LiDAR collection is pretty dense from a data perspective. When I was working with helicopter collection teams we would regularly collect 1TB of point data per day at around 120 points per square meter. Based on the density of the canopy in Central America, the teams are probably using a density greater than that, and are using "multiple return" depth mapping so that they can see through the forest.

Processing LiDAR data is also quite expensive in terms of the hardware required and expertise of the people needed to correct the data back at the office. Typically LiDAR processing workstations have around 128-256 GB of RAM and a large RAID array of SSDs to quickly perform corrections interactively.

After the initial correction, most of the data is thrown out or "filtered" down to just the points of interest to save on space. Since it's quite easy to end up with hundreds of TBs worth of point data within a few months of collection. If the team is using colorized LiDAR, orthographic, and oblique images then that can take up even more space.

So the short answer to why more of the world hasn't been mapped with LiDAR techniques is money, time, and storage space. But with self driving cars now making LiDAR more prevalent, and storage space becoming more affordable every year, mapping whole regions with LiDAR is a more reasonable task.

NOTE: One of my former jobs was working with aerial LiDAR collection teams to manage the data and processing of the collected assets. This was around 2012 so some of the numbers may have changed since then.

Thanks, this is the kind of information I was curious about receiving in a concise digest.

So ultimately it sounds like I have to wait a decade or so before we can really proliferate, and to speed it up need a few more eccentric billionaires to get into funding some cool projects.

I'll use your information as jumping points to find out more.

My hopes lie with the open sourcing of in-vehicle processing hardware and software that is being used within self driving cars. As of today my understanding is much of that is part of the "secret sauce" of companies like Waymo and others.

If it were possible to apply kinematic geo-correction along with point filtering and feature detection during collection then a lot of the "heavy" data could be reduced down to what is just needed for human verification.

If you want to start your investigation with existing projects you could have a look at what the United Kingdom is doing with their government LiDAR data sets[0][1].

[0] https://www.gov.uk/government/news/environment-agency-uncove...

[1] http://vterrain.org/Locations/uk/

Brilliant! Thank you. I wonder now if Canada has any projects remotely similar occuring

Maybe the technology will become accessable enough that everyone with a smart phone can start finding and destroying archaeological sites.

Well, username kind of checks out.

I like the tech because of its ability to prevent the destruction of archaeological sites while still allowing for discovery that would otherwise require an immense amount of digging and other labours.

If my hasty math is in the ballpark, the raw data for 500 trillion square meters would be a few hundred petabytes. Not cheap, but cheap enough that I would guess the cost and logistics of acquiring the data are much more of an obstacle.

Yes that is true, logistical support and processing have considerable costs. For aerial LiDAR collection via helicopter you're looking at about $20,000 per day of onsite collection including the helicopter team, insurance, and support. That gets you high density low altitude assets, not including the processing and storage costs.

It's cheaper to fly with a fixed wing aircraft, but you will typically get lower density assets and the possibility of obscured imagery due to cloud coverage because of the higher collection altitude. However the advantage is you're able to cover a wider collection swath with each pass, and it's a lot easier to fly a fixed wing vs. helicopter so the pilots are cheaper.

How much would the total weight be for the lidar equipment? I wonder if drones would be able to significantly reduce the price.

That depends a lot on the density of point coverage you need. Most LiDAR receptors weighed between 100-300lbs five years ago. They have become much smaller now, again due to self driving cars. However I am not sure that the hardware which is suitable for self driving applications would be "survey grade" appropriate for forest canopy and ground truth mapping. I'm also not sure what the power requirements of the newer LiDAR arrays are, but the weight for batteries is significant as it directly impacts your time "in the air" between ground support stations. In a helicopter you can power all the instruments off a DC power supply from the turbine.

Also typically you want an absolute vertical accuracy of +/- 5cm during flight in order to confidently map the surface features below foliated vegetation cover. This requires quite specialized differential GPS equipment as well as a radio/cellular connection to a commercial GPS ground station network. With the improvements in GLONASS/GALILEO constellations maybe the ground networks are not as critical anymore.

Met a Cal archeology ass’t prof doing this, quadcopter-mounted lidar.

Perhaps neural networks can help with filtering and segmenting the point clouds as well?

A few hints from the article: 810 sq miles mapped, a three year project and lidar flying by helicopter. Another giveaway is the search for MH370. In some dimensions Earth is still huge!

I wonder how much city life biases us to think the planet is smaller than it is. Perhaps it explains fears of overpopulation.

It’s both bigger and smaller than we often think.

In a straight line, it’s smaller than we might expect: People can (and do) walk across continents, sail across oceans in one-person boats, fly all the way around in 2 days, visit every country at least once during their life, etc.

But of course the surface is two-dimensional, we can’t see very far at a time. The land surface is about 150 million square kilometers. (For reference, New York’s Central Park is about 3.4 square kilometers.)

It’s also “fractal”, in the sense that as we look at smaller and smaller pieces, we keep noticing more and more details. Every little patch of soil is teeming with tiny animals, plants, fungi, and bacteria.

Or if you like, there are probably at least 10 million restaurants in the world (maybe an order of magnitude more, it’s hard to guess). If you eat at 3 different restaurants every day for your whole life, it will take more than 100 lifetimes to eat at all of them. Then there are the 150 million books to read. Etc.

To me, fears of overpopulation have less to do with how much of the earth serves as living space, but how much human activity currently impacts nearly every corner of the planet. I think it's plausible that we can lessen human impact without actually decreasing population. Not sure it's likely, though.

I recall a letter to Science from the mid 70s, rebutting The Limits to Growth. I have tried to find it more-or-less lately, with no luck. As I recall, its author argued that the Earth could support well over 10^8 people in Trantor style. He assumed unlimited energy from fusion, and substantial food synthesis. Only humans, pets, crops, livestock, and requisite insects and microorganisms would survive. The planetary surface would support a building ~16 km high. At that height, the surface would be in thermal equilibrium at ~1500°C.

But hey, if that were doable, why not just build space habitat's?

10^8 is a hundred million people, so yeah, it's safe to say the Earth can support well over that many humans.

Oops, I was brain dead there. I meant 10^11. Or maybe even 10^12.

In other words 100 billion to 1 trillion. The world’s land area, excluding Antarctica, is about 50 million sq. miles. With a comfortable population density of 1000 ppl/sqm (like in South Korea, Holland, Israel, or the state of New Jersey) you would get to 50 billion (same as Trantor). The population density of Singapore is close to 20k ppl/sqm, that would give you 1 trillion.

If you are willing to pad the Earth with buildings 16 km high, you could pack at least 1000 times more people, more likely 1 million times more.

Maybe that letter to Science mentioned 10^18?

I don't remember the number. What I recall is just that it was an absurdly huge estimate. And I'd love to have a copy of that ~1975 letter, if anyone has access to requisite tools.

And, in case it's not obvious, it's not a world that I'd care to live in.

Because it is cheaper and easier to build stuff down here.

Space habitats require thinking in terms of systems, dealing with inputs and outputs. Humans haven't mastered systems thinking yet.

https://en.wikipedia.org/wiki/Tragedy_of_the_commons

I guess. I've been reading Banks' Culture series ;)

I would say quite a lot.

A private small-aircraft flight over any populated area (other than the biggest cities) reveals just how sparse most towns are. Travel almost any direction away from a main highway and in a minute or two you'll be looking at trees, trees, trees (or desert/mountains/fields).

It helps give one an entirely new perspective when we're used to traveling along roads and seeing unyielding flows of civilization, when the reality is those are mere ribbons of development criss-crossing what is still mostly empty, disinterested nature.

Ahem, but the trees you then see are usually very civilized ones. Cultivated and harvested like an ordinary field.

There are very few wild areas left. Usually only at far off areas or very steep terrain.

see also:

https://xkcd.com/1338/

Where do you live? I live in the Pacific Northwest and the forests around here are definitely not cultivated.

Overpopulation at a certain level of consumption is a valid concern.

If all 7.4 billion people on the planet all consume 3 burger patties a day, all drive internal combustion powered vehicles, and all have a new iphone each, that would very quickly require an extreme amount of resource consumption - it is unsustainable.

Of course, if you just want "as high a population at any cost" the earth can probably sustain trillions of humans, but the bulk of those humans would have no clean water to drink, with no access to clean air, no access to a livable wage, no reasonable access to hospital and emergency services, etc.

You could have a trillion slaves, all pampering the lives of a the select few hundred billionaires who own everything. That's the real fear of overpopulation.

For scale, the Amazon jungle is a couple million square miles.

I guess I'll just have to study up on the tech. IIRC there was a fairly quick LIDAR demo put on by a team at MIT that scanned the campus and surrounding areas in a relatively brief amount of time using a drone, I think.

My ignorance on the subject of LIDAR has me wondering why we can't use a faster traveling vehicle, or would it be possible to do perform scans with something like a fleet of balloons that stream the data back vs one static device.

Some rudimentary googling, MIT campus is 168 acres or .26 sq miles. So the 810 sq miles in 3 years is still pretty impressive.

MIT also has wifi. And power outlets.

LIDAR data requires precise location-information of the scanning-device, otherwise the data you collect is at best useless, and at-most wildly skewed. This is because all the points are in-reference to the location of the device. A faster traveling vehicle inherently means less-accurate location measurements.

Can SLAM be applied in this case?

It definitely can, though accuracy would always be a big concern.

IIRC, the Google self-driving car used LIDAR. Though their localization algorithms included many sensors, with the LIDAR data was just one of them.

> In that Empire, the Art of Cartography attained such Perfection that the map of a single Province occupied the entirety of a City, and the map of the Empire, the entirety of a Province. In time, those Unconscionable Maps no longer satisfied, and the Cartographers Guilds struck a Map of the Empire whose size was that of the Empire, and which coincided point for point with it. The following Generations, who were not so fond of the Study of Cartography as their Forebears had been, saw that that vast map was Useless, and not without some Pitilessness was it, that they delivered it up to the Inclemencies of Sun and Winters. In the Deserts of the West, still today, there are Tattered Ruins of that Map, inhabited by Animals and Beggars; in all the Land there is no other Relic of the Disciplines of Geography.

for those (like me) wondering where this wonderful passage is excerpted from:

https://en.wikipedia.org/wiki/On_Exactitude_in_Science

""On Exactitude in Science" or "On Rigor in Science" (the original Spanish-language title is "Del rigor en la ciencia") is a one-paragraph short story by Jorge Luis Borges, about the map-territory relation, written in the form of a literary forgery."

In addition to what everyone else has said: even if we did we wouldn't necessarily know what to look for. I have a friend who's PHD is trying to figure out what ancient dams look like in order to find sites of ancient agriculture. But that's the thing "what does a 500 year old dam look like" is a lifetime worth of study and is only just now possible.

"Just scan the fuck out of it and look for wierd stuff" is the dream but at the moment (as the other comments point out) we can't just scan everything but ALSO we don't know what's wierd.