A MIXTURE MODEL OF ANNUAL PRECIPITATION IN PERU*

A two component mixed log-normal distribution effectively models annual precipitation totals at two stations in Peru characterized by widely differing interannual patterns of precipitation. Physical evidence supports the division of station records into two subsamples. Years with ENSO events and years without ENSO events identify the components of a mixed probability model. The mixed model produces a superior fit to the two parameter log-normal distribution. Model application provides a reliable means of precipitation prediction and also quantitatively describes the highly variable temporal and spatial pattern of annual precipitation in western Peru.     

Geostatistical analysis of field hydraulic conductivity in compacted clay

Hydraulic conductivity(K) of fractured or porous materials is associated intimately with water flow and chemical transport. Basic concepts imply uniform flux through a homogeneous cross-sectional area. If flow were to occur only through part of the area, actual rates could be considerably different. Because laboratory values ofK in compacted clays seldom agree with field estimates, questions arise as to what the true values ofK are and how they should be estimated. Hydraulic conductivity values were measured on a 10×25 m elevated bridge-like platform. A constant water level was maintained for 1 yr over a 0.3-m thick layer of compacted clay, and inflow and outflow rates were monitored using 10×25 grids of 0.3-m diameter infiltration rings and outflow drains subtending approximately 1×1 m blocks of compacted clay. Variography of inflow and outflow data established relationships between cores and blocks of clay, respectively. Because distributions of outflow rates were much less and bore little resemblance to the distributions of break-through rates based on tracer studies, presence of macropores and preferential flow through the macropores was suspected. Subsequently, probability kriging was applied to reevaluate distribution of flux rates and possible location of macropores. Sites exceeding a threshold outflow of 100×10^–9 m/s were classified as outliers and were assumed to probably contain a significant population of macropores. Different sampling schemes were examined. Variogram analysis of outflows with and without outliers suggested adequacy of sampling the site at 50 randomly chosen locations. Because of the potential contribution of macropores to pollutant transport and the practical necessity of extrapolating small plot values to larger areas, conditional simulations with and without outliers were carried out. Simulated scenarios based on all available data compared well with conditional simulations based on randomly chosen locations.This paper was presented at MGUS 87 Conference, Redwood City, California, 14 April 1987.     

Assessment of earthquake hazard in Panama based on seismotectonic regionalization

Assessment of seismic hazard in Panama is made using a seismotectonic regionalization model. The coefficients of Gumbel's Type-I distribution are calculated and return periods for several magnitudes are found. From these coefficients intensities, peak ground acceleration and earthquake hazard for a set of return periods and epicentral distances are estimated and substantial variations in the probability of occurrence are noted. The Panama Fracture Zone (PFZ) and the Panama-South America Suture Zone (PSZ) provinces are the most active in producing earthquakes with a magnitude of about 7.0 in less than 16 yr. Magnitude 7.0 earthquakes in the Azuero province have a return period of about 160 yr, whereas in the Panama Deformed Belt (PDB) province the return period for magnitude 7.5 events is about 175 yr.     

Geographical Information Systems and Geostatistics for Modelling Radioactively Contaminated Land Areas

The Chernobyl plume contaminated vast lands of Europe with radiocaesium (¹37Cs) in 1986 because of the deposition of radionuclides on the ground by wet and dry deposition processes. Nevertheless, in a nuclear emergency, contamination data may be very sparse and there is need to make rapid and scientifically supported decisions. Here we analyze the rainfall field, an important precursor of the wet deposition, during the passage of the plume. Thus, estimating rainfall spatial variability can help to identify possible contaminated areas and associated risks when rainfall exceeded a given threshold. In this paper, we show that the conditional probabilities of exceeding threshold rainfall values could be spatially assessed using the mutual benefits of linking geostatistical and geographical information system (GIS) to quantify the evaluation of the risk involved in decision making. In particular, the non-parametric geostatistic technique, termed Indicator Kriging (IK), enables one to efficiently estimate the probability that the true value exceeds the threshold values by means of the indicator coding transform. Afterward, GIS has been used to find the areas probably affected by wash-out (probability >0.5 that rainfall is above a certain threshold). The experimental study has been focused on a test site in Beneventan agroecosystem (Southern Italy) to model the spatial uncertainty over a continuous area from sparse rainfall data. This enabled to generate probability maps delineating area potentially affected by to contamination to be monitored after wet deposition of Chernobyl releases.     

Seismic hazard analysis in zones of time and space interdependence: An application to S?o Miguel Island, Azores

The Azores Islands are located in the mid-Atlantic region near the triple junction where the Euro-Asiatic-African-American plates join together. Seismic activity in the area is very high, as can be observed either from historical events since the fifteenth century, from present day microseismicity, and from direct and indirect measurements of recent tectonic deformation. Volcanic activity is also present throughout the region.All available information, even data exhibiting low quality, was used to develop hazard models of São Miguel Island. Source zones were established based on both the global tectonic behaviour of the region and on the local active fault structures. Recurrence laws for São Miguel Island, for which historical information seemed quite incomplete, were obtained from the large events in the entire archipelago and from their remarkable pattern of time and space dependence, and complimented by information on long-term fault deformation (for the longer recurrence periods) and on high precision instrumental network (for the very short recurrence periods).Attenuation laws were derived from data on events felt and/or recorded in the Island.Hazard maps were obtained through a modified version of McGuire's algorithm for several geometries of source areas and results compared with the maximum observed intensity of historical events.Abstract presented at the 21st General Assembly of the European Seismological Commission held in Sofia, 1988.     

The Neutral, Barotropic Planetary Boundary Layer, Capped by a Low-Level Inversion

The presence of a low-level, capping inversion layer will affect the height and structure of the planetary boundary layer (PBL). Results from models of varying levels of sophistication, including analytical, turbulent kinetic energy (TKE), second-order closure (SOC), large-eddy simulation (LES) and direct numerical simulation (DNS) models, are used to investigate this influence for the neutral, barotropic PBL. Predicted and observed profiles of stress and geostrophic departure components, and integral measures, such as the parameters of Rossby-number similarity theory, are compared for the KONTUR, Marine Stratocumulus, JASIN, Leipzig, Pre-Wangara and Upavon field experiments.Analytical models of the equilibrium value of inversion height z_i, which depend on the surface friction velocity u_*, and both the Coriolis parameter f and the free-flow Brunt-Väisälä frequency N, are found to give reasonable estimates of the PBL height. They also indicate that only the KONTUR and Marine Stratocumulus experiments were strongly influenced by N. More quantitative comparisons would require larger, more comprehensive datasets. The effects of the presence of a capping inversion on the profile structure were found to be insignificant for h_* = |f|z_i/u_* > 0.15.The simple analytical model performed quite well over all values of h_*; it predicted the profiles of the longitudinal stress component (in the direction of the surface stress) better than the lateral component. The more advanced models performed well for small values of h_* (for flow over the sea), but systematically underestimated the cross-isobaric angle for flow over land. These models predicted the profiles of the lateral stress component better than the longitudinal component. The profiles of the analytical model agreed with those of the advanced models when the constant eddy viscosity of the outer layer was increased.Agreement with DNS was achieved by increasing the eddyviscosity of the analytical model by a factor of 5.Zilitinkevich and Esau(2002, Boundary-Layer Meteorology, 104, 371–379)suggest that the neutral, barotropic values of A and B of Rossby-numbersimilarity theory are not universal constants, but depend on the ratio N/|f|. The dependence for A and B is calculated using the analytical model and TKE models. Over the sea (h_* 0.1; N/|f| 100, where we have used the Zilitinkevich-Esau relation to convert between h_* and N/|f|) there is agreement between the model predictions and observations; however over land where the equilibrium boundary-layer height is greater (h_* 0.35; N/|f| 10) the inconsistency between the advanced model predictions (TKE, SOC, LES, and DNS) and observations, as noted previously by Hess and Garratt, still exists. We attribute this disagreement to violations of the strict assumptions of steady, horizontally homogeneous, neutral, barotropic conditions implicit in the observations. At small values of z_i and a strongly stable background stratification (h_* 0.04; N/|f| 1000) both the TKE and analytical models predict that A and B depend significantly on h_*, however observations are unavailable to confirm these predictions. Zilitinkevich and Esau call this case the `long-lived near-neutral PBL', and state that it is found in cold weather at high latitudes.     

Selection of environmental conditions for nearshore structure design

1　Introduction　InChina ,coastalcitiesaremostlyregionaleconomicdevelopmentcenters.Becauseoftheirspecialgeogra phiclocations ,typhoonsandassociatedstormsurgescauseheavylossesoflivesandproperties.In 1 992 Qing daowasfloodedduetothecoincidenceoftheastrono m…     

嵌入式马尔科夫链的地质属性建模与应用

传统的地质统计学所形成的算法理论和方法(如克里格算法)过分依赖样品的数据,变异函数参数众多,给地质模拟造成很大困难。基于马尔科夫链的地质属性建模采用转移概率描述样品区域的各种参数变量,通过转移概率矩阵直接推导地质属性分布比例、平均长度,其简化了地质空间各向异性处理过程,克服了传统地质统计学中参数众多且复杂难以计算和地质体分布过程中存在不对称性等缺陷,使得整个地质属性建模的过程更简洁、清晰,容易理解,且建模的结果可以很好地反映地质体空间分布的复杂性。该文利用马尔科夫链对南京市河西地区的新近地质层进行了地质属性建模,实例应用表明,使用该模型进行地质属性建模可为进一步的数值模拟提供支持。     

基于位置误差的分布制作热带气旋路径袭击概率预报

为了探讨热带气旋路径预报的不确定性,利用2004-2007年4年中央气象台综合预报的误差分布状况,建立热带气旋路径综合预报的统计集合模型,制作热带气旋路径的袭击概率预报及产品,并为警报发布工作中有关登陆地段的预报提供有意义的参考.     

湖北地区云地闪电时空分布特征分析

为揭示湖北地区云地闪电时空分布特征及雷电活动规律,以满足雷电灾害防护和雷击风险评估工作需要,采用湖北不同地理位置的13个雷击探测仪组成的闪电监测定位系统获取的2006年3月至2009年2月的云地闪电资料,从闪电的极性分布、日变化、月变化、强度、闪电密度、累计概率分布等方面进行统计分析。结果表明:云地闪电中负闪电占闪电总数的96.2%;正闪电占闪电总数的3.8%;闪电频次的日变化呈明显的单峰单谷型,一日中,最大值出现在15—16时,最小值在09—10时;一年中,4—8月闪电次数占全年闪电总数的95.9%,其中7—8月闪电次数最多;正、负闪电的强度主要集中在10～50 kA,大于30 kA的累积概率在50%以下,大于60 kA的累积概率8.1%,大于100 kA的累积概率1.6%;拟合出湖北地区大于某雷电流强度累积概率方程,经统计分析,实测值与计算值相关系数达0.99998。闪电密度分布呈明显的地域性差异,鄂东南的嘉鱼县、咸宁、黄石、鄂州市一带为闪电高密度区,鄂西的远安县、当阳市附近为闪电次高密度区。山区与丘陵、平原交接地带,即地表状况发生明显变化的地带,是雷电多发地带。鄂西南山区县市雷电日数较多,闪电密度不一定随之增加,其原因可能是山区县市土壤电阻率较大和山区局地小气候环境共同影响的结果。