. Monitors showing very high levels of PM for thousands of miles. For reference, background levels are ~10ug, so these levels are way above normal.
Before twitter implodes, please check out our new paper on using satellites+ML to do large-scale public policy evaluation. Application is to impacts of electrification expansion in Uganda. Great team effort led by
@NWRAT
+
@atlasai_co
. quick thread below.
A few weeks ago, NASA published striking photos showing the reduction in key air pollutants over China as a result of the economic disruption caused by COVID-19 in the country: 2. .
LATE BREAKING: announcing 2 new cluster searches in Stanford's new school of climate & sustainability. One in sustainable development + environmental justice, aimed at senior social scientists. Other is open rank search in climate science. PLEASE RT WIDELY!! links and info 👇
Current wildfire smoke event in NYC is off the charts relative to anything in past two decades. Here's a chart showing daily PM2.5 from smoke back to 2006 versus today, averaged over pollution monitors in NYC. Last two days are remarkable.
Check out our new paper on the impacts of wildfire smoke exposure on test scores, expertly led by
@jeffliwen
. Smoke exposure during school year lowers test scores at end of year, w/ monetized costs in billions annually.
Are you an ML/AI whiz interested in working on problems that matter? Check out SustainBench, our new set of curated datasets + benchmarks for sustainability-related ML tasks. Great opportunity to improve measurement of poverty, ag, education, water, more
We've got a new paper out using public satellite imagery and machine learning to measure economic well-being over space and time at continental scale in Africa
For the average American, we
@StanfordECHOLab
calculate that cumulative smoke exposure (PM2.5 exposure on each day, summed across days) through mid-2023 is already way worse than total cumul exposure in every year since 2006. And main fire season in West is just getting started.
How well do markets price information about climate risk? We study this question in the context of flood risk and the US real estate mkt, in a new paper led by the awesome Miyuki Hino (not on twitter). Paper plus short thread
We have a new paper + dataset out today measuring daily wildfire smoke PM2.5 across US back to 2006, expertly led by
@marissalchilds
@StanfordECHOLab
. Below is an animation of 2020 data.
paper:
data:
Quick thread below
Come work with me as an RA! Looking for a couple folks with good 'metrics + spatial skills. RAs are treated like collaborators, often end up as co-authors on top pubs, go on to good PhD programs. Please apply!
Can we use satellites to measure and understand human livelihoods? Lots of folks now trying to answer this question, so
@DavidBLobell
@StefanoErmon
Anne Driscoll and I wrote a review paper, out today in Science. Answer is yes! thread 👇
The tyranny of national averages:
Blog post with our excellent summer intern
@ApoorvaRed
. We tend to measure poverty at the national level but this hides huge subnational variation. New methods and data are starting to allow us to do better.
I'm hiring 2 full time RAs/pre-docs for a summer 2023 start date. You will work in a collaborative lab environment, and nearly all past RAs have had their names on multiple pubs. Coding experience in R/python desired. Plz retweet!
#EconTwitter
We know that dirty air is really bad for health. Like really really bad. Could it be that this reduction in air pollution actually resulted in more lives saved than had so far been lost to the virus itself? (3/n)
Some thoughts on the monumental gift Stanford just received from Ann+
@johndoerr
to kickstart our new sustainability school. I will be employed at that school and am helping with hiring, organizing the social sciences etc so am biased, nevertheless... 1/n
Q: How much has climate change increased economic damages from flooding? A: a lot! Check out our new paper led by awesome
@fvdav
and Noah Diffenbaugh. Short thread below.
@StanfordEarth
How unusual is the recent wildfire smoke being experienced in US Rockies and midwest? Very. Here is a plot of daily smoke PM2.5 from monitors in Denver CO, for all days 2006-2022, vs the last 4 days. Smoke PM in last few days is a Denver record, and way earlier than normal
Forthcoming in the AER: "The Social Costs of Keystone Species Collapse: Evidence From The Decline of Vultures in India" by Eyal Frank and Anant Sudarshan.
Wow, very honored to receive Best Paper Award from
@AEAjournals
AEJ-Policy, for paper with Kyle Emerick on climate adaptation. Given the amount of cool stuff AEJPolicy has published in last 3 yrs, this is a real honor. Very psyched!
What are folks' favorite papers that evaluate a climate adaptation intervention? i.e. that tell you something to the effect of: intervention Z reduced the effect of some climate variable on some outcome by some amount. Self-citation encouraged!
Sol Hsiang and I are organizing a conference on climate adaptation + loss & damage, Apr 2023 at Stanford. Please submit your half- or fully-baked papers, deadline Feb 1! Some travel support available. Call for papers below, plz email/DM with questions.
Second paper on wildfire smoke from our
@stanfordecholab
group out this week. This one quantifies the contribution of wildfire to US PM2.5 trends. We find much broader wildfire influence than previously estimated, on both avgs and extremes. Quick thread.
Super proud to announce that our company,
@atlasai_co
, has launched our first product: continent-wide measurements of economic livelihoods across Africa. Read about it here and sign up for a free account at
So unbelievably shameful for the head of the most important development institution in the world, the
@WorldBank
, to be a climate denier - I had no idea. He's coming to
@Stanford
next week!
🌎🔥 Today I asked World Bank president
@DavidMalpassWBG
if he believed in the scientific consensus that the man made burning of fossil fuels is rapidly and dangerously warming the planet.
“I’m not a scientist,” he said.
Here’s a thread about the remarkable exchange. 🧵 1/
Announcing v2 of our climate adaptation conference at Stanford, this Oct 10-11. Please submit! Looking for empirical work on adaptation & will consider earlier-stage working papers as well as more polished work. Submission deadline July 15. Plz share, RT
Interested in wildfires, air pollution, or climate change? Check out our new PNAS paper on the causes and consequences of the growing wildfire burden in the US. Great author team: Anne Driscoll
@samheftneal
@jaburney
@MichaelWWara
Jenny Xue (1/n)
A new study coauthored w/
@carnegiescience
&
@DavidBLobell
shows that plant photosynthesis inversely mirrors air pollution patterns; if pollution was curtailed, increased photosynthetic activity could remove 40-60 megatons of CO2 from the atmosphere. 🌱
Interesting new pre-print on temperature and manufacturing productivity in US by
@jacopont
Yu Zeume. Productivity for all firms falls on hot days, effect maybe bigger for smaller firms. A hot week reduces annual productivity by 0.35%.
Using these numbers, I calculate 50k-75k reduction in premature mortality due to these air quality improvements. This is roughly 20x the current estimated number of direct deaths from COVID-19 in China. (9/n)
Heat waves are lethal, full stop. Even in CA. Here's the relationship between daily heat and mortality using recent CA data, from research with
@gould_cf
. Temps above 90-95F raise mortality meaningfully. Days like today expected to cause ~100 excess deaths/day across CA.
Currently, ~140 million Americans are under heat warnings & advisories as a long duration, intense heat wave scorches the West and sultry conditions engulf the South. Details on the "potentially lethal" California heat wave especially:
New working paper on climate, wildfire smoke, and mortality, expertly led by
@minghao_qiu
@StanfordECHOLab
. We find that by mid-century in the US, damages from mortality from wildfire smoke are about equal to the sum of all other climate damages in recent estimates 🧵
Mortality from increased wildfire smoke due to a warming climate could represent one of the largest costs of climate change in the US, from
@minghao_qiu
, Li, Gould, Jing, Kelp, Childs, Kiang, Heft-Neal, Diffenbaugh, and
@MarshallBBurke
New paper out today on impact of global biomass fires on infant health, expertly lead by
@HPullabhotla
@StanfordECHOLab
. We use satellite estimates of burning + variation in wind, and attribute large proportion of total infant deaths to biomass burning
Are these numbers right? Impossible to know now, but they will actually be testable in a few years when more comprehensive mortality statistics are available. But I think they are perhaps not crazy, and could be conservative. Air pollution is really bad. (10/n)
This excellent article by
@dwallacewells
will be read as a provocation, which it is not. The math, as he says, is simple. The rich owe the poor for the damages we have caused, and are causing.
What a triumph of empirical social science: a comprehensive comparison of the cost-effectiveness of 133 policy changes in the US . Finds immense returns to investing in children, relative to other interventions. What an amazing and useful paper.
To answer this question, I first pulled some air quality monitoring data from US government monitors in 4 Chinese cities, and compared PM2.5 concentrations in Jan-Feb 2020 relative to the previous 4 years. As you can see, in 3/4 cities, air pollution was way down on average:
Seconding this strongly. We just published work showing substantial increases in preterm birth risk for even short periods of moderate exposure. Please be careful and try to limit exposure however you can.
So this is my wildfire PSA for the day:
Pregnant women of Northern California, stay inside.
If you can, stay inside an environment with HEPA air filtration (MERV13).
If you must go outside, wear a properly fitted KN-95 mask. But better yet, just don't.
Another exciting Monday morning for COVID-19 vaccines! Moderna is reporting 94.5% efficacy for their mRNA vaccine.
Read on for a biostatistician's breakdown of the interim results. 10 tweets on severe disease, subgroup analysis, and more.
Press release:
Quick thread on a paper we had out last week on wildfire smoke in
@NatureHumBehav
. Take-home is that current policy approaches to protecting folks from wildfire smoke exposure are pretty inadequate. Paper here (1/n)
Sometimes-heard claim: increases in wildfire burned area under climate chg are self-limiting, because eventually we run out of stuff to burn.
Science: not really. Stuff grows back, reburns. See excellent new paper by
@climate_guy
@pyrogeog
and others
"Heat shocks make painters paint more brightly" was not on my climate impacts bingo card. Cool paper by
@ShuheiKitamura
. No wonder Dutch renaissance painting was so dark - it was freezing in the Netherlands in the 1600s...
Extra finally, reminder that wildfire smoke is not regulated under the Clean Air Act, our main (and historically very successful) tool for improving air quality. This current event from transboundary smoke, making it even harder. We are not prepared in any way for these events.
How should "effective altruists" - i.e. those trying to get the biggest bang for their donation buck - think about climate change? Some thoughts and crappy math. 🧵
Email: Dear Dr Burke, we'd like to invite you to present your research at...
Gmail spam filter: NOPE, LOOKS SUSPECT
Email: I am nite lovr i neeed somone cuddl will u send pix
Gmail: YES CERTAINLY LEGIT
Hopefully final update on the severity of this wildfire smoke event. We calculate that the last two days were the worst individual wildfire smoke days on record, in terms of the number of people exposed to the most smoke PM2.5. Quick 🧵 on importance, drivers, aftermath (1/n)
For change in mortality rate, I use excellent paper by He et al 2016, who used quasi-experimental variation in air pollution induced by regulations put in place around the Beijing Olympics. They find huge effects for young and old people (7/n)
New working paper out with a big team. We show how data from private pollution sensors, cell phones, social media, and search activity give new insights on exposures and behavioral responses during large wildfire smoke events. Quick thread.
“Loss and damage” is a central focus of recent global climate policy discussions, but somehow it still lacks a clear definition & framework for estimation. We try to provide both. Approach enables estimation of damages due to past or future emissions from any emitter.
A new framework for estimating past and future damage from climate change, and for linking this damage to specific emitters, from
@MarshallBBurke
, Mustafa Zahid, Noah Diffenbaugh, and Solomon M. Hsiang
First, how anomalous is this event? Near worst or worst event in last two-ish decades, I think. Using estimates of smoke PM2.5 we generated
@StanfordECHOLab
, here's the time series of PM2.5 from wildfires in Detroit, for all days back to 2006. Today is a historic anomaly.
Wowza- what an amazing climate adaptation paper. Shows that adaptation can be constrained by incorrect beliefs and that there are cost effective solutions. Like 8 papers in one.
My
@StanfordECHOLab
group is hiring a Research Data Analysis (aka pre-doc) to work with us full time starting next summer. Our RAs are co-authors and treated as colleagues. Strong programming background in R or python is required. Plz apply or RT!
New
@PNASNews
paper on quantifying fire-specific smoke impacts, expertly led by
@jeffliwen
. We link each big wildfire (2006-2020) in the US to its impact on PM2.5 and health. Complements typical fire impact measures (burned area, structures lost, etc). 🧵
Some exciting news to announce for AtlasAI (), a company I helped found that's using satellites and computers to better understand and improve livelihoods around the world.
Nice World Bank post on the importance of non-traditional data sources for adequately measuring economic progress around the world , including some of the work done by our great students
Beijing was a different story, and that story is nicely explained here: But averaged across all 4 cities including Beijing, I estimate a 15-17ug/m3 reduction in PM2.5 in Jan-Feb 2020 relative to past years. (5/n)
Articles like this make me proud to be a NYT subscriber. Critically important issue, rigorous original data collection, gripping and beautiful visualizations.
Couple of wildfire papers out from our echo lab group this week. First one out yesterday, expertly led by
@samheftneal
@StanfordECHOLab
, on smoke effects on emergency dept visits in California. Some surprises! Quick thread
ttps://www.pnas.org/doi/10.1073/pnas.2302409120
Finally, as Stanford faculty in the new school, we need to be held accountable. It's dumb that our society has people who can easily give us $1b. But now we have it, and it's up to us to not fuck it up. /n
@MichaelRoss7
I don't think this is a useful framing nor true to int'l or even US experience. Many countries + states have adopted pricing or C&T, and studies show benefits. Re US politics, here's a nice thread from someone who's been at this for a while
I did as much as anyone to analyze cap-and-trade bills when Congress last tried to enact sweeping climate legislation, and
@bobkopp
's tweet sparked quite a discussion, so I thought I’d offer my reflections. (Sorry in advance for the long tweet storm...)
We've got a new paper out in
@nature
on the health burden of air pollution in Africa. Made a little explainer video if slogging through scientific articles is not your thing:
Great new postdoc fellowship(s) announced today at
@stanforddoerr
. Search is broad area (incl social sciences) and is esp interested in folks from underrepresented backgrounds. Please apply!
New WP, expertly led by
@gould_cf
. We show that well-known U-shaped temperature-mortality relationship does not hold for morbidity: ED visits show linear relationship (in CA). This is bad under future climate chg: warming reduces mortality in CA (!) but increases morbidity. 🧵
Climate change could have opposing effects on mortality and morbidity, from Carlos F. Gould, Sam Heft-Neal, Alexandra K. Heaney, Eran Bendavid, Christopher W. Callahan, Mathew Kiang, Joshua S. Graff Zivin, and
@MarshallBBurke
Really nice short piece on climate adaptation by
@AndrewDessler
. Climate change will come as catastrophic extreme events, but also as a slow accumulation of little things that make life worse, with huge aggregate costs
Stanford's Center on Food Security and the Environment
@FoodSecurity_SU
now accepting applications for our postdoc fellowship. Open to candidates across social + natural sciences. Please apply! (or at least RT)
@FSIStanford
@StanfordWoods
I do worry that the combination of engineer lead donor and engineer initial dean could miss a substantial part of the solution space. If covid taught us anything, its that cutting edge tech is not enough. What policies support sustainable transitions and... 6/n
If you're interested in how satellites can improve the measurement and understanding of economic development, check out a few blog posts
@DavidBLobell
and I put together with our partners at
@usaid
@FeedtheFuture
: and
To be conservative, call this a 10ug/m3 reduction for 2 months. What were the health benefits of this very large air quality improvement? We need three numbers: change in mortality rate per change in ug/m3 PM, baseline mortality rate, and total population affected. (6/n)
Really terrible outdoor air around Seattle today, but just as important to look at what's going on inside buildings as well as outside. And that picture is ... also very bad. Here's outdoor & indoor purple air this morning - lots of indoor readings way into the 100s. Terrible!
Highest smoke PM readings of the year in Bay Area this morning. Per usual, only about half of folks seem to be running filters of some sort - outdoor AQ on left, indoor on right. Run those filters, folks!
Big news on the upcoming Stanford Freedom conference! Steve Koonin has dropped and been replaced with an actual climate expert, Noah Diffenbaugh. A win for climate science and, dare I say, academic integrity. Event is livestreamed, bring your popcorn
Recently published in REStud, ``Does Pricing Carbon Mitigate Climate Change? Firm-Level Evidence from the European Union Emissions Trading System,'' from
@JonathanColmer
,
@mondpanther
,
@MirabelleMuuls
, and Wagner:
Conservatively (?) assume that only 50% of Chinese population affected - equivalent to no rural effect and only some urban areas affected. Assume that air quality improvement lasted 2 months and then disappeared. And finally, assume no-one btwn 5yrs-70yrs of age benefitted.
Check out our new paper (preprint aka working paper) using satellites and ML to measure the impact of electrification on economic livelihoods, led by
@NWRAT
with colleagues at
@atlasai_co
. Quick 🧵
What are the health consequences of the terrible air quality experienced recently in CA due to wildfires? In a back-of-envelope, we calculate that smoke exposure has likely caused 1000-3000 excess deaths in CA in the last month. Blog+thread
Job posting
#2
for today: come work with us as a postdoc! Open to a broad range of fields + topics that touch on food security/ poverty/ environmental issues. Opportunities to collab across and outside our center. Please apply!
We're hiring more ML talent at
@atlasai_co
! You could destroy democracy at FB, spend your days selling ads, OR you could come work for an AI company trying to do good in the world.
HOWEVER: NYC is not yet experiencing what West Coast cities have experienced in the last few years. Here are comparable plots for SF and Seattle, who have both hit values this high multiple times recently from wildfire smoke.