How do we know about climate before thermometers existed?

Reliable instrumental temperature records reach back only about 170 years, so for anything older scientists read proxies — natural archives that recorded climate as a side effect and preserved it. These include rock types, fossils and microfossils, pollen grains, oxygen-isotope ratios in deep-sea shells and rock, ice cores, tree rings, corals, cave speleothems, and Aboriginal art and oral traditions. Reading these accidental records is the science of palaeoclimatology.

How do ice cores tell us about past climate?

Ice cores are cylinders of layered ice drilled from ice sheets, reaching back about 800,000 years at the EPICA Dome C site in Antarctica. As snow turns to ice it traps tiny bubbles of the actual atmosphere of that year, so past carbon dioxide and methane can be measured directly. The oxygen and hydrogen isotopes (δ18O and deuterium) of the ice itself record past temperature — colder years leave ice that is lighter in δ18O — giving two independent records, temperature and atmosphere, from the same layers.

How do oxygen isotopes (δ18O) record past temperature?

δ18O is the ratio of heavy oxygen-18 to light oxygen-16, and it reads in opposite directions in different proxies. In the shells of marine foraminifera and in deep-sea sediment, colder, more glacial periods lock light oxygen-16 into land ice, so the ocean and its shells become enriched in oxygen-18 (a higher δ18O). In ice cores, colder conditions deliver snow that is more depleted in oxygen-18, so the ice has a lower δ18O. Both are used as temperature and ice-volume proxies once that direction is understood.

Reading the past climate

Q: What is a climate proxy?

A climate proxy is a natural object or material — such as a tree ring, an ice bubble, a fossil shell or a layer of sediment — that captured a climate signal as a side effect and preserved it, letting scientists reconstruct past climate where no instrument was ever taken. A useful proxy must be datable and must have a known relationship between its signal and climate (for example, a wider tree ring usually means a better growing year).

Reliable thermometer records run back only about 170 years — yet we know the temperature of an ice age 20,000 years ago, and of a world before the dinosaurs. We read it from proxies: ice bubbles, tree rings, coral bands, fossil shells and the chemistry of ancient water — natural archives that recorded the climate without ever meaning to. This powerbook climbs back through them, deepest and oldest to finest and most recent.

~800k yr

ice-core air record

~170 yr

direct thermometer record

short sections

Reading the past climateSection 1 of 5

0 of 5 locked in

This book has 5 short sections. Read each one, answer the check, then use Next to move on — the bar above tracks how far you've come.

Reading a record nobody meant to keep

We want to know the temperature of long ago — but the thermometer is a recent invention, and reliable, widespread temperature measurements reach back only about 170 years. Before that, nobody was recording the weather with instruments. And yet we describe, with confidence, ice ages tens of thousands of years ago and steamy worlds before the dinosaurs. How? Not by digging up old thermometers, but by finding things that recorded temperature as a side effect — and learning to read them.

Such a thing is called a proxy: a natural archive that captured a climate signal and preserved it. A tree ring grew wider in a good year. A coral laid down a band each year. Air was sealed into snow that hardened into ice. A sea creature built its shell from the water around it. Pollen blew into a lake and sank into the mud. None of these was trying to record the climate, but every one of them did. Reading these accidental records is the science of palaeoclimatology.

Two things make a proxy useful: you must be able to date it (how old is each layer?), and you must know how its signal maps to climate (does a wider ring mean warmer, or wetter?). The best proxies lay down one countable layer per year, so they can be dated by counting — and each archive reaches a different distance into the past. This chapter climbs back through them, from the deepest and oldest to the finest and most recent.

Schematic (not to scale): the finest archives are short, the oldest are coarse. Together they cover the whole story.

New words

proxy

A natural record (tree ring, ice bubble, fossil shell) that captured a climate signal as a side effect and preserved it.

palaeoclimatology/PAY-lee-oh-kly-muh-TOL-uh-jee/

The study of past climates, reconstructed from proxy records rather than from instruments.

🔒 Lock it in

What is a climate proxy?

Answer the check to continue

Quick reference & sources

The interactive book above is the lesson. What follows is a plain recap you can revise from, plus the sources and how this maps to the syllabus.

Proxies: the whole idea

A proxy is a natural archive that recorded climate as a side effect and kept it — a tree ring, an ice bubble, a fossil shell, a layer of pollen. Direct thermometer records (the instrumental record) reach back only about 170 years; for anything older we read proxies, the science of palaeoclimatology. A proxy is useful only if it can be dated and its signal maps cleanly to climate.

Deep-time evidence: rock, fossil and isotope

Rock types record climate: coal = warm wet swamp, evaporites (salt, gypsum) = hot arid, tillite and dropstones = cold and icy. Fossils, microfossils and pollen record the environments organisms needed. And the oxygen-isotope ratio (δ18O) in deep-sea foraminifera shells tracks ice ages over millions of years — in a glacial world, light oxygen-16 is locked into land ice, so the ocean and its shells gain heavy oxygen-18 (a higher δ18O). These marine records confirmed the Milankovitch cycles.

Ice cores, and the annual archives

Ice cores (EPICA Dome C: ~800,000 years) are the gold standard — trapped bubbles are real samples of ancient air, giving past CO2 and methane directly, while the ice’s δ18O and deuterium give temperature (colder = lighter δ18O, the opposite of marine shells). Closer to now, three incremental archives count the years: tree rings (dendrochronology), coral bands (Great Barrier Reef cores), and cave speleothems (stalagmites/stalactites, uranium-dated).

Human records, Aboriginal evidence, and the multiproxy picture

The instrumental record (since ~mid-1800s) is precise but short. Diverse Aboriginal and Torres Strait Islander communities hold rock-art sites and oral traditions recording real past environments — now-extinct megafauna and coastlines drowned by post-ice-age sea-level rise (up to ~120 m) — a legitimate record best understood on the communities’ terms. No single proxy is the whole truth; the firm conclusions come from a multiproxy picture where independent archives agree.

For HSC students

This powerbook covers the Evidence for Climate Variation thread of Module 7 (Climate Science), framed by the inquiry question what scientific evidence is there of climate variations in the past? Working through it, you should be able to:

describe ancient evidence of past temperature — pollen grains in sedimentary rocks, changes in rock types, fossils and microfossils, and changing isotope ratios in rocks and deep-sea sediments;
explain more recent evidence — ice cores (gas bubbles and oxygen isotopes), dendrochronology, Aboriginal art sites showing now-extinct species and environments, human instrumental records, and isotope ratios in stalagmites, stalactites and corals.

The previous chapter covered the natural causes of climate change and their timescales; the next chapter asks whether human activity has since varied the climate. Explanations here are original and are a study aid, not a copy of the syllabus.

Sources

NOAA Paleoclimatology and the British Antarctic Survey — proxy data, ice-core records and how proxies are read.
The EPICA (European Project for Ice Coring in Antarctica) Dome C ice core — the ~800,000-year record of temperature and greenhouse gases.
Marine oxygen-isotope (δ18O) stratigraphy of deep-sea foraminifera — the record that confirmed the orbital pacing of the ice ages.
Published research on dendrochronology, coral and speleothem isotope records, including Great Barrier Reef coral archives.
P. D. Nunn and N. J. Reid, on Aboriginal oral traditions recording post-glacial sea-level rise — presented as a research argument; specific identifications are noted as debated.
NSW NESA Earth and Environmental Science Stage 6 Syllabus (2017) — used only to scope the Module 7 content; explanations above are original.

Keep going

The thermometer is barely older than your great-great-grandparents — but the planet has been keeping its own records all along, in ice, stone, wood and living memory, for anyone who learns to read them.