Comment by dredmorbius | Hacker News Reader

Comment by dredmorbius | Hacker News Reader

Parent Full thread

I've been keeping tabs on various atmospheric (and ocean and space) conditions via the Nullschool Earth viewer: https://earth.nullschool.org/

Different conditions and sensors can be selected, including both particulat (PM2.5 especially) and CO concentrations, both of which are strongly (though not exclusively) associated with wildfires. SO4 is another combustion-related emission, though also associated with power plants, shipping (major shipping lanes are very sharply visible), and volcanic erruptions (see especially the Kiril Islands, north of Japan).

Looking at PM2.5, Here's the current view over Brazil:

https://earth.nullschool.net/#2019/08/24/1700Z/particulates/...

And over Congo:

https://earth.nullschool.net/#2019/08/24/1700Z/particulates/...

And a year ago, 24 August 2018, over ...

Brazil:

https://earth.nullschool.net/#2018/08/24/0500Z/particulates/...

Congo:

https://earth.nullschool.net/#2018/08/24/0500Z/particulates/...

You can see other sensor information or skip forward or back a few days or weeks to see what trends are like.

I'd not done this until just now, and frankly, it appears that last year's smoke activity was actually greater, though the 2019 fires may have been worse several days ago (there's been pronounced global political backlash).

The "day of burning" was the 19th as I recall, let's look back a few days ... OK, unlike the timeslices above, this one is selected as a more active capture, on August 19, 2019, though a few later periods are comparable:

https://earth.nullschool.net/#2019/08/20/0300Z/particulates/...

OT: Watching for unusual activity can be interesting. For example, it appears one of the East African Rift volcanos may just have lit up in Ethiopia, probably Erta Ale (the most active continuously errupting volcano on Earth), though there are others in the general vicinity:

https://earth.nullschool.net/#2019/08/24/1500Z/chem/surface/...

(All Nullschool links should be to date-anchored snapshots. You can expand the controls by tapping "Earth". "Now" will show current / most recently reported conditions.)

Thanks for the links. Very interesting.

> I'd not done this until just now, and frankly, it appears that last year's smoke activity was actually greater

Someone linked to a NASA page a few days ago that said the same thing.

"As of August 16, 2019, satellite observations indicated that total fire activity in the Amazon basin was slightly below average in comparison to the past 15 years".

http://web.archive.org/web/20190820090023/https://earthobser...

Since the news story broke they have updated it, saying it was updated to clarify sources yet they substantially changed what they have said:

"Editor’s Note: This story was updated on 22 August 2019 to clarify our data source."

The live page now reads:

"As of August 16, 2019, an analysis of NASA satellite data indicated that total fire activity across the Amazon basin this year has been close to the average in comparison to the past 15 years. "

https://earthobservatory.nasa.gov/images/145464/fires-in-bra...

I would have been a lot more honest to say "we incorrectly stated that fires were below average and have updated our position".

Wow, side note, compare China/SE Asia PM2.5 to the rest of the world.

Yeah, that's pretty stunning. Sulfur Oxide emissions also.

Though if you hit up January and look at SO4 emissions over the US and Europe, they're pretty terrifying as well.

Spotting wildfires by PM2.5 emissions is a part-time hobby, as is looking at MSLP and watching cyclonic storms developing. It looks as if the Atlantic hurricane season may start cooking off in the next week or so.

Not all channels have forecast values, but temps, winds, pressures, and precip do, so you can roll out a few days to see what the models are projecting. I've followed (and anticipated) most of the big storms over the past few years. Some false starts (a lot of swirls never really develop), but it was painfully obvious that Harvy and Florence were going to be massive storms.

In the Pacific, the unrelenting assaults on the Philippines, Taiwan, Japan, and mainland China are impressive.

China tends to buy light crude instead of light sweet crude, hence the oil and derived fuels have more sulfur than is typical for European or American vehicles, industry, etc.

This is something of a problem when a country with as many people as China and an economy as growth-oriented as China's is, uh, firing on all cylinders.

What is a good source of data for spotting wildfires? Any APIs that can give me hot spots in a region?

Generally, I'll take a look at the PM2.5 channel on occasion. Swapping that for CO (carbon monoxide) is usually a good confirmation.

While wood smoke contributes a lot of PM2.5, other sources can as well -- dust over the Sahara (and much of the Atlantic), salt spray from hurricanes, and a few others, which aren't direction or at all combustion related.

CO tends to hover near urban areas -- you'll spot concentrations especially near Los Angeles, New York, and Chicago.

Sulfur Dioxide is also related to combustion, usually heavy fuel oils. Shipping traffic lights up especially, but so do industrial sources (generators), many volcanos, and some others I haven't worked out.

Nullschool doesn't have an IR channel, but Nasa's Modis can turn that up. There are generally national or regional wildfire websites, particularly for the US, Canada, and Australia, as well as California and other western US states.

(Fire activity so far this year has been pretty low in the US.)

If you spend a few weeks looking at data (or browsing through historical data), you get a pretty good idea of what patterns are typical.

Note also: some sensors / channels get recallibrated from time to time. There was a big adjustment in CO2 measurements a year or so back, which kind of freaked me out. Whilst CO2 concentrations are gradually increasing, they did not take a massive jump in the past 5 years, despite appearing to do so from the Nullschool history.

Also, as noted, sometimes stuff shows up that's not fires. The current Ethiopian volcano erruption would be a case in point (I think it's a volcano, seems very likely).

> major shipping lanes are very sharply visible

once you're in international waters, you get to burn whatever cheap dirty fuel you want, without regulation.

.. just another problem to add to the list

Reply on news.ycombinator.com