The most valuable unbuilt-map on the planet.
What would an ocean map look like and do for the world? (Part 3 of 3)
I love maps. I think they’re one of the most elegant and effective infographic visual formats humans have ever created. The volume of information that can be effectively crammed into a single frame (or now 3D environment), while making multiple complex underlying datasets approachable, digestible, and valuable — is elegant, even beautiful.
Humans have been making maps for at least 5,000 years and since then, we’ve come up with millions of different types of maps to convey and make use of all sorts of information. For most of that time, it was important to visualize this data correctly because humans were always the end consumer of the information. And today, we usually only interact with a very select few types of these map products: road & highway (Google Maps), Imagery (Google Earth), topography, and perhaps weather.
However, in the past 10 years, something in the world has fundamentally changed.
Humans used to be the ONLY consumers of map products. But that is no longer the case. And in fact, the value behind many of the maps we use today (and even more so of the future), lies within the accessibility, usability, and general amount of underlying data a map was built off of.
Now, computers and software are the primary consumers of map data, not people.
This allows further insights and conclusions from the same underlying data — far beyond the visual aspects which humans require and are restricted too, to consume it.
Humans are now secondary. However, we are the primary consumers of the deeper insights locked within the data.
Why? Advancements in general compute & AI means we can extract new and unbountifully more from the exact same dataset previously used simply to create an image for us. It’s an unfortunate limit of us being creatures using solely vision as our consumption technique for map data and not something like an ethernet port— at least to our knowledge (#theMatrix).
What does this shift of consumer mean? It means that beyond making visual maps humans can easily interpret — it’s critical the underlying data can be easily read by a piece of software, anywhere in the world. It means APIs, and cloud-native systems become the cornerstone way to quickly distribute and enable everything to be consumable to any piece of software in the world.
Ocean Maps have not evolved with technology.
Google Earth’s oceans are about as good as it’s gotten, and I’ve already written about how terrible that actually is. It’s not because Google Earth isn’t an incredible piece of software (because it is), but because humans just don’t have a lot of the necessary data to make a great ocean map. Not surprisingly, humans have collected far less data on Earth’s oceans than we have of Earth’s land (for a variety of reasons) — ocean maps, and the technology behind them, have fallen severely behind. Google tried and failed with Google Oceans. Still, to date, there is no one company that dominates this space.
Why? Because the data behind our ocean maps is hard-to-collect, sparse, or non-existent, and it too hasn’t evolved at the pace of new technology. Currently, there is no easy way for a computer to access human’s current public collection of underlying raw ocean data.
How do we know that? We tried asking the government for it.
Specifically, we asked the National Center for Environmental Information (NCEI). It’s the current world repository for publically available ocean data and they said it will take A YEAR to get us a comprehensive set…..
A year. I asked for a download link or an API and documentation to do it ourselves, and they laughed. Non-existent. While they have map products (tilesets and images of the finished map products), the underlying data is not online and not accessible to computers.
We learned that the data is stored on magnetic strips inside a mountain in Colorado. One-by-one they have to convert the strips from analog to digital and then write the data onto an external hard drive, which they then ship to us over FedEx. Given the volume of data, and the speed of their converter, the minimum time to complete a full raw data transfer is ~1 year. I was stunned.
We started receiving hard drives having no clue what state the data was in. Worse news, the data is not stitched together geographically (though, that is a broader initiative within NOAA at the moment) rather organized by survey mission. It’s a mess.
What this means is even for the 15% of the ocean we do have useable data on, it’s never actually been looked at all at one time — in its raw format — by a piece of software or a computer.
An analogy: do professional photographers do their editing on a processed jpeg image? NO. They do it in RAW format because if not, you immediately limit the potential of the photo if you start in an already processed format.
Seafloor data shouldn’t be treated any different in today’s world. But it is because, without a cloud system, you couldn’t possibly manage and effectively build software over this enormous raw dataset in multiple locations.
Almost everyone in the ocean data world works exclusively with post-processed data.
Taking the implications of this a step further — We’re proposing there’s probably information locked within the raw underlying data that we actually already have, but don’t know about because we can’t work across and leverage the whole dataset.
Ocean maps and data are not yet ready for computers and software to be the primary consumers.
What could an ocean map look like?
If you believe in pushing the bounds of human ambition and capability — which is exactly what we’re out to do —I’ve put together a list of what we’d be trying to make available to every data scientist, software engineer, oceanographer, commercial enterprise, government, and military interested. Enjoy:
1. Bathymetric maps
It is the bedrock of the map. Think of this as topography for the seafloor. We need high-resolution (<10m) bathymetric maps to begin understanding what may actually be there. Whether it’s a seamount with minerals around it, a ship, or hell — Atlantis. Without high-resolution data, we simply won’t know it’s there.
1.1 LIDAR maps
This is more for detailed scanning of objects in an area. Because the range of LIDAR underwater is only about 30–45m depending on the clarity of the water, this is not good for subsea mapping. However, it’s great for getting extreme detail once you know something is there.
This is used for industrial equipment and infrastructure inspections, archeological scanning, and ultra-high resolution localized seafloor mapping — which is like bathymetry.
This is used to identify objects and see smaller features beyond broad geophysical geometry. Not surprisingly, this is useful for verifying what an object we believe to be there actually is. It also helps enrich bathymetric data with visual confirmation. It’s generally collected with cameras (< 30m away) or by side-scan sonar. More on sonar imaging here.
This is used ubiquitously across the subsea landscape because we almost always don’t know what’s there.
This defines what the substrate on the seafloor is. Is it sand, mud, wood, metal, rock, or coral? This data will tell you.
This information is critical for any engineering projects involving the ocean floor. It’s also critical to understanding the geological environment necessary for finding certain fisheries and minerals.
4. Seafloor geomorphic feature identify & name mapping
Each area of the seafloor can be defined by a geomorphic name: (abyssal plain, shelf, seamount, etc) — think back to elementary geology class. Moreover, each discovered feature has a unique name, like the Rockies vs the Alps! Not surprisingly, there are many more yet to be discovered and still named. Find a new seamount? You can submit a name here. I mean… how often do you get to name a mountain range you found on Earth?! The only caveat, it would be underwater.
This information is critical for transatlantic subsea cables and pipeline routing, as well as the starting point for identifying areas of mineral value. It’s also ultra-critical for predicting subsea landslides (yep — they happen) which create freak tsunamis.
If we do a geospatial analysis with the extremely low-resolution (0.5–5km) satellite altimetry data we already have we get this:
We’ve already mapped the different features of Mars at cm resolution — but this is easier because we can see them all clearly from Earth’s outer orbit.
However, the ramifications of doing this in higher resolution across the whole seafloor would uncover completely new features we’ve yet to find. At the bottom of the ocean, these features would be significantly sized comparatively to land features (hills & valleys) that most of us would think of. Exposing more granular information here gives high-level clues to geological hotspots for minerals and other items of value. You’d also be able to significantly improve subsea infrastructure routing with more detailed information here — dramatically reducing the cost of the process now because you’d be starting with a better baseline of information.
5. Magnetic anomaly mapping
Things get really interesting when you can begin to map out where magnetic anomalies lie not just on the seafloor, but below it as well. This begins to paint a picture of areas of high metallurgic content, but also the total volume of a specific deposit. Sometimes things we’ve lost in the ocean can get buried underneath the sand — it happens. Magnetics can show us if there are structures hiding in the sand.
Previously this data was hard to collect, but there are new companies pioneering new, cheaper, 3D sensing magnetometers, that allow this to be done cheaply now.
6. Sub-Bottom Profiling (SBP) mapping
This is like X-Ray for the seafloor. It shows the layers of sediment and rock that lay below the seafloor. This is useful for oil and gas prospecting as well as getting another lens of data on what may lie below the surface.
7. Biomass density mapping
This is a specific type of water column mapping. It’s a rough measurement of the volume and density of sea life in a certain part of the ocean at a certain time.
This is a depiction of biomass indexing where you’re looking to get a sample of groups of sealife.
This data can be used to build biomass density maps:
8. Gas seep mapping
Also a type of water column mapping. This is used to find oil and gas reserves, as well as understand the health of already dug wells. In an environmental context, it helps identify what chemicals could be in the local ocean area which affects the type and health of the sea life.
9. Conductivity, Temperature, and Density (CTD) mapping
A specific sensor that measures the three values in the immediate area of the sensor. By processing CTD up and down the water column, you can get salinity, density, and sound velocity of the surrounding water. This is critical for accurate sonar post-processing, as well as understanding what is going on in the local ecosystem.
10. Plastic density mapping
Also created by processing the noise in the MBES data. As we continue to use the ocean as our dumping ground — there is value in understanding just how bad we’re actually screwing things up.
Where big markets are hiding.
Everything talked about above represents single datasets.
However, it’s when you combine them that they become particularly useful to some of the largest industries in the world: mining, oil & gas, fishing & fisheries, telecom, marine archeologists (treasure hunters or treasure “preservers” as we’ve heard it called), and government/military intelligence.
Finding correlations and broadening the total dataset unlocks a whole new level of potential from the same datasets collected from only a handful of sensors.
1. Nautical Charts
Humans have viewed “ocean data” like this for quite some time. Most of the data in nautical charts were collected prior to modern sensors and dates back before the 1950’s. People are desperately trying to update this with the most accurate representation of what’s there.
While the safety of navigation at sea is important, it just scratches the surface of what can be useful within subsea data.
2. Marine Mineral Resource Mapping
While a budding industry, seafloor mining will almost inevitably be the way we sustain human existence on this planet at the population numbers we’re predicting in the next 20+ years. Subsea mining is expected to grow 37% CAGR over the next 10 years due to rising costs to extract minerals on land.
It certainly will be part of most country’s political strategies to reduce their reliance on critical foreign metals needed to sustain their populations. Surprisingly (or maybe not) the US has no clue what minerals lie just offshore of our own waters like the EU does (see below). This was exhibited by the recent presidential decree from the Trump administration.
This is the US’s marine mineral portal for comparison — notice how empty it is…
3. Oil & Gas site probability mapping
We rely on plastic — for the worst — but alas, until we truly replace this material we will need petroleum. The goal is to identify where there are natural geophysical traps and conditions that have allowed a certain chemical to have formed/collected. While there are several methods for prospecting these different chemicals, everyone is looking to identify the lithology, quality of the reservoir, and the actual chemical within. Being able to do any of these prior to sending ships out to sea is an ENORMOUS efficiency gain.
4. Habitat maps
Anyone in fisheries, aquaculture, or ocean engineering knows that to get permits to work in certain areas, you need to provide habitat maps. Beyond the commercial value in being able to generate them quickly, governments are beginning to realize we need to preserve their fisheries to continue feeding their people.
We see this as an enormous growth area. We have a burning belief that we must make sure the ocean is used responsibly. As humans inevitably do more business in the ocean, it’s essential that we monitor the effects of what we are doing and STOP if we see irreparable damage occurring. The only way to do that is by having a 3rd party habitat and environmental monitoring services.
5. Item location/hazard location mapping
Humans have intentionally put quite a lot in the ocean. But we’ve also lost a lot in the ocean. To find items of archeological or commercial importance we need to use multiple different datasets to build probability maps of where we think certain objects may be. Multiple datasets are also required to verify that an object is what you think it is.
There’s obvious value in knowing where all these assets actually are — whether intentionally or unintentionally put there.
There are currently ~3,000,000 estimated missing shipwrecks out in the ocean.
Currently, humans have only found 1% of the world’s documented lost ships. And if you look at the currently documented shipwrecks just by the US, it’s a bit staggering:
And there’s much, much more out there.
6. Unexploded Ordnance (UXO) Mapping
In WWII and many subsequent years throughout the cold war, planes could not return and land with armed bombs so, before landing their payloads are jettisoned into the ocean. Conquer a nation and the winners need to dispose of their enemy’s weapons? Throw them into the ocean! This is exactly what happened.
Anytime you want to do anything that involves the seafloor (dredging, drilling, laying pipes, etc.) you want to be sure you’re not going to run into any explosives. This is when UXO maps become critical. To see the process of what this is like, check this out.
7. Military object mapping/detection/monitoring (military assets — classified not for public)
There are plenty of secrets in the deep. Navies are constantly trying to figure out the ones from other countries. Persistent, reliable, low-vis monitoring and mapping is a critical component of subsea naval strategies and missions.
8. Mapping what we don’t know exists.
The unknown will almost undoubtedly show us things we don’t know yet. 60% of our PLANET (85% of the ocean) hasn’t even been looked at with modern equipment. We’re excited to see what we find. We’re excited to see what human creativity and ingenuity will create.
So we’re going to build it.
All this represents one of the largest uncollected geospatial datasets known to mankind. In 2019, how often is that the case?
Hell — we’ve mapped the universe — THE UNIVERSE. But we have yet to collect this valuable information from our own ocean. It’s astonishing.
Right now, bits and pieces of this information sit siloed in large corporations and governments. While these entities will support early movers, they have also limited growth. The last thing we want is this enormously valuable and Earth-shifting dataset to only be used for political and commercial purposes.
So, we believe opening access to the troves of subsea information we can collect could lead to completely new ways to utilize our planet for the continued survival of the human race, growth & evolution of businesses, and rapid advancement of ocean science & understanding. We think this is more than necessary, it’s imperative.
We intend to build tools that make accessing this vast information simple and affordable.
We’re particularly excited to allow computers and software to easily consume, use, and extract further meaning from underlying subsea datasets.
Whether you’re a lone software engineer, data scientist or geologists in a large organization, oceanographer at a non-profit, or the largest government in the world — we hope to make this hidden and non-existent data readily available to you. Reach out for beta access.
While our exact plan is still a secret, we certainly don’t want the problem to be. You now know.
We’re currently organizing a group of some of the best people from the software, robotics, oceanographic, government, and finance world — and hopefully many more that we have yet to meet — to bring this data and map into existence.
Bedrock — over and out.
This is Part 3 of 3 in a series written to explain the ocean mapping problem better. Hopefully, if you’ve read the others and gotten here, you’re a little more aware than when you started.
If you’re starting with part 3, no worries — Part 1 of 3 dug into what our current map of the ocean is actually made up of and how little we know about it.
Our map of the ocean explained.
Isn’t the ocean already mapped? Nope. 85% nope. (Part 1 of 3)
Part 2 of 3 dug into why the systems we use collect information about the ocean floor have held us back from truly collecting a meaningful amount of information here.
It’s time we mapped the ocean. Here’s how.
With today’s mapping systems, it will take us 970 ship years. We need other options. (Part 2 of 3)
If you found this interesting, please give it a clap or four and please, please please share this with people. If you want this in your inbox, feel free to sign up here.