Not before time, the world's population is focusing its attention on threats arising from humankind's impacts on its planetary habitat. But underlying whatever abilities we have to mitigate the impacts of habitat change lies the essential ability to understand our planet's structure and dynamics at all levels and on all time-scales.

The slow but steady progress in understanding Earth is the momentous achievement of the geosciences since their foundations were laid well over a century ago. Nature Geoscience is Nature's newest monthly sister journal, and is intended to capture the best of those sciences and serve all of the interested research communities (see http://www.nature.com/ngeo). As with all Nature research journals, this launch in no way dilutes Nature's own commitment to these disciplines.

Launching in January 2008, the journal can be expected to track important research currents. For example, an expanding branch of climate science is the investigation of past analogues for current change, such as rapid sea-level rise, warm periods, ocean acidification or such crucial links as the coupling between atmospheric carbon dioxide concentrations and global temperatures over geological time-scales. Traditionally rooted in geology, palaeoclimate researchers are now interacting more intensively with modellers and with investigators of modern climate.

Solid-Earth scientists are still exploring the implications of the 2004 discovery of the high-pressure phase of the mineral MgSiO3, the post-perovskite that dominates the lowermost mantle, for our understanding of the structure of this region. This discovery affects our understanding of the temperature in the lower mantle and mantle convection, as well as iron and heat exchange between Earth's core and the mantle. All these, in turn, have implications for our reconstructions of Earth's history, the evolution of its core and the geodynamo.

Like all the Nature research journals, Nature Geoscience will also have its informal aspects. The 'Backstory' section, about the hard work that comes before a publishable research paper, will describe, for example, what it takes to reconstruct 15 million years of Arctic ocean circulation, to map the floor of the Arabian Sea or to study an Alaskan glacier. One of these articles will be published on the final page of each printed issue, with additional articles published online each month.

The coming year sees the start of a period of international programmes focusing on our planet, not least the International Year of Planet Earth. The new journal could not arrive at a more auspicious moment.