Intensity–duration rainfall thresholds are commonly used in regional-scale landslide warning systems. In this manuscript, 3D thresholds are defined also considering the mean rainfall amount fallen in each alert zone (MeAR, mean areal rainfall) in Emilia Romagna region (Northern Italy). In the proposed 3D approach, thresholds are represented by a plane instead of a line, and the third dimension allows to indirectly account for the influence of complex rainfall patterns. MeAR values are calculated according to different time periods ranging from 7 to 30 days, and all threshold parameters are calibrated independently for the 8 alert zones in which the region is divided. The approach was validated and compared with classical intensity–duration thresholds, finding that the 3D threshold may be used to get better performances, especially in terms of a consistent reduction of false alarms:− 20 to − 86%, depending on the alert zone and the selected MeAR duration. These results open new encouraging perspectives for the development of the regional warning system that is operated in the study area.

中文翻译：

---

定义 3D 降雨阈值以提高滑坡预警系统的性能

强度-持续时间降雨阈值通常用于区域尺度的滑坡预警系统。在这份手稿中，3D 阈值的定义还考虑了艾米利亚罗马涅地区（意大利北部）每个警报区（MeAR，平均面积降雨量）的平均降雨量。在提议的 3D 方法中，阈值由平面而不是线表示，第三维允许间接考虑复杂降雨模式的影响。MeAR 值根据 7 至 30 天的不同时间段计算，所有阈值参数均针对该区域划分的 8 个警戒区独立校准。该方法经过验证并与经典的强度-持续时间阈值进行比较，发现 3D 阈值可用于获得更好的性能，尤其是在持续减少误报方面：- 20% 至 - 86%，具体取决于警报区域和所选的 MeAR 持续时间。这些结果为研究区域内运行的区域预警系统的发展开辟了新的令人鼓舞的前景。