Today, we go back many millennia in order to protect ourselves for the coming centuries: Emily Hart speaks to two Colombian scientists, Carlos Jaramillo and Camila Martínez, time-travellers of the smallest imaginable time machines: fossilised pollen and tree cells.

Climate change has been a constant feature of Planet Earth: at points in history, the planet has been both much cooler and much warmer than it is today - if we know which plants occupied an ecosystem the last time the Earth was a certain temperature or had a certain level of CO2 in the atmosphere, we can predict what our ecosystems will look like in the conditions that we will soon be living in.

Using tiny fossilised clues, Carlos and Camila are doing exactly this.

The climate change we are currently living through is unprecedented in speed – and water and rain cycles are a major concern for humanity's continued existence on the planet, so one focus of this work is the Amazon rainforest – both Colombia's slice of it and further afield.

Predictive models currently disagree about where the Amazon is headed as the earth warms – some models predict it will get wetter, others say it will become grasslands or scrub. One way to find out is to work out which plants lived in the area the last time conditions changed in the ways they are currently changing, and look at how that ecosystem and its inhabitants changed and adapted during that time. Drilling deep into the earth to find fossil records from 12 million years ago, Carlos is now studying the fingerprints left by Amazonian life from that time – particularly pollen. Camila is studying fossilised trees, whose cells – frozen in time – can show us how much water was in the environment. But pollen and other microscopic clues are in such abundance in places like Colombia that there simply isn't enough time in a human life to study and identify all of the species being found. Luckily, artificial intelligence is opening up huge possibilities – Carlos has been digitalising massive fossil collections and training AI to identify and catalogue samples. So today, we travel from the microscopic fingerprints of a distant ecological past resting in rocks and trees deep underground through to the futuristic methods made possible by new machine learning and digital processing. Carlos and Camila span multiple disciplines and vast timeframes, all in the hopes of getting us the information we need to survive the climate crisis which will change the face of the planet within our lifetimes. They'll be telling us how - and why it's so important. Support the podcast: www.patreon.com/colombiacalling