For many important aspects of weather forecasts, predictive skill has improved by about one day per decade in recent years. A seven-day forecast now, for example, is approximately as accurate as a three-day forecast was four decades ago. Thus, progress in improving forecasts during about four decades has resulted in extending the useful forecast range by about four days. This is a remarkable accomplishment. However, any error in specifying the initial conditions, and we know errors are inevitable, will make the forecast go wrong after a certain time. As for our current forecasting skill, we can predict weather for at least several days. For the largest-scale features – very large weather systems, highs and lows on continental scales – the limit is thought to be a few weeks. Recent progress in research has led to major advances in our understanding of climate. These advances have greatly increased the confidence of scientists in their ability to make skillful and useful forecasts of how the climate system will respond to increased amounts of greenhouse gases in the atmosphere.