This document presents a collection of selected papers produced for and discussed at the Second International Conference on Early Warning, in Bonn, Germany, in 2003, and four regional conferences in Africa, Asia/Pacific, America and Europe. The conferences focused on integrating early warning into sustainable development policy. The document notes the failure of scientists, policymakers, local authorities and other relevant actors to use early warning systems efficiently, and makes suggestions for improvement. It highlights challenges, lessons learnt and possible trajectories for further development of early warning systems, and outlines key steps towards strengthening frameworks, finding resources, and designating responsibilities. Key areas for action include improving data collection, ensuring that warning systems focus on people and achieve gender balance, and creating platforms for early warning dialogue.

This special issue of the journal Current Opinion in Environmental Sustainability presents a collection of interdisciplinary scientific articles and commentary on biodiversity. It includes new research in key areas such as food security and climate change.

It also reviews major initiatives that will be released or discussed during 2010, the International Year of Biodiversity. These include the Intergovernmental Platform on Biodiversity and Ecosystem Services, a new remote sensing project called the Group on Earth Observations — Biodiversity Observation Network, and key issues such as access and benefit sharing.

This article describes how a network of sensors, linked by software and the Internet, can provide an automatic satellite-based surveillance system for disasters such as volcanoes, wildfires and flooding.

The system, or 'Sensorweb', uses data from low resolution, high frequency sensors to trigger imagery from high resolution instruments. The low-resolution data is collected regularly (twice a day or more) from instruments such as NASA's Moderate Resolution Imaging Spectroradiometer (MODIS).

Anomalies, such as hotspots in the case of fires and volcanoes or surface water for floods, are automatically detected. The SensorWeb then sends a request to a higher resolution satellite such as Hyperion, which is very sensitive in the infrared spectrum, to request data over the area of interest. These data can then provide disaster managers with early warnings of adverse events.