ABSTRACT
Way back in Chapter 1, it was noted that the use of models to predict the response of an environmental system was normally done for a purpose, and in many cases it was done to inform a decision in managing the environment, whether at local, national or global scale. That is why assessing uncertainty in such predictions is important. Uncertainty in prediction implies a risk of being wrong in assessing the response to some change or management strategy. In situations where a decision must be made, a risk of being wrong may imply important serious economic consequences. At the time of writing, the United States and some other nations had not ratified the Kyoto agreement on reducing greenhouse gas emissions. This decision was widely reported as being based on an argument that, since the scientific predictions of the effects of greenhouse gas emissions on future climate are still so uncertain, there is no justification for taking a decision to reduce emissions when it will have such a serious impact on the economy of these countries. At the time of Kyoto their politicians concluded, perhaps dangerously, that the risk of short-term economic recession is greater than the risk of possible long-term climate impacts.
