Abstract Coastal hazards often result from the combination of different simultaneous oceanographic processes that occur at multiple spatial and temporal scales. To predict coastal flooding and erosion, it is necessary to accurately represent hydrodynamic conditions. For this reason, here we present a stochastic, climate based wave emulator that provides the hydrodynamic conditions needed for these predictions. The emulator can generate an infinitely long data series maintaining its statistical properties at different time scales, from intra-storm to inter-annual variability, and its link to large scale climate patterns. The proposed methodology relies on the use of weather types and an autoregressive logistic regression model forced with different variables to simulate daily scale chronology. Considering the dependencies of wave conditions on the different weather types, the intra-storm chronology is solved by means of shuffling and stretching historical wave sequences. To demonstrate the replicability of this emulator worldwide, we have applied the model to 3 different locations and found good agreement when compared to the historical data. Furthermore, to illustrate and explain the strengths and limitations of the emulator, we present a different application for each of the different locations.

中文翻译：

---

用于海岸线变化建模的多变量、随机、基于气候的波浪模拟器

摘要 沿海灾害通常是由在多个空间和时间尺度上发生的不同同时发生的海洋学过程组合而成的。为了预测沿海洪水和侵蚀，有必要准确地表示水动力条件。出于这个原因，我们在这里展示了一个随机的、基于气候的波浪模拟器，它提供了这些预测所需的水动力条件。模拟器可以生成无限长的数据系列，在不同的时间尺度上保持其统计特性，从风暴内到年际变化，以及它与大尺度气候模式的联系。所提出的方法依赖于天气类型的使用和强制使用不同变量的自回归逻辑回归模型来模拟每日尺度年表。考虑到波浪条件对不同天气类型的依赖性，通过对历史波浪序列进行混洗和拉伸来解决风暴内时间序列。为了证明该模拟器在全球范围内的可复制性，我们将该模型应用于 3 个不同的位置，并且与历史数据相比发现了良好的一致性。此外，为了说明和解释模拟器的优势和局限性，我们为每个不同的位置提供了不同的应用程序。