The Rio+20 conference (20–22 June 2012) will be held in Rio de Janeiro, Brazil. It will focus on two main themes — development of a green economy while fostering sustainable development and poverty eradication, and the institutional frameworks necessary to achieve these goals, including the strengthening of international environmental governance. Areas identified for 'priority attention' are employment, energy, sustainable cities, food security, and sustainable agriculture, water, oceans and disaster readiness.

Brazil — the fourth biggest greenhouse-gas emitter in the world — has so far made limited progress in limiting emissions, and, as discussed by Eduardo Fernandez Silva (page 379), the measures so far announced under the National Policy on Climate Change Law (enacted in December 2009) are unlikely to achieve much more. The intention of the law is to set in place measures to reduce emissions. The authorities hope that by 2020, emissions will have fallen to around 2005 levels. Many question whether this target will be achievable in the absence of attractive economic incentives for environmentally responsible behaviour along with strong sanctions against misbehaviours.

In essence, Brazil needs to move beyond good intentions and start to implement practical measures. To be fair, the nation has started to reduce deforestation in the Amazon and limit damaging land-use change through sustainable management. However, as discussed by Silva, what has so far been announced in relation to other sectors such as industry look less promising.

Apart from obvious greenhouse-gas sources such as heavy industry, considerable amounts of carbon dioxide and methane are released from tropical dams. Much of the electricity consumed in Brazil comes from hydroelectric power plants. And yet, as discussed by Philip Fearnside and Salvador Pueyo (page 382), emissions from tropical dams rarely feature prominently in national or global greenhouse-gas inventories.

Even when greenhouse-gas emissions from tropical dams are considered, they are often underestimated or misreported. There is no excuse for this — reliable methods exist for measuring emissions from reservoirs, as well as from turbine outlets and downstream river flows, and for up-scaling estimates. It does not help when, as documented by Fearnside and Pueyo, a major Brazilian electric utilities company gets its sums wrong, thereby seriously underestimating total reservoir surface emissions from Brazil's largest dams. But even setting aside such sloppiness — and indeed the loss of river and forest habitat caused by dam construction — tropical hydroelectric plants are not quite as green as often portrayed, and certainly no panacea for dealing with the problem of emissions from the energy sector.

This leads on to a more general point. It is often naively assumed that energy from fossil fuels can be replaced in a one-to-one manner with energy from renewable sources — that is that one unit of renewable energy displacing one unit of fossil-fuel energy. Indeed, this is an implicit assumption of reports prepared by the Intergovernmental Panel on Climate Change. However, as reported by Richard York (page 441), the assumed relationship has not held for most nations of the world over the past 50 years.

York finds that, on average, less than one-quarter of a unit of fossil-fuel energy use is displaced by each unit of total national energy use from non-fossil-fuel sources. Focusing solely on electricity, the situation is considerably worse. As Andrew Jorgenson puts it in an accompanying News & Views (page 398), “York's findings contradict the widely held assumption that the expansion of alternative energy production will proportionally suppress fossil-fuel energy production.” Jorgenson goes on to outline both the implications and limitations of the study, but echoes the call of many that the best way of reducing emissions is for all of us, individually and collectively as societies, to consume less energy.

As already mentioned, one of the areas highlighted for attention by Rio+20 is disaster readiness. On page 462, Ning Lin and colleagues consider the likelihood of increased hurricane threat under climate change, taking as a case example New York City, which is increasingly vulnerable to flooding under storm-surge conditions. Using computer simulations, the researchers show how the combined effects of changes in storm climatology and sea-level rise could greatly increase the frequency of surge flooding over the next century. Their findings suggest that now is the time to initiate long-term adaption planning so as to avert the worst impacts on property, infrastructure and the safety of the city's citizens — a theme expanded on by Jeroen Aerts and Wouter Botzen on page 377.

Meanwhile, as discussed by Allan Findlay on page 401, a recently published study (C. L. Gray and V. Mueller Proc. Natl Acad. Sci. USA 109, 6000–6005; 2012) has shown that climate change-related flooding in a rather different context — that of rural Bangladesh — may be relatively unimportant compared with non-climate-related crop failure in driving human migration. So, could it be that warnings from environmentalists that climate change will inevitably lead to forced mass human migrations are mere scaremongering?

That is probably going too far, but as Findlay notes, even if migration is a practical option, the decision to 'up sticks' and move is likely to involve multiple, complex and intertwined factors. This view is given further credence, by a study conducted by Dominic Kniveton and colleagues in the landlocked west-African country Burkina Faso (page 444). They find that owing to complex interactions between climate, rainfall variability and various socio-economic factors, migration could increase or decrease over the next few decades, depending on the scenario considered — issues further discussed in an accompanying News & Views by Etienne Piguet (page 400).

Moving to agriculture, there is now a not-so-small industry of researchers looking to predict the possible impacts of climate change on regional, national and global scales. Much of this work is intended to inform policy. However, a dispute about projected future wheat yield trends in the United Kingdom (pages 378 and 380) highlights, at the very least, the critical importance of clarity in communicating climate change evidence. We trust that whatever comes out of Rio+20 will be reported in a clear and unambiguous manner, reducing the potential for unnecessary misunderstanding and rancour.