AI and satellite imagery reveals expanding footprint of human activity at sea
globalfishingwatch.org
globalfishingwatch.org
First off, a big reason as to why - is because the laws differ from country to country. Some countries, like Norway for example, require shipping vessels over a certain length to broadcast their positions through AIS, VMS.
This may not be the case for, say, UK.
There are arguments to be had from both systems as to WHY you wouldn't want to broadcast your positions at all times - a typical one is that other could easily infer your fishing fields (fishing patterns are trivial, if you've worked with this you can easily spot what tools they are fishing with, and likely what species they are going for - don't need ML-based systems for that. Any fisher or fisheries analyst can spot the patterns), and thus go after that.
The other is privacy.
A very typical thing is that ships turn off their AIS as soon as they enter international waters. There is no enforcement of that, and many developed countries have practically zero resources to fight illegal fishing, from a technological point of view. UNDP has a program which is aimed at helping developing countries with the tech and training to detect illegal fishing, but there's a long way to go. Developing countries desperately need the data, which is either owned by governments, or private actors. AIS data is either picked up by satellites, or base stations. VMS systems are expensive, but also allow for ships to transfer catch reports and similar - but is unfortunately not always enforced in a good way.
But tech is becoming better. Satellites with NAV/marine radar detectors are in orbit. Long-range drones with sensors are a thing. Countries are hammering through laws that force ships to have certain sensors on them. Lots of ML-assisted tools for automating detection and analysis is being introduced and used.
True in the aviation world. I've noticed that most people under, say, 50 or so, like ADS-B (airplanes broadcasting their location) for the safety reasons. But most of the pilots over 50 will say "It's just the government coming up with another way to track you."
I don't understand the impetus to allow someone to fly a massive, deadly hunk of metal over a community without clarifying their telemetry to that community.
I think it's an act of wanton corruption that law enforcement is now exempt from this requirement in some situations.
> I think it's an act of wanton corruption that law enforcement is now exempt from this requirement in some situations.
The point I am trying to make: most of things flying seem more dangerous (higher speed, higher reach, different fuel) than cars.
And yet somehow flying is safer than driving! As a sibling comment mentions, a lot of that has to do with all of the safety controls in place.
My point is that driving at modern speeds in modern vehicles is actually pretty dangerous, even if not as dangerous as aviation, and yet we as a society exercise very little oversight over the activity.
There are good reasons to not have "ADS-B for cars", mostly related to privacy. Unlike flying, many people heavily rely on cars to carry out their day-to-day activities. But I think it's at least worth thinking about.
Hanging around FBOs/flight lines/bars with pilots that also ride, it's come down to this:
After countless attempts and arguments while trying to normalize accident/fatal accident rates by converting hours flown/miles ridden, statistics do support that GA does seem to be a bit more dangerous than riding.
That having been said, my opinion is that most of the GA risks are in the hands of the pilot, while most of the risks of riding are external to the rider. Most of the worst motorcycle accidents I've seen or heard of involved the rider getting schwacked by a car or truck, while most of the worst GA accidents involved the pilot schwacking him or herself with poor judgement and/or skills that are below MIF. I put a lot of effort into minimizing pilot-induced risk when I fly, so scientifically speaking and all, I -feel- safer when flying than riding, and probably am.
But I do think there's a difference in risk. Aviation has a pretty dang good record, and I think that the tradition of transparency in comms and telemetry is a huge part of that.
While this is true about commercial airliners it is not true for general aviation. The safety of general aviation is roughly comparable to motorbikes.
I believe that plenty of people would object to that level of tracking citing privacy reasons.
(I'm not against ADS-B myself.)
You could argue it's not a very good one, but it does exist:
"This rulemaking provides an exception to ADS–B requirements, removing the transmission requirement for sensitive operations conducted by Federal, State and local government entities in matters of national defense, homeland security, intelligence and law enforcement."
https://www.govinfo.gov/content/pkg/FR-2019-07-18/pdf/2019-1...
I think it's wrong that the government exempts themselves from it if they're going to require it of everyone else, but that's so far down my list of grievances with the government it's barely worth thinking about :)
For more than a decade I worked airborne geophysical surveying - hundred thousand to a million+ line kilometre grid draping at 80m ground clearance on routes planned in advance.
The planes ran stable known instrument configurations and were calibrated weekly to measure their individual magnetic and radiometric signature.
Aircraft positioning signals on long haul flights Australia -> India -> Fiji -> Mali weren't an issue.
Extraneous non required signalling during grid surveying was an issue.
A reasonable comprimise was to blip a quick position fix at the end of every fourth or fifth 20 kilometre line run (say) ... otherwise no broadcasts during instrument recording times .. and glow in the dark divers watches for the pilots or random chunks of metal casing there one day and gone the next.
In fact I wonder if it's even a licensed pilot—William Bushling apparently flew for 20 years without a current FAA license and also didn't use ADS-B; when he finally lost the dice game it took them days to find the wreckage: https://youtu.be/69NvK6YbNtg
Other sites like https://globe.adsbexchange.com or https://globe.adsb.fi don't filter aircrafts, they even show military planes (well, if they have ADS-B enabled). Though they don't have coverage above oceans, as satellite ads-b is expensive. (These sites look nearly the same, as they both use tar1090 as the webui, but they may have different coverage)
These sites, may sometimes have less coverage, as the sites you mentioned are way more popular. But if your area has coverage by these sites, you can be sure that you'll see every plane if it has ADS-B.
You could even replay, if you know the time and date when the plane passed, by appending ?replay to the URL.
A 30usd rtl-sdr dongle is enough to receive local ads-b signals, if one wants to test it.
(The fields are San Martin and Watsonville in the Bay Area, FWIW.)
Not wrong, but the tracking is to help with reducing collisions.
That said:
* https://www.faa.gov/air_traffic/technology/equipadsb/privacy
You need to go with 1090ES(-only), and not the 'consumer focused' UAT, though.
guess there is the corruption issue?
- Fisheries may not (historically or present day) be a huge part of their economy, and has therefore been neglected for years.
- Not enough infrastructure due to economic neglect.
- Huge revolving door of officials. One month you deal with one person, six months its another one. You don't know if they are real professionals, or someone placed there (party stooge, nepotism, you name it).
- No cooperation between fisheries/marine agencies, and navy/police/coastal guard.
- And, yes, some corruption I'd assume.
I have a concrete example:
We were invited to a developing country to assess their systems, and consult them on how to move forward. They were/are losing millions and millions due to illegal trawling.
Arrive at the HQ, which was a run down office where a handful of people were working. All data was shared via excel spreadsheets, no real systems to work on, lots of paper forms that someone had to digitize. Someone looking at MarineTraffic from time to time.
That's the state of some of these countries.
The illegal fishers are long out of their EEZ before anyone can react.
This can be very efficient if you spot check people from time to time, and simply cut the hands off of anyone found to be violating the law.
Tell people to calculate and pay their own taxes too, with the same punishment if they falsify stuff.
That's how Afghanistan dealt with opium farmers before 2001, and it was amazingly effective.
however, one (of many) problems is that the poachers make more money than the employed bureaucrats .. step 2 - follow the money
There are EU regulations requiring fishing vessels over 15m to use AIS (the UK also abides by these) so within Europe the standards are reasonably consistent. Regulations are considerably more relaxed or less effectively enforced in other jurisdictions which is reflected in the data; it's possible the 98% of all fishing vessels operating off Thailand without AIS are all legal. But given the reputation of larger vessels operating in that region for slavery, perhaps not.
Agree there are privacy-related reasons why you might ideally not want competing vessels seeing which waters you fish, but AIS alone doesn't tell anyone whether it's any better than other waters (and fishermen have no shortage of other data points to plan their activity or other boats to follow if they really don't have any better heuristics). Regulatory evasion is rife even in effectively-regulated areas not least because the industry is convinced some of the regulations are actually wrong, especially when it comes to localised quotas. You knows there's an issue when a former operator of a commercial fleet in Europe tells you candidly, that their skippers (who are all tracked by AIS/VMS) would definitely expect any catch logging system to allow them to lie about which points on their voyage yielded most of the fish....
Ya, they are called spy sats. For the better part of a century, very smart people have been funded by people with very deep pockets to track ships at sea that don't want to be tracked. If anyone thinks their startup sat company is going to start tracking ships with passive radar, they are entering a very very mature field.
https://en.wikipedia.org/wiki/Lacrosse_(satellite)
https://en.wikipedia.org/wiki/Naval_Ocean_Surveillance_Syste...
>> The Naval Ocean Surveillance System (NOSS) is a series of signals-intelligence satellites that have conducted electronic signals intelligence for the U.S. Navy since the early 1970s.[1] The first series of satellites were codenamed "White Cloud" or "PARCAE", while second- and third-generation satellites have used the codenames "Ranger" and "Intruder".
And for anyone saying "well, our sats are going to be in lower orbit" ... that has been done too. If you ever read about satellites powered by nuclear reactors, those were low-orbit radar sats looking for ships.
But, as someone who's working on exactly that, it is not easy - and there's a lot of red tape to wade through. The data gets filtered by the military, and then you get whatever scraps are left. Scans/passes of areas with too much military activity tend to be off-limits.
A big threat is that people could start to fingerprint radar signatures, which in turn isn't ideal if you're a, say, navy or intelligence ship.
Not that your adversaries (read: China, Russia, etc.) probably don't have that exact type of data, but it's just standard practice to minimize exposure.
Yeah but as long as no one dares to stand up to the worst offenders - particularly China, who have been under fire for years from their neighborhood in the Philippines [3] all the way to Africa [1][2] - with actual navies to stop and either seize or sink non-compliant ships all of this will be in vain.
The open sea is the last remaining place on Earth where darwinism in its purest form still rules. Might makes right...
[1] https://www.telegraph.co.uk/global-health/climate-and-people...
[2] https://www.voanews.com/a/fishy-business-report-details-chin...
[3] https://apnews.com/article/china-canada-philippines-illegal-...
Luckily I have other safety systems in place but damn is that a buzz kill at 3am on a new moon night.
Also another reason you would turn off AIS is if you are hauling precious cargo and don’t want pirates to know your whereabouts. I know this sounds cliche but it’s true. Cargo vessels traveling near pirate activity will often “go dark” to avoid detection and dealing with a Capt. Phillips situation.
So yeah, it’s difficult to get a fix on just how many vessels are out there, big or small. I’ve seen a lot of weird designs and contraptions out at sea.
https://earth.nullschool.net/#current/chem/surface/level/ani...
Wild when you see just how much emissions are still being released and still presumably cooling the Earth, meaning the effects of climate change we're seeing now are still likely a dampened version of the true long term impact.
Was there an outright study of the "main culprits" part of this? As I recall there was some evidence but then the main discourse was based on a lot of extrapolation by a viral tweet.
I consider "suspected" to be the least level of evidence while still taking something into consideration as a potential cause. A suspected murderer might not even have been arrested, let alone convicted.
We do know that sulphur emissions have a global cooling effect, and we do know that sulphur emissions recently were reduced, so it's a reasonable hypothesis from first principles.
Starting in 2020 there's a limit:
* https://www.imo.org/en/MediaCentre/PressBriefings/pages/02-I...
What is your specific PhD interest?
Seems like a really good application for Jerry Pournelle's "Rods From God kinetic weapons concept[0].
Trying to make protected marine animals disappear? You might just disappear.
(Zero sympathy for any kind of poaching; we're making ecosystems go extinct fast enough)
[0] https://nationalinterest.org/blog/reboot/rods-god-strange-su...
Edit: cool examples here https://github.com/GlobalFishingWatch/pyseas/blob/master/pys...
And thanks to multiple factors, it's becoming easier by the day.
https://subredditstats.com/r/avoidchineseproducts/#:~:text=c...
> The data shows that factories are producing less and hiring fewer people," said Dan Wang, chief economist at Hang Seng Bank China. "(The data) could also show a loss of confidence in government policy," she added, warning factory activity was unlikely to improve anytime soon as other economic problems dominate. [0]
> A growing number of companies are looking to move their manufacturing out of China. Companies are examining their heavy concentration in China. Boards and companies are reevaluating their risks and reviewing mitigation strategies. [1]
[0] https://www.reuters.com/world/china/chinas-factory-activity-...
[1] https://www.forbes.com/sites/betsyatkins/2023/08/07/manufact...
But I would rather have ex. the problem of how to fix plastic in the oceans, than not have access to plastic at all.
Cause all problems are soluble and all solutions to problems create new problems. They are however better problems.
Disclosure: I work for a satellite imagery company. As far as I know, our data is not usable for the level of tracking described in your comment. However, I likely have a personal bias towards data transparency. Opinions expressed are my own.
The issue is that would need constellation covering the whole Earth. That is doable, but not needed for observation so nobody has done it. Ships have an empty background and can be tracked intermittently while people and cars require continuous tracking.
Drones would be better, but would still need a huge number to provide continuous tracking. CCTV cameras would be easiest but limited in area. Finally, tracking cell phones is doable, and easier to do with satellites.
What is the actual ML behind this?
I'm currently in the middle of my phd where I am working with these types of data and methods, its extremely interesting.
[1] David A. Kroodsma et al. ,Tracking the global footprint of fisheries.Science359,904-908(2018).DOI:10.1126/science.aao5646
In that case it is more about identifying the actual fishing activity than the fishing boat itself. Each mode of fishing (which requires different types of gear and types of ships) will leave a distinct signature. For example bottom-trawling will require a certain (low) speed (here a big net is dragged behind the vessel), purse seines will make somewhat circular patterns (they lay out a net in a circle around a school of fish). So often you want to identify these fishing activities to, for example link them up with a later delivery of catch, but also to investigate and estimate things like fuel usage.
Even if you don't have the IMO number of the ship you can look for these patterns, so it can be redused to whether the ship is in proximity of a fishing ground, if they display the movement patterns that is the signature of fishing activity, and later if they meet up with other vessels or go in to land to deliver their catch. A recreational boat (given that it is large enough to be suspected to be a fishing vessel just by where it is and how large it is) would perhaps have a constant high velocity (or complete stand still) for instance.
A simplified example: a person is jogging on the road that passes your apple trees you would not suspect them of stealing apples, if they are slowly walking - maybe (but if they stay no the road - probably not), if they are jogging through your garden of apple trees its certainly suspicious, but if they are slowly walking through your garden I would send a coastal guard ship ;)
A less simplified one would be to try to match up fishing related activity with how much fish a vessel is delivering (is it delivering 20 tonnes but have spent two full weeks with fishing activity?). Then maybe see if they have the right permissions to do that kind of fisheries in that zone at that time of the year etc.
I know nothing about the industry but I did wonder if an innocent excuse is that as a fishing vessel you wouldn't want to tell everyone else where you are fishing?
I have several relatives that do fish. This includes long line tuna fishing and reef fishing.
More likely they don't want to follow quota, rules about what they can or where they can fish.
https://en.wikipedia.org/wiki/List_of_countries_by_seafood_c...
There are actually satellites that do it. Over the course of many days. And it's not easy or quick to act on their feedback. Something like a military or police patrol simply can not happen.
The problem is this would cost millions of dollars and take weeks or months unless you got very lucky. The Navy could do it in their sleep but why would they? Nobody who cares enough about this problem has the means to execute it, and nobody with the means cares.
The reason they don't is the Navy is not a law enforcement agency and this is not their mission.
https://web.archive.org/web/20121220000930/http://www.afp.co...
> This Chinese fleet is also categorized by The Global Initiative Against Transnational Organized Crime as the largest purveyor of illegal fishing in the world. Our reporting revealed Chinese vessels illegally entering the waters of other countries, disabling locational transponders in violation of Chinese law, breaking U.N. sanctions that prohibit foreigners fishing in North Korean waters, transmitting dual identities (or “spoofing”), finning of protected shark species, fishing without a license, and using prohibited gear.
In fact, bet a few of them are here on HN right now and could probably give us a first hand account while on the high seas trying to browse the only bit of the internet that still works with a 32kbps high latency internet connection...
Sink the boats, problem solved. Presumably another, much larger, problem will be created. But the illegal fishing might stop.
Send in the coast guard, impound the ship, assess big fines. If the fines go unpaid, sell the ship.
https://globalinitiative.net/wp-content/uploads/2021/12/IUU-...
For flag state responsibility (Table 16), Taiwan (#2), South Korea (#4), Japan (#6), Spain (#7) are also among top-10 global worst offenders.
https://globalinitiative.net/wp-content/uploads/2021/12/IUU-...
1 China 2 Russia 3 South Korea 4 Somalia 5 Yemen 6 Taiwan 7 Ukraine 8 Eritrean 9 Egypt 10 Lybia
In case it's not obvious, the entire point is that the actual statistics is not very convenient for the "good-guy v. bad-guy" framing. Why else would I only highlight countries considered in the US as part of the "good guys" camp?
While we're on this topic, US's own ranking drastically worsened between 2019 and 2021, from being #66 worst globally (see p.121) to #27 worst (see p. 111):
> When considering prevalence, five countries – South Korea, Seychelles, the USA, Senegal and Saint Vincent & the Grenadines – entered the list of worst-performing countries in 2021 for the first time, with China, Taiwan, Vietnam, Thailand and Ecuador remaining in the worst performers. (p.28)
But you’re right - everyone is screwing the planet up in their own messed up ways. I think rather than making bogey men out of the “bad guys,” we should realize this fact. But with that fact realized, China does act with a particularly egregious lack of responsibility and accountability that belies their stature.
I agree with your larger point, but I'm not sure this is entirely accurate. If you look at table 4 on p. 32 (i.e. printed page number; PDF page number is 21, and the complete data can be found at the end of the PDF, starting on (printed page number) p. 110), you'll see that the scores are derived from three indicators: (i) the prevalence of IUU (Illegal, unreported, and unregulated) fishing, (ii) each country's vulnerability to IUU fishing, and (iii) their response to it. Only in the Prevalence table for 2021 does China "weight 25-30% worse" than #2 (South Korea). In the Vulnerability table, China's score is only marginally worse (3.3-5%) than Japan, the second-worst in both 2019 and 2021. In the Response table, China (3.31) is not even among the top 10 worst countries, scoring comparably to Israel (3.29) and better than Singapore (3.87) and UAE (3.82).
Also worth noting: the Prevalence score is calculated based on the number of vessels recorded on IUU lists (see p.12), so we'd expect countries with a larger number of fishing vessells to score higher (i.e. worse) in respect of Prevalence, which is why even the US made it to #8 worst in the Prevalence table for 2021.
> For some reason your deep linking isn’t bringing me to the tables you’re referring to
I don't have the machine to test this right now, but it's possible that the ?page=N search parameter trick only works in Chromium.
Ocean fish are being literally depleted, with bluefin tuna alone being down 97% from historical levels. [1]
Commercial fleets use sonar to isolate and annihilate entire populations. Nets have been as wide as 50 miles, resulting in massive indiscriminate marine death as bycatch. Only 2.5km nets are legal now- still more than wide enough to result in a lot of bycatch, and there is no reason to assume these dark fleets are abiding by UN law. [2]
[1] https://www.nationalgeographic.com/science/article/sea-runni...
[1] https://en.wiktionary.org/wiki/sunlight_is_the_best_disinfec...
Because I want there to still be fish to eat in 20 years and enforced public view will make illegal overfishing harder.
1. https://en.wikipedia.org/wiki/Collapse_of_the_Atlantic_north...