12 | New Scientist | 21 March 2020MANY cities in China normally
experience dangerous levels
of PM2.5, the tiny particulate
matter linked to some of the
worst health effects from
air pollution. Now it seems
artificial intelligence could
help us avoid the stuff.
Most air pollution forecasts
are based on models that
use maps of annual average
air pollution emissions, the
application of weather models
and assumptions about
chemical reactions. These
can’t account for unforeseeable
events, such as extra pollution
from a traffic jam or an
accidental chemical release.
Baihua Li at Loughborough
University, UK, and her
colleagues took a different
approach to see whether AI could
do a better job. They used
machine learning to train a
model for predicting PM2.
levels on three years’ worth of
data from Beijing, chosen due
to the number of Chinese cities
with a PM2.5 problem.
Around 26,000 data
points were used, including
average PM2.5 readings fromfour roadside pollution sensors
and historical weather data. Li
says the resulting forecasts for
the locations near these sensors
were highly accurate, based
on how closely they matched
historical observations. One
hour ahead, they were 95 per
cent accurate, falling to 85 per
cent for 6 hours ahead.
Using this approach also
helped the researchers tease
out which factors were the
most important in predicting
dirty air. Sunlight, air pressure,
the season and wind speed
were found to be key. Wind
blowing from the north-westoften indicated a bad pollution
episode was likely, perhaps due
to the industrial facilities to the
north-west of Beijing.
Unlike most air quality
forecasts, which usually warn
the public of a single likely level
of pollution for their area, Li’s
system provides a likely rangeof PM2.5 levels to give an idea
of the uncertainty. The model
could be adapted for other cities
affected by PM2.5, such as Delhi,
provided it is trained with local
data. The developed system is
going to be tested in Shenzhen,
China, using live data from
pollution sensors.
“If we can warn people
of impending smog, then
they can take steps to protect
themselves,” says Gary Fuller
at King’s College London. He
says machine learning could be
useful where data on pollution
sources is lacking. “Alerting
systems around the world,
such as that run by the UK Met
Office, rely on understanding
air pollution sources in a city.
Data-driven models, such as this
example for Beijing, can be used
where the knowledge of local
sources is sparse or uncertain.”
The coronavirus outbreak
means that air pollution in
China is currently down, as
people are staying at home.
Satellite analysis reveals that
levels of PM2.5 in February
2020 were down 20 to 30 per
cent versus previous years.
A separate study by the
European Environment
Agency (EEA) last Thursday
concluded that relatively low-
cost air pollution sensors used
by individuals and groups can,
when used well, provide data
meeting the same quality
standards required of official
roadside monitors. The EEA
said such citizen science
initiatives could “help raise
public awareness of air quality
issues in communities and
trigger behavioural changes
to reduce emissions”. ❚IT IS possible that life started
on Mars billions of years ago in
certain regions where conditions
were unusually hospitable.
If life ever existed on Mars, it is
conceivable that it began with RNA,
which is similar to the DNA that
carries our genes. Many scientists
suspect that RNA played a key role
in life’s beginnings on Earth, andsome even argue that the first
life was based solely on RNA.
To see if that could have played
out on Mars, Angel Mojarro at
the Massachusetts Institute of
Technology and his colleagues
simulated the early environment
of Mars. Its current frigid climate
allows for little liquid water, but
there is evidence that Mars had
running water billions of years ago.
Mojarro’s team placed RNA-like
molecules in water that contained
dissolved metals and other
chemicals. These solutions mimic
water on early Mars, which would
have contained metals leached out
of rocks, and would have been
altered by atmospheric chemistry.
They tracked how long the
molecules lasted before breaking
apart. The RNA-like molecules were
most stable in mildly acidic water
containing magnesium ions, which
would be expected if the water were
resting on basalt: a rock that forms
when lava from a volcano cools
(bioRxiv, doi.org/dprt).
The authors say that means
Martian life developing from RNA
is plausible. Jack Mustard at Brown
University in Rhode Island says the
results should be treated with
caution, because there are “very
large uncertainties” about what
conditions on Mars were really like
when the planet was young. ❚Some Chinese cities
have high levels of
air pollutionSolar system Air pollutionMichael Marshall Adam VaughanDA
VIDGOURHAN/ALAMYNews
PM2.
Tiny particulate matter that
can cause health problems“ Mars is currently too frigid
for much liquid water, but
it may have had running
water billions of years ago”
AI forecasts harmful
particle levels in Beijing
Early Mars may have
had right conditions
for RNA to start life