What are the giant blobs at the base of the mantle, and do plumes rise from them?

What we know. Seismic waves slow down in two huge regions sitting on the core, one beneath Africa and one beneath the Pacific. They are called LLSVPs (large low-shear-velocity provinces). Each is thousands of kilometres across and rises up to about 1,000 km above the core. Many hotspot volcanoes, like Hawaii and Iceland, sit above broad, slow columns in the mantle that look like mantle plumes: hot rock rising from deep down.

What we don't. Whether the blobs are just hotter than their surroundings or chemically different too; how old they are; whether they are piles of ancient sunken ocean floor or even, as one recent idea suggests, remnants of the planet-sized body that hit Earth and formed the Moon. Whether every hotspot is fed by a plume from such depths is also debated.

Why it matters. Several of the largest lava floods in Earth's history, possibly including the Deccan Traps, seem to have erupted above the edges of these blobs. They may steer how heat leaves the core.

How old is the inner core, and why does its spin vary?

What we know. The solid iron inner core is growing slowly as the liquid outer core cools and freezes onto it. That freezing helps drive the churning that makes Earth's magnetic field.

What we don't. When it began to freeze. Estimates range widely, most often from about 0.5 to 1.5 billion years ago, which would make it young compared with the planet. Studies of repeated earthquakes from the same places suggest the inner core rotates slightly faster than the mantle at some times and slower at others, with a change around 2010. How to interpret those signals is hotly debated.

Why it matters. The magnetic field shields the surface from charged particles from the Sun. Its history is tied to the core's.

How much water and carbon are hidden in the mantle?

What we know. Between about 410 and 660 km down is the mantle transition zone, where pressure squeezes olivine into denser minerals such as ringwoodite, which can hold water in its crystal structure. In 2014 a tiny ringwoodite crystal trapped in a Brazilian diamond was found to contain about 1.5% water. Most of Earth's carbon is also stored deep inside the planet, not at the surface.

What we don't. How much water the transition zone holds overall; estimates range from a fraction of the oceans' volume to several times it. Whether subduction carries more carbon down into the mantle than volcanoes breathe back out is also unresolved.

Why it matters. Deep water and carbon cycles over billions of years help set how much ocean and atmosphere a planet keeps.

When did plate tectonics start, and why only here?

What we know. Plate tectonics operates today, and rocks show it has done so for at least the last billion years, and probably much longer. None of the other rocky planets has it. Venus, almost Earth's twin in size, has a single unbroken outer shell, a "stagnant lid".

What we don't. When it began. Published estimates range from more than 4 billion to under 1 billion years ago; many geologists favour roughly 3 billion years, possibly as a gradual switch-on. Nor do we know for sure why Earth has it. Leading ideas involve liquid water weakening rock, and the planet's size and internal heat.

Why it matters. Plate tectonics recycles carbon, builds continents and concentrates ore. It may be one reason Earth stayed habitable.

Why wasn't the young Earth frozen, and where did life begin?

What we know. Stars brighten as they age. Early in Earth's history the Sun gave out only about 70% of today's light, enough to freeze the oceans with today's atmosphere. Yet tiny ancient zircon crystals suggest liquid water by about 4.4 billion years ago, and widely accepted fossil evidence of life, such as layered microbial mats called stromatolites, dates to about 3.5 billion years ago.

What we don't. How the planet stayed warm, the faint young Sun paradox. More CO₂ or methane, fewer clouds and a darker surface are all candidates. Where life started, perhaps at hot springs on the sea floor or in pools on land, and older claimed signs of life, remain contested.

Why it matters. It is the original test of the weathering thermostat, and it tells us where to look for life on other worlds.

Can we predict earthquakes?

What we know. We know where big earthquakes are likely: along plate boundaries and major faults (see Earthquakes & volcanoes). Seismologists make forecasts, probabilities like "a 30% chance of a large quake in this region over 30 years". Early-warning systems can give seconds to tens of seconds of warning once a quake has started.

What we don't. How to make a prediction: a specific place, size and date, in advance. No reliable warning sign has been found, and many scientists doubt one exists. Parts of the Himalaya have not had a great earthquake for centuries and are thought to be storing strain, but when it will be released is unknown.

Why it matters. Hundreds of millions of people live in the Himalayan foothills and Indo-Gangetic plains.

How fast are we draining our aquifers?

What we know. An aquifer is rock or sediment that holds groundwater. India pumps more groundwater than any other country, roughly a quarter of the world's total. Since 2002 the twin GRACE satellites, and their successors, have measured tiny changes in Earth's gravity caused by water moving. They showed that north-west India (Punjab, Haryana, Rajasthan and Delhi) was losing groundwater at roughly 4 cm of water depth a year in the 2000s.

What we don't. How much usable water is left, how fast aquifers recharge, and how depletion will respond to changing monsoons, crops and policy. Gravity satellites see large regions, not individual farms or wells.

Why it matters. Groundwater irrigates much of India's food. It also links to soil chemistry and to how dissolved carbon moves underground.

Is there useful hydrogen underground?

What we know. Rocks make hydrogen gas naturally. The main way is serpentinisation, in which water reacts with iron-bearing minerals such as olivine, which is abundant in ultramafic rocks like peridotite (and present in some rocks used for enhanced weathering). A well in the village of Bourakebougou in Mali has produced nearly pure hydrogen since 2012, and companies are now exploring for natural (or "white") hydrogen elsewhere.

What we don't. How much gets trapped in reachable reservoirs rather than leaking away or being eaten by microbes, and whether wells can flow for decades. Some modelling suggests the total underground is enormous, but most is likely too deep, too scattered or too far offshore to recover.

Why it matters. If it can be found and produced cheaply, it could be a low-carbon fuel and industrial feedstock.

Can we mine enough critical minerals, fast enough?

What we know. Batteries, wind turbines, motors and power lines need copper, nickel, lithium, cobalt and rare earth elements in quantities far beyond today's. Copper ore grades have been falling for decades, and it often takes more than 15 years from a discovery to a working mine. Processing is concentrated in a few countries; China refines around 90% of rare earths. See What makes an ore deposit.

What we don't. Where the next generation of large deposits is, especially those buried under cover, and how fast they can be found, permitted and mined responsibly. Recycling and new chemistries could change demand in ways that are hard to forecast.

Why it matters. The energy transition depends on rocks we haven't found yet.

How much will the planet warm, and what do past climates say?

What we know. Climate sensitivity is how much the world eventually warms if atmospheric CO₂ doubles. The IPCC's 2021 assessment put it likely between 2.5 and 4 °C, best estimate 3 °C. The rocks record natural experiments. About 56 million years ago, during the Palaeocene–Eocene Thermal Maximum (PETM), thousands of billions of tonnes of carbon entered the atmosphere and ocean over a few thousand years, and the world warmed by roughly 5–8 °C.

What we don't. The exact source of the PETM carbon and how fast it was released, and how clouds and ice will behave in a warmer world, which is why the sensitivity range is still wide.

Why it matters. The PETM took something like 100,000–200,000 years to recover, much of that through silicate weathering. Today we are adding carbon far faster than during the PETM.

Can we remove carbon at gigatonne scale, and prove it?

What we know. Most scenarios that hold warming to 1.5–2 °C need billions of tonnes of CO₂ removed from the air every year by mid-century, on top of deep emission cuts. Enhanced rock weathering speeds up a natural process we understand well in the lab, and model studies suggest farmland worldwide could remove on the order of a billion tonnes of CO₂ a year.

What we don't. How fast crushed rock actually weathers in real, messy soils, which is often much slower than in lab beakers; how much of the captured carbon is delayed or lost on its way to the sea; and how to measure all this cheaply and robustly.

Why it matters. Carbon removal is only worth paying for if it is real, permanent and measured honestly.

Introductory geology

Earth: Portrait of a Planet by Stephen Marshak, and Understanding Earth by John Grotzinger and Thomas Jordan. Both are written for first-year university students with no background, and are beautifully illustrated. Older editions are cheap and still excellent.

Courses and geological surveys

The Open University's free OpenLearn geology courses. The websites of the Geological Survey of India, the British Geological Survey and the US Geological Survey, which publish maps, explainers and data on everything from earthquakes to minerals.

Geology as literature

Annals of the Former World by John McPhee, a Pulitzer Prize-winning journey across North America with geologists, in which McPhee coined the phrase "deep time". Read a chapter at a time.

Look at rocks

The best way to learn is to look. Stop at road cuttings and riverbanks, pick up pebbles and ask where they came from, and use the rules from Stratigraphy to read a cliff. Ask the geologists at Alt Earth what they are working on.