Called AlphaEarth Foundations, the geospatial AI model pulls together satellite images and other environmental data to create a single picture of the planet, according to a blog post.

Satellites orbiting Earth collect large amounts of information every day. While this data is valuable, it’s often stored in many different formats and comes from different times and sensors, making it hard to combine.

AlphaEarth Foundations acts like a “virtual satellite” that can merge all this information into one consistent view, the post said.

For example, the model can see through persistent clouds in Ecuador to map agricultural plots in various stages of development, according to the post. It can also map the surface of Antarctica, normally a tough place to image.

The AI model can spot changes in land use, such as new construction, deforestation or crop growth, in 10-meter squares, per the post. It also stores the information much more efficiently, using 16 times less space than comparable AI systems.

To make it available to the public, Google is releasing yearly snapshots from 2017 through 2024 in a new Satellite Embedding dataset within Google Earth Engine. This dataset contains more than 1.4 trillion data points each year and is ready to use without extra processing work, the post said.

Over 50 organizations have already tested the system, including the United Nations’ Food and Agriculture Organization, Stanford University, Oregon State University and others, according to the post.

For example, the Global Ecosystems Atlas, an initiative that aims to comprehensively map and monitor the world’s ecosystems, is using AlphaEarth Foundations to help countries classify unmapped ecosystems into categories such as coastal shrublands, per the post. In Brazil, environmental mapping group MapBiomas is using the tool to track farmland and forest changes.

However, even though the model is a “cutting-edge technological breakthrough” in Earth mapping, it is dependent on high-quality satellite data, according to a GoGeomatics Canada blog post.

“While it is known for effectively filling gaps in missing or incomplete data, interpreting poor-quality inputs in critical situations can lead to misdirection,” the post said.

Read also: T-Mobile and SpaceX to Test Satellite-to-Cell Service

How AlphaEarth Foundations Works

AlphaEarth Foundations combines information from many different satellite and environmental sources into one clear, consistent picture of Earth. It works a bit like stitching together thousands of puzzle pieces into a single image, except the puzzle pieces come from different satellites, sensors and even time periods.

The system takes in a variety of public data, including the following:

• Optical satellite photos, like those available on Google Earth
• Radar scans that can penetrate cloud cover
• 3D laser mapping
• Climate and environmental readings, such as temperature and rainfall
• Elevation maps and gravity measurements
• Descriptive information linked to locations

It treats the images from the same location over time like frames in a video. This lets it “understand” changes through the seasons or from one year to the next, like crops being planted and harvested, forests being cleared or cities expanding, the Deepmind blog post said.

AlphaEarth Foundations then condenses all this into what Google called a “64-dimensional representation” for each 10-meter square, whether land or coastal water.

Consider that 3D only provides latitude, longitude and elevation; 64 dimensions provide richer detail, not just location but also appearance, environment and behavior over time.

“What is interesting is that they’re able to get down to 10-by-10 meter squares, which is phenomenal,” Christopher Seeger, professor and extension specialist of landscape architecture and geospatial technology at Iowa State University, told The Register. “It’s going to be great for decision makers.”

“This breakthrough enables scientists to do something that was impossible until now: create detailed, consistent maps of our world, on demand,” the DeepMind blog post said. “…[T]hey no longer have to rely on a single satellite passing overhead. They now have a new kind of foundation for geospatial data.”

The system can be used for many purposes, including monitoring wildfires, tracking water levels in reservoirs and spotting urban growth. It can also help create detailed maps with fewer samples, saving time and resources.

Google plans to explore combining AlphaEarth Foundations with its Gemini multimodal model to expand the system’s capabilities further.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

Read more:

Amazon Plans to Offer Satellite Internet Service in Late 2025

Amazon’s Space Lasers Redefine Global Data Connectivity

The Cyber Hack From Space

“It’s more exhausting,” says Afer, a deputy commander of the “Da Vinci Wolves”, describing how one of the best-known battalions in Ukraine has to defend against constant Russian attacks. Where once the invaders might have tried small group assaults with armoured vehicles, now the tactic is to try and sneak through on foot one by one, evading frontline Ukrainian drones, and find somewhere to hide.

Under what little cover remains, survivors then try to gather a group of 10 or so and attack Ukrainian positions. It is costly – “in the last 24 hours we killed 11,” Afer says – but the assaults that previously might have happened once or twice a day are now relentless. To the Da Vinci commander it seems that the Russians are terrified of their own officers, which is why they follow near suicidal orders.

Reconnaissance drones monitor a burnt-out tree line west of Pokrovsk; the images come through to Da Vinci’s command centre at one end of a 130-metre-long underground bunker. “It’s very dangerous to have even a small break on watching,” Afer says, and the team works round the clock. The bunker, built in four or five weeks, contains multiple rooms, including a barracks for sleep. Another is an army mess with children’s drawings, reminders of family. The menu for the week is listed on the wall.

It is three and a half years into the Ukraine war and Donald Trump’s August peace initiative has made no progress. Meanwhile the conflict evolves. Afer explains that such is the development of FPV (first person view) drones, remotely piloted using an onboard camera, that the so-called kill zone now extends “12 to 14 kilometres” behind the front – the range at which a $500 drone, flying at up to 60mph, can strike. It means, Afer adds, that “all the logistics [food, ammunition and medical supplies] we are doing is either on foot or with the help of ground drones”.

Further in the rear, at a rural dacha now used by Da Vinci’s soldiers, several types of ground drones are parked. The idea has moved rapidly from concept to trial to reality. They include remotely controlled machine guns, and flat bed robot vehicles. One, the $12,000 Termit, has tracks for rough terrain and can carry 300kg over 12 miles with a top speed of 7 miles an hour.

Land drones save lives too. “Last night we evacuated a wounded man with two broken legs and a hole in his chest,” Afer continues. The whole process took “almost 20 hours” and involved two soldiers lifting the wounded man more than a mile to a land drone, which was able to cart the victim to a safe village. The soldier survived.

While Da Vinci reports its position is stable, endless Russian attempts at infiltration have been effective at revealing where the line is thinly held or poorly coordinated between neighbouring units. Russian troops last month penetrated Ukraine’s lines north-east of Pokrovsk near Dobropillya by as much as 12 miles – a dangerous moment in a critical sector, just ahead of Trump’s summit with Vladimir Putin in Alaska.

At first it was said a few dozen had broke through, but the final tally appears to have been much greater. Ukrainian military sources estimate that 2,000 Russians got through and that 1,100 infiltrators were killed in a fightback led by the 14th Chervona Kalyna brigade from Ukraine’s newly created Azov Corps – a rare setback for an otherwise slow but remorseless Russian advance.

That evening at another dacha used by Da Vinci, people linger in the yard while moths target the light bulbs. Inside, a specialist drone jammer sits on a gaming chair surrounded by seven screens arranged in a fan and supported by some complex carpentry.

It is too sensitive to photograph, but the team leader Oleksandr, whose call sign is Shoni, describes the jammer’s task. Both sides can intercept each other’s feeds from FPV drones and three screens are dedicated to capturing footage that can then help to locate them. Once discovered, the operator’s task is to find the radio frequency the drone is using and immobilise it with jammers hidden in the ground (unless, that is, they are fibre optic drones that use a fixed cable up to 12 miles long instead of a radio connection).

“We are jamming around 70%,” Shoni says, though he acknowledges that the Russians achieve a similar success rate. In their sector, this amounts to 30 to 35 enemy drones a day. At times, the proportion downed is higher. “During the last month, we closed the sky. We intercepted their pilots saying on the radio they could not fly,” he continues, but that changed after Russian artillery destroyed jamming gear on the ground. The battle, Shoni observes, ebbs and flows: “It is a war of drones now and there is a shield and there is a sword. We are the shield.”

A single drone pilot can operate 20 missions in 24 hours says Sean, who flies FPVs for Da Vinci, for several days at a stretch in a crew of two or three, hidden a few miles behind the frontline. Because the Russians are on the attack the main target is their infantry. Sean frankly acknowledges he is “killing at least three Russian soldiers” during that time, in the deadly struggle between ground and air. Does it make it easier to kill the enemy, from a distance? “How can we tell, we only know this,” says Dubok, another FPV pilot, sitting alongside Sean.

Other anti-drone defences are more sophisticated. Ukraine’s third brigade holds the northern Kharkiv sector, east of the Oskil River, but to the west are longer-range defence positions. Inside, a team member watches over a radar, mostly looking for signs of Russian Supercam, Orlan and Zala reconnaissance drones. If they see a target, two dash out into fields ripe with sunflowers to launch an Arbalet interceptor: a small delta wing drone made of a black polystyrene, which costs $500 and can be held in one hand.

The Arbalet’s top speed is a remarkable 110 miles an hour, though its battery life is a shortish 40 minutes. It is flown by a pilot hidden in the bunker via its camera using a sensitive hobbyists’ controller. The aim is to get it close enough to explode the grenade it carries and destroy the Russian drone. Buhan, one of the pilots, says “it is easier to learn how to fly it if you have never flown an FPV drone”.

It is an unusually wet and cloudy August day, which means a rare break from drone activity as the Russians will not be flying in the challenging conditions. The crew don’t want to launch the Arbalet in case they lose it, so there is time to talk. Buhan says he was a trading manager before the war, while Daos worked in investments. “I would have had a completely different life if it had not been for the war,” Daos continues, “but we all need to gather to fight to be free.”

So do the pilots feel motivated to carry on fighting when there appears to be no end? The two men look in my direction, and nod with a resolution not expressed in words.