A multivariate return period (MRP)-based ground motion selection methodology was recently proposed to identify hazard-consistent ground motions by holistically incorporating the joint rate of exceedance of vector conditioning intensity measures (IMs). Despite these advances, the high computational cost involved in target spectra identification hinders the use of high-dimensional vector conditioning IMs and especially for large return periods, and thus may limit the practical application of this ground motion selection methodology. In this paper, we introduce a two-step adaptive refinement procedure to substantially alleviate the computational hurdle, while still offering reliable target spectra selection. The refined MRP-based ground motion selection then further facilitates the development of a MRP-based seismic risk assessment (SRA) framework to extend its applicability into risk-based assessment. The efficacy and computational efficiency of the proposed refinement are evaluated through comparison with the original unrefined approach, and the unbiasedness and variability of the risk estimates under different use cases of SRA are thoroughly examined and validated through case-study analyses. Aiming at facilitating more confident and reliable seismic risk estimates, implications and advantages of leveraging the MRP-based ground motion selection in SRA are elaborated.

最近提出了一种基于多元回归期（MRP）的地面运动选择方法，通过整体纳入矢量调节强度测度（IM）的联合超出率，来确定与危害一致的地面运动。尽管取得了这些进步，但目标光谱识别中涉及的高计算成本却阻碍了高维矢量调节IM的使用，尤其是在较大的返回周期内，因此可能会限制这种地面运动选择方法的实际应用。在本文中，我们引入了两步自适应细化过程，以在很大程度上缓解计算障碍的同时仍提供可靠的目标光谱选择。完善的基于MRP的地震动选择进而进一步促进了基于MRP的地震风险评估（SRA）框架的开发，以将其适用性扩展到基于风险的评估中。通过与原始未改进方法进行比较来评估所提出改进方案的功效和计算效率，并通过案例研究分析对SRA不同使用案例下风险估计的无偏性和可变性进行了彻底检查和验证。为了促进更加自信和可靠的地震风险估算，阐述了利用SRA中基于MRP的地震动选择的含义和优势。通过与原始未改进方法进行比较来评估所提出改进方案的功效和计算效率，并通过案例研究分析对SRA不同使用案例下风险估计的无偏性和可变性进行了彻底检查和验证。为了促进更加自信和可靠的地震风险估算，阐述了利用SRA中基于MRP的地震动选择的含义和优势。通过与原始未改进方法进行比较来评估所提出改进方案的功效和计算效率，并通过案例研究分析对SRA不同使用案例下风险估计的无偏性和可变性进行了彻底检查和验证。为了促进更加自信和可靠的地震风险估算，阐述了利用SRA中基于MRP的地震动选择的含义和优势。