Mapping large territories for earthquake-induced soil liquefaction hazard may sound like an oxymoron since soil liquefaction is by itself a spatially highly localized phenomenon. Thus, the zonation of liquefaction hazard at the continental scale (i.e. megazonation) is a truly hard facing challenge even if accepting a low level accuracy. Yet, the availability of a megazonation chart of liquefaction hazard could be useful to identify territories that in case of an earthquake may display this phenomenon of soil instability distinguishing them from the regions where soil liquefaction is not expected even in case of strong ground shaking. A representation of the spatial variability of liquefaction hazard potential within a single country is within reach considering the resolution and accuracy of geological and geotechnical information that is currently available in the most developed nations. The LIQUEFACT project fully addressed in a specific work package the problem of constructing georeferenced, earthquake-induced soil liquefaction hazard maps in continental Europe for various return periods. They were built using homogeneous datasets in Europe on the expected seismic hazard and on the geological, geomorphological, hydrogeological, shallow lithology and digital terrain information. A probabilistic prediction model based on a logistic regression for liquefaction occurrence was purposely developed using a set of optimal geospatial predictors (explanatory variables) which include the weighted-magnitude peak ground acceleration, the weighted-mean shear-wave velocity in the top 30 m (V_S30) and the compound topographic index. The optimal geospatial descriptors were selected based on the Luco and Cornell methodology, namely on the criteria of efficiency, practicality and proficiency. To calibrate and successively validate the logistic regression, a database of liquefaction manifestations occurred in continental Europe was used. Although the level of accuracy provided by these models at a local scale is low, still the availability of continental charts of liquefaction hazard may help policy makers and administrators to prioritize which urbanized territories should be investigated further to assess the seismic risk of structures and infrastructures associated to the occurrence of soil liquefaction.

由于土壤液化本身是空间上高度局部化的现象，因此为地震引发的土壤液化危害绘制大领土的地图听起来像是矛盾。因此，在大陆范围内液化危害的分区（即兆分区））即使接受较低的准确性也确实是一个艰巨的挑战。然而，提供液化危险的巨型分区图可能有助于确定那些在地震情况下可能会显示这种土壤不稳定现象的地区，从而将其与即使在强烈地面震动的情况下也不会发生液化的地区区分开来。考虑到大多数发达国家当前可获得的地质和岩土信息的分辨率和准确性，在一个国家范围内表示液化危险可能性的空间变异性是可行的。LIQUEFACT项目在一个特定的工作包中完全解决了在不同回归期在欧洲大陆构造地理参考，地震引起的土壤液化危害图的问题。它们是使用欧洲的均匀数据集构建的，这些数据集涉及预期的地震危害以及地质，地貌，水文地质，浅层岩性和数字地形信息。使用一组最佳地理空间预测变量（解释变量），有针对性地建立了基于逻辑回归的液化概率预测模型，其中包括加权幅度峰值地面加速度，顶部30 m（ V_S30）和复合地形指数。基于Luco和Cornell方法，即基于效率，实用性和熟练程度的标准，选择了最佳的地理空间描述符。为了校准并逐步验证逻辑回归，使用了欧洲大陆发生的液化表现数据库。尽管这些模型在本地范围内提供的准确性水平很低，但是大陆液化危害图的可用性仍然可以帮助决策者和管理者确定应该进一步调查哪些城市化区域的优先级，以评估与建筑物和基础设施相关的地震风险到土壤液化的发生。