[al_biglan]

"preindustrial" being tagged at 1900 seems more than a bit off. Steel boomed in the 1880's (starting in 1850's with Bessemer). I would think a better data set would be to look back to 1850 or better 1800 to be sure to capture a decent number of years that show the "preindustrial" time period.

All that said. It is nice to see the trend over the past 30 years!

[maxerickson]

Half the steel ever made has been produced in the last couple decades.

1913 production was ~60 million tons (http://www.freerepublic.com/focus/f-chat/2892944/posts ).

2015 production was ~1500 million tons.

[al_biglan]

True, and we are getting cleaner production methods. The early production was incredibly inefficient and produced way more pollution than now. I can say _that_ with some certainty, but I can't say that they produced x% of the current production back then and were N times more polluting. Plus, it isn't clear how the atmosphere absorbs/reacts to the pollutants. (is there a lag? why the spike in 1970s? did the levels really fall off so quickly? I thought once steel left the US production it got much "dirtier" to produce?) Lots of questions, I'm mostly asking to show me a bit of the range that is clearly outside the definition of "preindustrial"

[kogus]

For the record, steel production has not left the US (although many of the jobs have due to automation).

The US is #4 in the world for steel production[1], although China produces about 17x more.

[1]https://en.wikipedia.org/wiki/List_of_countries_by_steel_pro...

[dpark]

US steel production in absolute terms has dropped 45% (edit: 31%) from 50 years ago, when we made about 25% of steel globally. Now we produce less than 5% of steel globally. I'd say steel production has largely left the US.

[DougWebb]

The Wikipedia article shows US production dropping 30% from 1967 to 2015. It's a loss, yes, but 70% of the production is still here. What's changed is that China and other countries are producing much more now, so as a percentage of world-wide production the US has dropped. But that'd be true even if we were producing more domestic steel than before.

[philipkglass]

The other big thing is that US per-capita steel consumption (including imports and finished products) is significantly down since the 1970s.

This says that US steel consumption per capita went from 622 kilograms in 1979 to 451 kilograms in 1988: https://www.worldsteel.org/en/dam/jcr:373b3927-26ff-4591-8e7...

More recent numbers here say 348 kilograms per capita in 2003 down to 297 kilograms in 2015: https://www.statista.com/statistics/185031/apparent-steel-us...

With a US population of 225 million in 1979 and 322 million in 2015, that means that total US steel consumption is only 68% of what it was in 1979. It doesn't look like domestic production has declined much (if any) faster than domestic consumption since the end of the 1970s.

Cars are lighter than in the 1970s, cars last longer than in the 1970s, and non-steel materials are increasingly used in roles that steel was used for in the 1970s. Like parts of cars and making beverage cans.

[dpark]

> The Wikipedia article shows US production dropping 30% from 1967 to 2015. It's a loss, yes, but 70% of the production is still here.

Fair, I plugged my number in upside down. The production 50 years ago was ~46% higher than today, or today's is ~31% lower.

> What's changed is that China and other countries are producing much more now, so as a percentage of world-wide production the US has dropped. But that'd be true even if we were producing more domestic steel than before.

Yes, but we've also dropped 30% of our own production. When our absolute production is down this much and our overall share of production has plummeted, I think it's fair to say the industry has largely left. The population is also up by over 50% since 1967, so per capita our steel production has basically been cut in half.

In comparison, the absolute number of consumer electronics that America produces might actually be higher than 50 years ago, but as a share of the market, production has plummeted, to the point that everyone would likely agree that consumer electronics manufacturing has largely left the US.

[Retric]

The backdrop for this is each location has gotten more productive in part because most of this is recycled steal. So, the US has lost far more than 30% of it's steal mills.

AKA, net production of new steal in the US from ore is what has has really declined.

[jcranmer]

To understand steel production, you need to look at three numbers, not one: production, capacity, and consumption.

One of the things that has happened since 2000 is that China has expanded its steel production capacity a lot. The capacity utilization--which should optimally be around 80-90%--is at 75% right now and plummeting. China basically built its steelmaking capacity assuming consumption growth rates would increase instead of what they actually did (decrease).

The obvious solution is to close down steel mills. Unfortunately, many governments consider steelmaking to be a point of national prestige (the EU started out as the European Coal and Steel Community!), to the point that democratic governments often idolize steelmakers (with other industrial workers) as the "every-man" [1]. Which means no one wants to cut their capacity.

[1] I find this somewhat ironic given that, now that I think about it, no one in my family that I can trace ever worked in a factory of any kind. Although my grandparents and some of my second and third cousins are farmers, which was historically the other idolized profession.

[Roodgorf]

Looking at the amount produced by year it seems like U.S. production has been relatively steady over the last year and not too much lower than it was in previous decades. I think looking at percentages doesn't really tell the full story here, since China's production has increased so much as to overshadow most other countries' output. The "drop" from 25% to 5% wasn't as much due to production leaving the US as it is production not increasing at breakneck pace.

[dpark]

> I think looking at percentages doesn't really tell the full story here, since China's production has increased so much as to overshadow most other countries' output. The "drop" from 25% to 5% wasn't as much due to production leaving the US as it is production not increasing at breakneck pace.

But that is extremely relevant. When our production is flat (actually declining) while world production is soaring, it means that we control very little of the market. We don't control supply. We don't control pricing. This puts our production at higher risk because it's much easier for China and India to price our industry out of business.

Our production per capita has also been cut in half since 1967. So in addition to the world producing more, the US is actually producing a lot less. Our consumption is also up, and we're producing only about half of what we consume now.

[kaybe]

If you don't inject it into the stratosphere, most pollutants such as acids or particles are precipitated out (dry or wet) pretty quickly (on the order of days to weeks/months), so I'd expect a lag of max 2 months due to the transport time to Greenland (and it will be a bit lower due to the distance and loss over time). [1]

(If China did a better job of filtering the SO2 from exhaust, this particular air problem there would be gone within a week.)

[1] http://acmg.seas.harvard.edu/people/faculty/djj/book/bookcha...

[anexprogrammer]

The spike in the 70s will have been the steady growth of cars and re-industrialisation since the war. The fuel crisis was 73, for a couple of years, which led to queues at pumps, rationing and introduction of speed limits in many places. That led to recession and the decline of some heavier industries... and was a factor leading to the election of Reagan and Thatcher at the end of the 70s, start of 80s. Their restructuring led to the terminal decline of some heavier industry.

In 79 there was a UNECE conference[1] on pollution that most members signed up to. 1985 gave us the Sulfur agreement that UK and US (amongst others) never ratified. The mid 90s saw further reduction of sulfur, that finally the UK adopted, though the US didn't.

That led to wider use of SO2 scubbers and tighter emissions, though initial efforts only stemmed the increase due to rising car numbers during the 80s. The agreement has had regular updates ever since, leading to lower targets, adding pollutants, adoption of catalysts etc. Most of the acid rain is now in the oceans of course.

It's still all under the UNECE so excludes the growth economies and the developing world so in today's world it excludes much of the world's industry.

That was nearly all from memory - there will be mistakes. :)

It's worth noting that at the start of the 80s acid rain, its causes and effects, were on the news almost like climate change is today. It was every week - seeing pieces showing dieback in Scandanavian forests, lake acidity, some previously unknown effect and on and on. In the UK it was a weekly reminder the Thatcher govt didn't believe. We learned the UK had tried to limit Sulfur in the 30s and had started fitting SO2 scrubbers on power stations - between the wars! The war meant that was the last heard of it until the 80s and 90s.

[1] https://en.wikipedia.org/wiki/Convention_on_Long-Range_Trans...

[Ericson2314]

Ozone and acid rain, things from the biology book that we no longer need to worry about. Hopefully C02 has a nice clear peak in retrospect someday too :|.

[maxerickson]

Presumably China adopted some of the pollution control technologies developed in the 1970s as it industrialized.

edit: It's pretty likely that steel production since 2010 is greater than all steel production prior to 1900. I would think that the energy used to make steel since 2010 is at least comparable to the energy used for all metal production prior to 1900.

Industrialization certainly got started prior to 1900. I would say it took a while to really hit it's stride.

[nickff]

Thanks for taking the time to give a clear view of what's happened. Lots of us spend time trying to make a point (on one side or the other), and not enough spend time trying to clarify what's actually going on.

[rjsw]

I think it needs to start a lot earlier. I would expect that a lot of the acidity would have come from the UK, we were to blame for a lot of the acid rain that affected the Black Forest.

[btilly]

A combination of short smokestacks and global weather patterns mean that while England created a lot of local problems for itself, it wouldn't show up well in the Greenland ice sheet record that is being sampled here.

[olkid]

Agreed. The direction we're trending is good, but isn't that entire range [4.99-5.75] considered "acidic"?

[jplasmeier]

Rain is normally acidic due to dissolved Carbonic acid.

Source: http://chemistry.elmhurst.edu/vchembook/190acidrain.html

[olkid]

"Normal rainfall is slightly acidic..." ~ thanks for the info