78 Science & technology The EconomistMarch 28th 2020
2 er airlines are willing to continue to be
bound to something rather different from
what they originally thought they were
signing up to.
Ultimately, the outcome of covid-19 for
the climate will depend on two unknowns.
One is how long the pandemic lasts—
something over which politicians have
some control, but certainly not as much as
they would like. The other is how govern-
ments choose to pull their economies out
of the hole now being created, which is en-
tirely a matter of politics.
The lazy way, the easy way, to boost
countries’ economies in response to the vi-
rus would be for governments to throw
money at established versions of big in-
dustries like energy, transport and con-struction. They could, though, if they
chose to do so, spend the cash instead on
encouraging climate-friendly versions of
these industries: more solar energy (or
even, heaven forfend, nuclear power) in-
stead of bungs to oil and gas; more batteries
for cars, and money for research into hy-
drogen-powered fuel cells; cash prizes for
ways of making steel and cement without
releasing CO 2 ; and so on.
Fatih Birol, head of the International
Energy Agency, an intergovernmental or-
ganisation that operates under the aegis of
theoecd, has already put out a plea for
green stimuli of this sort. Similar calls,
though, were made in 2008 and 2009. They
fell on deaf ears. Perhaps this time around
it will be different. 7S
ticking a thermometerinto an arm-
pit, mouth, ear or other body cavity is
the most accurate way to take someone’s
temperature. Understandably, though, this
cannot be done at airports or checkpoints
set up elsewhere to screen the masses for
feverish victims of covid-19. So, in a bid to
detect the warmth produced by a fever
without touching any bodies, officials have
opted for alternatives.
The hand-held “thermometer guns”
now ubiquitous in China, among other
places, are one option. These instruments,
known technically as spot pyrometers, use
a device called a bolometer to estimate an
object’s temperature. A bolometer’s electri-
cal resistance depends on how hot it is.
That, in turn, depends on the amount of in-
frared radiation falling on it from whatever
it is pointing at.
Spot pyrometers are used widely in in-
dustry to check equipment for signs of
overheating, but the infrared signals they
rely on can be muddled by dust, moisture,
smoke, a change in ambient temperature, a
smudge on the device’s lens or even by ra-
dio signals. Beyond all this, an official
checking a stream of foreheads may, for
reasons of personal safety, be reluctant to
hold the gun close enough to obtain an ac-
curate reading.
An alternative technology, the thermal
camera, is costlier. But it can operate from
farther away. Instead of a single bolometer,
it has an array of them. These form the pix-
els which generate the camera’s image,
thus building up a heat map of whatever
that camera is looking at.One popular thermal camera, the A320,
made by flir Systems, an American firm,
can detect variations in radiation which
correspond to temperature differences
within an image of just 0.02°C. But this
merely shows whether one part of the ob-
ject being examined is warmer or cooler
than another. When measuring the object’s
actual temperature, the A320 is accurate
only to about 2°C, says Giovanni Scaglia,
flir’s head of sales in Italy. For a single
reading, this is enough to mistake normalbody temperature for a raging fever. In
practice, however, the camera’s software
looks for deviations from the average tem-
perature of passers-by. Those noticeably
hotter than their fellows can then be select-
ed for closer investigation.
Identifying individuals who are infect-
ed is important, in order to stop them pass-
ing the virus on. But temperature data can
also be used to track the epidemic as it
spreads. This is the goal of Kinsa Health, a
firm in San Francisco that has sold or do-
nated more than 1m smartphone-connect-
ed thermometers to households in Ameri-
ca. The phones these thermometers are
linked to carry an app that transmits back
to base each body-temperature recording
which its user makes. The app can then
give simple medical advice (for example,
based on age, sex and so on, does a user
with a particular temperature need to see a
doctor or not?) The app also provides epi-
demic information about the neighbour-
hood, including such things as how badly
local schools are affected.
Using its thermometers and apps, Kinsa
has built up a trove of data on past fevers in
America. Besides being good for public
health, this information has commercial
value. One American pharmacy chain pays
Kinsa for fever data in order to avoid partic-
ular stores selling out of things the feverish
may wish to purchase.
In the past, Kinsa’s focus has been on in-
fluenza. But now the fear is that any elevat-
ed body temperature is the result of co-
vid-19. On March 18th the firm began
posting relevant data, duly anonymised,
but more or less in real time, on a website
called Health Weather. In contrast to this
“nowcasting”, a government site called Flu-
View, which is run by the Centres for Dis-
ease Control and Prevention, relies on
sources like hospitals to provide it with
data, and tends to lag an epidemic by about
two weeks.
Once a fever kicks in, people tend to
take their temperature frequently. Kinsa
therefore knows a lot about how an epi-
demic is progressing, and how severe it is,
in lots of different places around America.
In the past, this has enabled it to make ac-
curate forecasts of disease burden down to
the level of individual zip codes. According
to Nirav Shah of Stanford University, who
advises the firm, such forecasts can look up
to 20 weeks into the future for influenza.
Covid-19 is not influenza, of course, and
presumably has different patterns of
spread. But even before these have been
worked out properly, data from Kinsa’s re-
mote thermometers are flagging up useful
warnings. As The Economistwent to press,
atypically high numbers of fevers had
popped up in much of peninsular Florida.
Time, then, for Floridians to get off the
beaches and start keeping their distance
from one another. 7Taking people’s temperatures can help to fight the spread of the coronavirusDetecting elevated body temperaturesAnd now here is the fever forecast
Bolometers to the fore