Rui Chen

The probability distribution model for principal displacement accommodated on the surface main trace is a critical input to the fault displacement hazard analysis. This article presents a new model for strike-slip ruptures in the moment magnitude ( M ) range of 6 to 8.3. The new model is the outcome of a multi-year research effort to update the widely used model developed by Petersen and others in 2011. Updates include the adoption of the Fault Displacement Hazard Initiative database and enhancements to rupture and displacement data preparation. Statistical formulation and estimation have also been updated substantially. A three-parameter modified normal distribution that we refer to as the negative Exponentially Modified Gaussian distribution is adopted to model the probability distribution of the natural logarithm of principal displacement. Formulation for the mean parameter of the modified normal includes a random earthquake term, a nonlinear scaling relation with M , and an ellipse function for along-main-trace variation. The aleatory variability of the updated model now depends on M as well as site’s along-main-trace position. These updates not only significantly improve the fit to the distribution of the observed displacements but also yield reasonable 95th percentile predictions for M > 7.5 events. Alternative models representing the estimation uncertainty of the M -scaling relation are also developed. These new models are compared to the previous model in terms of percentile predictions and the calculated hazard curves. The steeper hazard curves from the new models yield a lower exceedance rate than the normal-distribution based model developed previously by Petersen and others.

Majority of the time series carry seasonal effects and show cyclical fluctuation patterns. The Census-X12 seasonal adjustment method and the HP filter method are based on factor decomposition theory, which divides the original time series into four parts: long-term trend, cyclical fluctuation, seasonal characteristics and random fluctuation terms. This paper selects monthly price data of China's major meat products: chicken, beef, mutton and pork from 2009 to 2020 to derive medium and longterm trends, seasonal fluctuations and cyclical patterns in the prices of major meat products. It is found that the X12 seasonal adjustment method and the HP filter method are able to accurately extract the characteristics of different factors in the prices of major meat products in China. The price fluctuation patterns of beef and lamb are basically the same, with a cyclical cycle of 24 and 29 months respectively; the cyclical cycle of pork prices is consistent with that of chicken prices and fluctuates more frequently at random.