The Public preservation measure is a Kind of legal sanctions other than criminal punishments, which is to be imposed on a person who should be protected, because of his (her) future risk on the ground of his(her) behaviors open to the public, for the main purpose of giving medical treatment or educating or reinstating. Therefore in the view of a possibility of risk according to the responsibility, generally a criminal punishment has been understood as a treasure of peace preservation. That is to say that a criminal punishment is a retributive justice to a crime based on the responsibility, on the other hand the public preservation measure is a legal sanction for the social protection as well as his(her) correction and education related with social danger. A sharp line between the two legal viewpoints mentioned has been drawn. Accordingly as mentioned above judging from this point of view of distinction between a criminal punishment and the public preservation measure, the responsibility system is to be linked with a criminal punishment, but is not to be linked with the public preservation measure. For all that recently a view that the public preservation measure is able to be imposed even to the matter of responsibility has been on the rise. So the purpose of this study consists in the matter of the theory that the enforcement of preservation measure even to the responsibility is possible or not. In other words, it's a question that whether the responsibility system and the public preservation measure can be linked, if possible, how is it coming along with each other and if not possible, for what reason is it? On such problems a study has been pursued from the point of view of the criminal policy.

In this study, instead of economic estimation of forecast value, we evaluated the value score(VS) of the user satisfaction using the concept of satisfaction/dissatisfaction. We compared the collective Value Scores (cVS) based on outputs of probabilistic forecasts of precipitation seasonally in Seoul and Busan during the period of 2004 to 2013 and ÿnally found the optimum threshold that can improve cVS of both cities. When using 30% threshold, the users can expect a higher cVS compared with those using other thresholds. When using the seasonal optimum threshold in Seoul, the cVS is additionally higher by 9%. These results show the level of satisfaction of the forecast that can be improved when the meteorological communities inform the users the correct threshold of the rainfall probabilistic forecast.