Propensity of rivers to extreme floods:climate-landscape controls and early detection (PREDICTED)

世界各地在短时间内发生了几次主要河流洪水，这表明用于评估洪水规模和频率的曲线(称为洪水频率曲线)出现了急剧上升的偏差(称为阶跃变化)，这让公民、工程师和科学家想知道洪水是否发生了变化。对于社会事业和经济部门来说，这是一个重要问题，它们需要对河流流域的洪水危险进行可信的评估，如城市规划、水资源管理和保险业。气候变化和人类活动(例如集约化农业和城市化)造成的土壤储存损失常常被认为是引发这些现象的原因。然而，河流流域的正常运行也可能导致看起来极端的洪水，因为它们还没有被记录在现有的数据系列中。该项目旨在确定在外部驱动因素(如降雨)没有变化的情况下，气候和流域景观之间的相互作用是否能够产生极端洪水。此外，它的目的是提供方便的方法，根据在尚未观察到极端事件和没有洪水长期记录的地区也可以表现出的特征来评估这种可能性。将开发客观地检测观测和模拟的洪水频率曲线明显上升的起点(即阶跃变化的位置)的方法。在验证了由于可用数据的长度有限，阶跃变化可能不仅仅是数学上的假，并且阶跃变化不一定需要改变降雨量或土地利用之后，将调查与其外观有关的流域属性。重点将放在气候和地貌特征上，最近提出的关于河流流域内发生的过程的简化描述表明，这些特征与产生径流有关。将得出一个索引，该索引总结了这些特征如何相互作用以产生(或不产生)阶跃变化。这一指数预测阶跃变化位置的能力将与来自不同地理区域的一大组河流流域的观测结果进行对比测试。还将评估采用阶跃变化作为正常洪水和极端洪水之间的分界线的可能性。该项目的最终目标是编制图表，说明观测到的时间序列的可靠性，以评估受不同气候和景观环境影响的河流流域的洪水风险，或为获得可靠的洪水估计所需的监测时间。还将以情景的方式对这些因气候变化而建议的具有实际意义的指标进行修改。

The occurrence around the world of several major river floods in a short period of time, pictured by sharp upward deviations (called step changes) of the curves used to assess how large floods can be and how often they will occur (called flood frequency curves), recently led citizens, engineers and scientists to wonder whether floods have changed. This is an important issue for social endeavors and economic sectors which require trustworthy evaluations of the flooding hazard of river basins, such as urban planning, the management of water resources and the insurance industry. Alterations of the climate and loss of soil storage due to human activities (e.g. intensive agriculture and urbanization) are often claimed to trigger these phenomena. However, the normal functioning of river basins might also cause floods that appear as extreme because they were not yet recorded in the available data series.The project aims to ascertain whether the interactions between climate and landscape of river basins are able to produce extreme floods, also when no changes of the external drivers (e.g. rainfall) occur. Moreover, it aims to provide handy methods to assess this possibility based on features that can be characterized also in areas where extreme events have not been observed yet and long records of floods are not available. Methods to detect in an objective way the beginning of distinct rises of observed and modeled flood frequency curves (i.e. the position of step changes) will be developed. After verifying that step changes may be more than mere mathematical fakes due to the limited length of the available data, and that modifications of rainfall or land use are not necessarily required for having step changes, attributes of river basins linked to their appearance will be investigated. The focus will be placed on climatic and landscape features that have been suggested as relevant for streamflow generation by a recently proposed simplified description of the processes taking place within river basins. An index which summarizes how these features interact to produce (or not) step changes will be derived. The capability of this index to predict the position of the step change will be tested against observations in a large set of river basins from different geographical regions. The possibility to adopt the step change as a divide between normal and extreme floods will be evaluated as well. The ultimate objective of this project is the development of charts describing the reliability of observed time series for assessing flood hazard in river basins subject to different climatic and landscape settings, or the required length of monitoring to obtain trustworthy estimates of floods. Modifications to these indications of practical interest recommended as a consequence of climate change will be also provided by means of scenarios.