Page not found | The Economist

The Greenland ice sheet has melted past the point of no return

Even if global warming stopped today, the ice would keep shrinking

Graphic detail
Aug 25th 2020

ANNUAL SNOWFALL can no longer replenish the melted ice that flows into the ocean from Greenland’s glaciers. That is the conclusion of a new analysis of almost 40 years’ satellite data by researchers at Ohio State University. The ice loss, they think, is now so great that it has triggered an irreversible feedback loop: the sheet will keep melting, even if all climate-warming emissions are miraculously curtailed. This is bad news for coastal cities, given that Greenland boasts the largest ice sheet on the planet after Antarctica. Since 2000 its melting ice has contributed about a millimetre a year to rising sea levels. The loss of the entire ice sheet would raise them by more than seven metres, enough to reconfigure the majority of the world’s coastlines.

A new study shows that SARS-CoV-2 can linger in the air for hours and on some materials for days

AT A TIME when many people have taken to washing hands and sanitising the objects they hold dear—frequently—a pesky question has loomed. How long does the SARS-CoV-2 virus stick around? A new paper in the New England Journal of Medicine, one of the first to examine the lifespan of the virus on common surfaces, offers some answers.
Like the common cold, covid-19 spreads through virus-laden droplets of moisture released when an infected person coughs, sneezes or merely exhales. A team of researchers, including scientists from America’s National Institute of Allergy and Infectious Diseases, simulated how an infected individual might spread the virus in the air and on plastic, cardboard, stainless steel and copper. They then measured how long the virus remained infectious in those environments.

They found that SARS-CoV-2 stays more stable on plastic and steel than on cardboard or copper. Traces of the virus were detected on plastic and steel up to three days after contamination. SARS-CoV-2 survived on cardboard for up to one day. On copper, the most hostile surface tested, it lasted just four hours (see chart). In the air, the team found that the virus can stick around for at least three hours. In the air, as elsewhere, the virus’s ability to infect people diminished sharply over time. In the air, for instance, its estimated median half-life—the time it takes for half of the virus particles to become inactive—was just over an hour. And the levels of the virus that do remain in the air are not high enough to pose a risk to most people who are not in the immediate vicinity of an infected person.

These findings are likely to assuage some fears. Homebound consumers worried about contagion from cardboard delivery boxes may have less to worry about the next time Amazon rings (unless they are used to same-day delivery). At the same time, the findings will amplify concerns about airborne transmission, which some experts had not considered possible. The research may change the way medical workers interact with infected patients, who with close contact may transmit the virus onto protective gear.

Why the virus can survive longer on some surfaces rather than others still remains something of a mystery. Maybe it has to do with the consistency of the object playing host to the virus. Cardboard, of course, is much more porous than steel, plastic or copper. But the authors noted that there was more variation in their experiment for cardboard than for other surfaces, and the results should be interpreted with caution. No doubt consumers are used to treating their surroundings that way by now.

►https://www.nejm.org/doi/full/10.1056/NEJMc2004973

“Sacrifice the weak”, urged a sign at a protest against Tennessee’s lockdown on April 20th—though whether the person holding it was trolling the other protesters is unknown. Some claim social distancing is pointless, since covid-19’s elderly victims would soon have died of other causes. In Britain many pundits have said that two-thirds of the country’s dead were already within a year of passing away. They cite an estimate made in March by Neil Ferguson, an epidemiologist at Imperial College London who advises the government.

Mr Ferguson notes that two-thirds was the upper bound of his estimate, and that the real fraction could be much lower. He says it is “very hard” to measure how ill covid-19’s victims were before catching it, and how long they might have lived otherwise. However, a study by researchers from a group of Scottish universities has attempted to do so. They found that the years of life lost (ylls) for the average Briton or Italian who passed away was probably around 11, meaning that few of covid-19’s victims would have died soon otherwise.

First the authors analysed data for 6,801 Italian victims, grouped by age and sex for confidentiality. About 40% of men were older than 80, as were 60% of women. (The virus has killed fewer women than men, perhaps because they have different immune responses.) The authors excluded the 1% of victims under 50. Then they calculated how much longer these cohorts would normally survive. Life expectancies for old people are surprisingly high, even when they have underlying conditions, because many of the unhealthiest have already passed away. For example, an average Italian 80-year-old will reach 90. The ylls from this method were 11.5 for Italian men and 10.9 for women.

Then the authors accounted for other illnesses the victims had, in case they were unusually frail for their age. For 710 Italians, they could see how many had a specific long-term condition, such as hypertension or cancer. The authors used a smaller Scottish sample to estimate how often each combination of diseases occurs among covid-19 victims. Finally, they analysed data for 850,000 Welsh people, to predict how long somebody with a given age and set of conditions would normally live.

Strikingly, the study shows that in this hybrid European model, people killed by covid-19 had only slightly higher rates of underlying illness than everyone else their age. When the authors adjusted for pre-existing conditions and then simulated deaths using normal Italian life expectancies, the ylls dropped just a little, to 11.1 for men and 10.2 for women. (They were slightly lower for Britons.) Fully 20% of the dead were reasonably healthy people in their 50s and 60s, who were expected to live for another 25 years on average.

The researchers warn that their data exclude people who died in care homes, who might have been especially sickly. Nor can they account for the severity of underlying illnesses. For example, covid-19 victims might have had particularly acute lung or heart conditions. More complete data could produce a lower estimate of ylls. Mr Ferguson also points out that tallies of all-cause mortality will contain clues. If the pandemic has merely hastened imminent deaths, there should be fewer than usual once covid-19 is under control.

Still, the available evidence suggests that many covid-19 victims were far from death’s door previously, and cut down at least a decade before their time. Allowing the virus to spread freely would sacrifice the strong as well as the weak. ■

Sources: “Covid-19 – exploring the implications of long-term condition type and extent of multimorbidity on years of life lost: a modelling study”, by P Hanlon, F Chadwick, A Shah et al., 2020; Istituto Superiore di Sanità (Italy); SAIL Databank (Wales); Public Health Scotland
This article appeared in the Graphic detail section of the print edition under the headline “Before their time”

▻https://wellcomeopenresearch.org/articles/5-75

Le papier (librement accessible, très technique, pas UN graphique pour représenter l’effet…)
The fertility effect of catastrophe : U.S. hurricane births
▻http://www.econ2.jhu.edu/people/hu/fertility_jpope2010.pdf

Abstract Anecdotal evidence has suggested increased fertility rates resulting from catastrophic events in an area. In this paper, we measure this fertility effect using storm advisory data and fertility data for the Atlantic and Gulf- coast counties of the USA. We find that low-severity storm advisories are associated with a positive and significant fertility effect and that high-severity advisories have a significant negative fertility effect. As the type of advisory goes from least severe to most severe, the fertility effect of the specific advisory type decreases monotonically from positive to negative. We also find some other interesting demographic effects.

#merci !

joli exercice de régression. Si le résultat de l’exercice donne une différence entre les deux types de régime «  hautement significative statistiquement  », il y a vraiment beaucoup de choses à dire sur son utilisation «  politique  ».

Note : j’arrive à retrouver sans trop de problèmes les données en utilisant les deux sources mentionnées, ce qui donne un nuage de points a priori à peu près similaire (mais il faudrait voir dans le détail…) ; mais je ne dispose pas d’une classification des pays par année et par type de régime.
Apparemment, il s’agit de l’ensemble des désastres biologiques, soit les épidémies, celles-ci pouvant être bactériennes, parasitaires ou virales.

Entre 1960 et 2020, je trouve 1026 «  désastres biologiques  » ayant provoqué au moins 1 mort (il y en a 187 où le nombre de décès n’est pas renseigné) et quelques anomalies (ex. 523 morts pour 2 épidémies virales en Afghanistan en 2000 pour 11 personnes atteintes (affected) soit 4750% de mortalité… Il y a aussi 91 pays (dont la Martinique en 2010) pour lesquels le PNB/hab. n’est pas connu.

Le modèle utilisé établit donc un lien entre la mortalité de l’épidémie, donc une sorte de réponse sociale au désastre en fonction de la richesse du pays, en fait sa capacité à produire et son type de régime. Indépendamment donc du type de l’épidémie et de la létalité de l’agent pathogène. Avec un effet en niveau (il y a moins de morts dans les démocraties) et en pente (le surcroît de richesse diminue plus la mortalité dans les démocraties).

Je remarque, en premier, qu’il y a très probablement déjà un lien entre PNB/hab. et type de régime. Et qu’un facteur prépondérant de la réponse est la qualité du système de santé du pays (elle aussi, certainement corrélée au PNB/hab. ce qui fait que les auteurs pourraient affirmer que le PNB/hab. est un bon substitut, mais pour laquelle le lien avec le régime politique n’est pas forcément aussi clair).

Surtout, si la liaison statistique est hautement significative (les zones colorées nous montrent l’incertitude sur les droites de régression) la capacité prédictive du modèle est vraiment faible. Il n’y a qu’a voir la très grande dispersion des observations autour des 2 droites. De ce fait, les performances du modèle utilisé comme classificateur (prévoir le type de régime en fonction de la mortalité et de la richesse) seraient, me sembleraient, vraiment mauvaises. (calculs à faire…)

Enfin, il est probable que les pays à la plus forte richesse (> 30 000 $/hab.) soient déterminants (points influents) dans le résultat des deux régressions.

La variable année me semble aussi bizarrement absente du « modèle », alors qu’il me semble qu’un trait important de cette histoire est qu’à chaque épisode on apprend à mieux gérer, ce qui fonctionne et ne fonctionne pas.

En tout cas ça va à l’encontre de l’idée de la « bonne dictature », avec un effet présenté comme négatif (même si on peut argumenter sur la classification à priori et tout un tas d’autres choses).

C’est toujours binaire et tendancieux, la définition d’une dictature. #démocrature

Oui, le PNB/tête est aussi corrélé avec le temps (en gros, à quelques exceptions près, depuis 1960, ça croît) et donc, il y a confusion des facteurs et on ne peut pas isoler ce qui vient de l’avancement du temps (expérience accumulée) de ce qui vient de la « richesse » ou de l’amélioration (ou la meilleure préparation) du système de santé.

hmmm ce graphe vieillit-il bien ?