Risk: Vulnerability is measured in terms of expected harm/damage and so is risk. How can those terms be delineated from each other? Risk always involves the notion of probability of occurrence. So information on when or on how often indicates we are talking about risk. That could be captured in a continuous damage-frequency relationship or just the definition of the return period for a particular event scenario. While vulnerability informs about the consequences of possible adverse events, risk also provides information on how often or with what probability those scenarios have to be expected. For example: information on expected losses for an event during which the water level reaches 5 m above normal refers to hazard and vulnerability. Information on expected losses for a 200 year event during which the water level reaches 5 m above normal refers to risk. In another context: projecting the consequences of a 15 m tsunami is important but in order to make informed disaster management decisions it is necessary to know how often such an event can be expected. Disaster management decisions are based on risk and not only on hazard. Despite all the known shortcomings of databases of historic events, they do provide some means to create a magnitude-frequency relationship over a range of event magnitudes. This magnitude-frequency relationship can be an important tool for supporting the decisionmaking process with respect to the level of acceptable risk. Responsible disaster managers have to decide for what type of event a community should be prepared. To get prepared for the biggest possible event would be the safest way to go but it is rarely economically feasible; such high levels of protection are simply unaffordable and the benefits would not justify the costs. In addition, maintenance and alertness would be unmanageable over such long periods of time because the largest events can only be expected to occur after many years of quiescence. To summarise, risk is understood as a function of hazard, vulnerability, exposure, and resilience (See also: The Figure Below): Risk = f (hazard, vulnerability, exposure, resilience). Figure 4. Risk seen as a function of hazard, vulnerability, exposure, and resilience, while the mathematical relationship between the variables is unknown. The frequency or return period of adverse effects allows the individual or official decision maker to define a level of acceptable consequences. This is only possible if the decision maker understands what events to expect over time. Decisions will be different for a 10-year event as compared to a 5,000-year event. For decision making, information on the probability of occurrence is crucial. Often the historical record is too short to provide reliable magnitude-frequency relationships for particular hazards and regions. In addition, climate change has started to alter those relationships. This can be seen in Germany where the return period of the 100-year event for the Rhine and the Danube had to be revised as a 20-year or even a 10-year event (Alt, 2002). Or in the US where the Missouri River has had six 100-year floods since 1946 (Albright Seed Company, 1998). Fluke of nature or real trend? It is hard to decide. But many scientists agree that the trend is strongly supported by data. In situations of uncertainty it would be most appropriate to heed the precautionary principle. After all, if we are not even prepared to deal with the current risk situation, how shall we cope with and adapt to a deteriorating situation due to climate change?