A stepwise method for determining the number of component distributions in a mixture

This paper considers the problem of estimatingm, the number of components in a finite mixture of distributions from a parametric family. A step-up procedure using the bootstrap method is proposed. Some properties of the procedure are illustrated with simulation studies. An example of the method, applied to orientation of beach clasts, is given.     

Statistics of nonlinear wave crests and groups

Groups of large nonlinear waves with sharper higher crests can pose hazards to ships, induce harbor resonance and cause wave-overtopping of fixed and floating structures. Past interest in wave groups has mostly been focused on the statistics and modeling of linear wave groups. Studies on nonlinear wave groups are surprisingly few, and address deep water waves only. Here, statistics of nonlinear wave crests and wave-crest groups in deep and transitional water depths are considered, using an appropriate second-order representation for crest heights and the continuous wave-envelope approach. In particular, theoretical expressions describing the statistics of nonlinear wave crests and their groups are posed in the form of a simple second-order transformation of well-known results on linear waves. Predictions from the transformation so posed compare well with nonlinear wave data gathered in the North Sea, and demonstrate that nonlinearities do affect the statistics of large wave crests and their groups significantly.     

The Mw = 6.0, 7 September 1999 Athens Earthquake

On 7 September 1999 at 11:56 GMT a destructive earthquake (M_w = 6.0) occurred close to Athens (Greece). The rupture process is examined using data from the Cornet local permanent network, as well as teleseismic recordings. Data recorded by a temporary seismological network were analyzed to study the aftershock sequence. The mainshock was relocated at 38.105°N, 23.565°E, about 20 km northwest of Athens. Four foreshocks were also relocated close to the mainshock. The modeling of teleseismic P and SH waves provides a well-constrained focal mechanism of the mainshock (strike = 105°, dip = 55° and rake = -80°) at a depth of 8 km and a seismic moment M₀ = 1.010²⁵ dyn·cm. The obtained fault plane solution represents normal faulting indicating an almost north-south extension. More than 3500 aftershocks were located, 1813 of which present RMS < 0.1 s and ERH, ERZ < 1.0 km. Two main clusters were distinguished, while the depth distribution is concentrated between 2 and 11 km. Over 1000 fault plane solutions of aftershocks were constrained, the majority of which also correspond to N–S extension. No surface breaks were observed but the fault plane solution of the mainshock is in agreement with the tectonics of the area and with the focal mechanisms obtained by aftershocks. The hypocenter of the mainshock is located on the deep western edge of the fault plane. The relocated epicenter coincides with the fringe that represents the highest deformation observed on the differential interferometric image. The calculated source duration is 5 sec, while the estimated dimensions of the fault are 15 km length and 10 km width. The source process is characterized by unilateral eastward rupture propagation, towards the city of Athens. An evident stop phase observed in the recordings of the Cornet local stations is interpreted as a barrier caused by the Aegaleo Mountain.     

Foreshocks and aftershocks of the M W = 7.1, 1992, earthquake in the Atrato region, Colombia

We study the October 18, M _W = 7.1, 1992 Atrato earthquake, and its foreshocks and aftershocks, which occurred in the Atrato valley, northwestern Colombia. The main shock was preceded by several foreshocksof which the M _W = 6.6, October 17 earthquacke was the largest. Inparticular, we examine foreshocks and aftershocks performing joint-hypocenter relocations using high quality Pn and Sn wave readingsfrom permanent regional networks. We observed a few hours prior to the main shock a sudden increase of foreshocks. Maybe this could be used as a predictor since foreshocks have been known for other major events in the region. Our locations align for 90 km with a trend of 5^° ±4^° in agreement with the Harvard CMT solution showing the faultplane trending 9^° to be the plane of rupture. In relation to theepicenter of the main shock, maximum intensities were located to thesouth, consistent with a rupture that traveled from north to south witha larger energy release in the south as suggested by an empirical Green'sfunction study (Li and Toksöz, 1993; Ammon et al., 1994). The boundarybetween the Panama and North Andes blocks has been placed close to thePanama-Colombia border as either a sharp boundary or a diffuse zone. TheAtrato earthquake, however, shows that the plate boundary between thePanama and North Andes microblocks is a diffuse deformation zone. Thiszone has a width of at least 2^° stretching from 78^°W to 76^°W. Quantification of earthquake moment release (during the past30 years) in this zone shows a similar amount of moment release in thewestern and eastern parts of this zone.     

基于DEM的黄河流域天文辐射空间分布

基于1 km×1 km分辨率的数字高程模型(DEM) 数据,利用建立的起伏地形下天文辐射分布式计算模型,计算了黄河流域1 km×1 km分辨率各月天文辐射的空间分布。结果表明:局部地形对黄河流域年和四季天文辐射的空间分布影响明显;在太阳高度角较低的冬季,地理和地形因子对天文辐射的影响相当强烈,山区天文辐射的空间差异大,1月份向阳山坡(偏南坡) 天文辐射可为背阴山坡(偏北坡) 的2~3倍,极端天文辐射的差异可达10倍以上;而在太阳高度角较高的夏季,天文辐射空间差异较小,7月份不同地形极端天文辐射的差异仅在16%左右;四季中,地形对天文辐射影响的程度为冬季>秋季>春季>夏季。     

Fault plane solutions of the january 26th, 2001 bhuj earthquake sequence

A 12-station temporary microearthquake network was established by the Geological Survey of India for aftershock monitoring of the January 26th, 2001 Bhuj earthquake (M _w 7.6) in the Kutch district of Gujarat state, western India. The epicentres of the aftershocks show two major trends: one in the NE direction and the other in the NW direction. Fault-plane solutions of the best-located and selected cluster of events that occurred along the NE trend, at a depth of 15–38 km, show reverse faulting with a large left-lateral strike-slip motion, which are comparable with the main-shock solution. The NW trending upper crustal aftershocks at depth <10 km, on the other hand, show reverse faulting with right-lateral strike-slip motion, and the mid crustal and lower crustal aftershocks, at a depth of 15–38 km, show pure reverse faulting as well as reverse faulting with right-lateral and left-lateral strike-slip motions; these solutions are not comparable with the main-shock solution. It is inferred that the intersection of two faults has been the source area for stress concentration to generate the main shock and the aftershocks.     

Extreme value theory based on the r largest annual events: a robust approach

In Smith (1986, J. Hydrol. 86, 27–43), a family of statistical distributions and estimators for extreme values based on a fixed number r > = 1 of the largest annual events are presented. The method of estimation was numerical maximum likelihood. In this paper, we consider the robust estimation of parameters in such families of distributions. The estimation technique, which is based on optimal B-robust estimates, will assign weights to each observation and give estimates of the parameters based on the data which are well modeled by the distribution. Thus, observations which are not consistent with the proposed distribution can be identified and the validity of the model can be assessed. The method is illustrated on Venice sea level data.