资料同化在空气质量预报中的应用

随着数值模式的不断完善和观测技术的不断提高，资料同化逐渐成为能够进一步提高数值预报水平的一种有效方法。20世纪70年代，资料同化开始引入空气质量预报领域，成为当前大气环境科学研究的一个新方向。简要介绍了资料同化的含义，较详细地介绍了Kalman滤波法、四维变分同化法、牛顿松弛法的基本思想和优缺点，重点阐述了国内外资料同化在空气质量预报中的研究应用情况，最后指出资料同化应用于空气质量预报时存在的问题和今后的研究方向。     

数值天气预报检验方法研究进展

数值天气预报检验是改进及应用数值模式的重要环节。近年来,模式检验中的观念不断更新,适用于不同预报产品及不同用户需求的模式检验方法也不断涌现。首先简单回顾了以列联表为基础的传统的模式检验方法。其次重点总结了伴随高分辨率数值预报而出现的空间诊断检验技术,按照检验目的的不同,诊断方法可以归纳为:①基于滤波技术的分辨模式在不同时空尺度上预报能力的邻域法、尺度分离法;②利用位移偏差诊断模式预报位置、面积、方位、轴角等与观测差异的属性判别法、变形评估法。然后阐述了集合样本成员的概率分布函数(PDF)、集合预报与观测概率分布函数相似程度、事件发生的概率预报等集合预报检验方法。最后论述了空间诊断技术、集合预报检验方法的适用领域,并讨论了模式检验中存在的一些问题及未来的发展方向。     

MODE检验在天气预报中的应用研究进展

客观评价高分辨率中小尺度变量的预报表现是数值天气预报模式应用和发展的重要环节。传统点对点的检验在高分辨率数值预报模式评估中存在显著不足。空间面向对象或目标法MODE检验方法利用卷积函数和给定的阈值在预报和观测场中识别目标对象、诊断模式的预报表现,在天气预报中有着广泛应用。首先系统回顾了MODE检验方法的学术思想、技术架构、算法流程和检验指标;详细归纳了MODE检验在降水预报、天气雷达、卫星云图、集合预报以及其他不同要素或物理量场中的典型应用,阐述了检验结果在数值模式预报质量评价中的意义和对天气预报结果主客观订正的影响;介绍了近年来MODE检验方法的更新发展,主要包括考虑不匹配目标对象属性的综合评价指数MCS,以椭球体为目标的三维时域空间对象追踪和MODE扩展的时域检验方法 MODE-TD;最后总结和讨论了MODE检验方法的适用性、优势和不足,同时对MODE检验方法未来的发展方向和应用前景进行了展望。旨在为更好地应用MODE检验方法诊断数值模式的预报性能提供参考。     

基于GRAPES的西北地区沙尘暴数值预报模式及其应用研究

介绍了中国气象局兰州干旱气象研究所和中国气象科学研究院数值预报研究中心合作研制的耦合于GRAPES（Global/Regional Assimilation and Prediction Enhanced System）的沙尘暴数值模式GRAPES_SDM。该模式包括沙尘的起沙、传输、吸湿增长、并合、干沉降与云下清洗等详细的物理过程，可以对沙尘暴的起沙和空气中沙尘浓度进行模拟和预报。在此基础上形成了西北地区的沙尘暴数值预报模式系统，并于2005年4月开始在兰州中心气象台试运行，同时还将模式结果与卫星遥感资料反演的沙尘暴监测结果进行了对比验证。以2005年4月17~19日和5月28~29日发生在西北地区的2次强沙尘暴为例，利用GRAPES_SDM对这两次沙尘天气的起沙、传输、扩散直至消散进行了数值模拟。结果表明，模式输出的沙尘时空分布与实况观测和卫星云图监测的沙尘分布范围基本一致，说明该模式系统对西北地区沙尘暴天气的起沙、传输有较好的模拟和预报能力。     

全天候卫星微波观测资料变分同化研究进展

云和降水区微波观测包含大量与天气系统,特别是台风、暴雨等灾害性天气系统发生发展密切相关的大气信息,因此微波资料的全天候同化应用成为当前数值预报领域的热点研究问题。过去20年间,全球几大数值预报中心逐步开展了全天候同化技术的研究和业务应用,证实了全天候卫星微波观测资料能够改进模式中的质量、风场、湿度以及云和降水场的初始信息,从而改进数值预报模式的预报效果。通过梳理和评述全天候卫星微波观测资料同化方法,分析其中的关键技术问题和目前存在的困难和挑战,为未来在我国数值天气预报领域开展全天候同化研究提供依据。随着我国新一代数值天气预报模式的发展应用,加强我国全天候资料同化技术的研究将会在业务中发挥更大的科学效益和应用效益。     

基于台风气压模型的沿海天文潮与风暴潮耦合数值模拟试验研究

结合4种台风气压模式,采用改进的超浅台风风暴潮联合模型对辽宁沿海天文潮与风暴潮进行数值模拟试验。试验结果表明:在4种台风气压模式下,高桥模式或捷氏模式误差相对较小,在台风路径预报中有最大风速信息时可以直接运用以上两种模式进行台风增水影响;相比于传统算法,改进后的沿海天文潮与风暴潮耦合数值模型模拟的台风增水相对误差小于30%的保证率提高29%,达到规定的台风增水预报的精度要求。研究成果对于沿海地区防洪决策具有重要的参考价值。     

降水临近预报及其在水文预报中的应用研究进展

随着全球气候变化、自然变迁及陆表生境改变,极端天气频发且呈现出多尺度时空变异特征,对其进行预报和预警一直是气象水文领域关注的焦点。临近预报可较准确地预报未来短时间天气显著变化,是当前预报强降水等极端事件的主要手段。从基于天气雷达0~3 h外推临近预报、融合数值模式0~6 h临近预报的发展历程梳理了临近预报的研究进展,阐述了雷达外推算法的发展进程、雷达外推预报与数值模式预报融合技术进展,指出"取长补短"的0~6 h融合预报在提高降水预报精度、延长降水预见期等多方面有较大的发展潜力,进一步探寻及提升融合技术是未来融合预报发展的核心。将临近预报以气象水文耦合的方式引入水文预报是从源头提高水文预报精度、保障水文预报效果的主要途径,总结了现阶段主流耦合方式、空间尺度匹配技术、水文模型不确定等陆气耦合中的关键问题,阐述了外推临近预报、融合临近预报作为水文预报输入的研究进展,明确了融合临近预报在延长洪水预见期、提高洪水预报精度中存在优势,并讨论了未来的研究重点及发展方向。     

基于进化神经网络混凝土大坝变形预测

根据丰满大坝多年变形观测数据,建立了基于进化神经网络混凝土大坝变形预测方法。经典的BP神经网络的缺陷在于收敛速度慢和泛化能力弱等特性。与普通的多元回归方法和传统的BP神经网络相比,采用遗传算法训练的人工神经网络预测模型预报大坝的变形具有精度高和全局收敛的特点。在丰满大坝工程实际应用表明,所建立的基于进化神经网络混凝土大坝变形预报方法与广泛采用的统计方法相比,可以显著提高大坝变形预报精度。     

论地震数值预报——关于我国地震预报研究发展战略的思考

随着中国工业化和社会现代化进程的加快 ,中国城市化进程必将进一步加速。如何减轻地震灾害的问题正变得日益严峻。尽管地震预测是一个举世公认的国际性科学难题 ,但在强化各种减轻地震灾害措施的同时 ,仍须大力推进地震预测研究。为此 ,需要打破长期徘徊在以地震前兆异常监测为基础的经验性预测局面 ,把注意力尽快转向研究以动力学为基础的数值预报。以GPS为代表的空间对地观测技术 ,岩石圈巨型高分辨率宽频带流动地震台阵观测技术以及数值模拟技术已经为地震数值预报研究提供了前所未有的技术基础。以地震数值预报为目标的GPS阵列地壳形变连续观测 ,高分辨率地壳上地幔结构探测 ,地壳动力学 ,地震孕育和破裂过程的理论、模拟试验和实际观测 ,数据同化和计算软件的开发应成为今后研究的重点。现在的问题是 ,需要积极借助数值天气预报的经验 ,强化多学科 ,多部门的组织协调 ,尽早在有条件的地区开展地震动力学数值预报的科学试验。地震数值预报研究必将极大地促进中国地震科学基础研究和地震预报的进一步发展。     

安塬水源地地下水流三维数值模拟

安塬水源地为干河冲洪积扇型水源地,含水层由第四系冲洪积层构成,厚15～30m,非均质,各向同性.受含水层底板坡降的影响,潜水面坡降较大,为5％～8％.三维流特征明显,尤其是抽水试验时在抽水井周边区域不同深度的水头差别更为显著.为了观测不同深度的水位,在群孔抽水试验时在同一观测孔中按上、中、下不同深度设置了3组观测管分别进行水位观测,从而为三维数值模拟提供了水位数据,并以该资料为依据建立了三维有限差分模型,对6种开采方案进行了数值模拟计算和预报.通过11a的开采,证明建立的三维有限差分模型可靠,为水源地的地下水开采建立了较为可靠的数学预报模型.     

Modified Weibull-derived spectrum for deep water significant wave height estimation

The modified Weibull spectrum is utilized to calculate the zeroth spectral moment (m_o) using Monte Carlo integration methods. Then significant wave height (H_s) is calculated using the formula This is validated with observed buoy data and numerical wave model (WAM) predicted significant wave heights. The Weibull parameters have been calculated using energy densities from observed spectra recorded by DS5 buoy (13.80° N, 82.52° E, depth 3355.48 m) by the method of maximum likelihood (MLE).The relative root mean square error (RRMS) and relative bias error criteria show that modified Weibull spectrum estimated significant wave heights are better than those predicted by WAM model. The monthly averaged observed wave power spectra for the year 2005 recorded by deep water buoy DS5 is considered in this work. The spectra exhibit bimodal sea states for several months of the year.     

Estimation of wave period statistics using numerical coastal wave model

We used field and model wave data to investigate that zero crossing and average wave period distribution follow Gamma distribution. Since Gamma CDF is an infinite power series, further mathematical treatment is difficult. Hence its shape parameter is approximated to the nearest integer to arrive at Erlang distribution. An expression is derived from Erlang distribution to estimate various mean wave periods and significant wave period and validated by relative root mean square (RRMS) error criteria. It is shown by mathematical logic that the significant wave period distribution follows Erlang (or Gamma) distribution and is validated. The average of one-third and one-tenth highest wave periods (T _s ) estimated from Erlang distribution are in accordance with the values computed from recorded buoy and numerical coastal wave model wave period data. The predicted T _s values from coastal wave model wave period data underestimates the values from buoy wave period data.     

Application of Weibull model for redefined significant wave height distributions

It is well accepted that the parent distribution for individual ocean wave heights follows the Weibull model. However this model does not simulate significant wave height which is the average of the highest one-third of some ‘n’ (n- varies) wave heights in a wave record. It is now proposed to redefine significant wave height as average of the highest one-third of a constant number (n-constant, say,n = 100) of consecutive individual wave heights. The Weibull model is suggested for simulating redefined significant wave height distribution by the method of characteristic function. An empirical support of 100.00% is established by Χ²-test at 0.05 level of significance for 3 sets of data at 0900, 1200 and 1500 hrs at Valiathura, Kerala coast. Parametric relations have been derived for the redefined significant wave height parameters such as mean, maximum one-third average, extreme wave heights, return periods of an extreme wave height and the probability of realising an extreme wave height in a time less than the designated return period.     

Parametric wave prediction model based on time delay concept — An evaluation

In the early stages of wave growth it is seen that wave heights are underestimated by presently available models especially in a low wind regime. Parametric wind-sea relationships of significant wave height (H₂) and zero-crossing period (T ₂) for slight to moderate sea-states were proposed earlier on an analysis of wind and wave data. This model is based on the concept of time delay between the wind speed (U) and wave evolution process. It is simple and requires less computational effort compared to the spectral method. The present paper attempts to test and evaluate the performance of the proposed model with additional field data of wind and waves measured off the Indian coast. MeasuredU,H ₂, andT ₂ ranged between 1 and 15 m/s, 0·5 and 2·7 m and 4 and 10 s respectively. By and large, the comparison between model output and field observations are encouraging. A hindcast study was carried out earlier using a spectral wave prediction model (TOHOKU) for Indian Seas using field measurements which include the data sets utilized in this study. Comparison between these two models reveals a good agreement.     

Demarcation of inland vessels’ limit off Mormugao port region,India: a pilot study for the safety of inland vessels using wave modelling

The Ministry of Shipping desires to revise the inland vessels’ limit (IVL) notification based on scientific rationale to improve the safety of vessels and onboard personnel. The Mormugao port region extending up to the Panaji was considered for this pilot study. Measured winds and wave parameters (AWS and moored buoy) as well as NCEP re-analysis and NCMRWF winds were used for the analysis and input to regional and local models. The results of wave model were validated with measured significant wave heights (SWHs) and the comparison shows a good match. The analysis indicates that SWHs do not exceed 2.0 m during non-monsoon months, and in monsoon months exceed 5.0 m, and even 7.0 m, especially during extreme events. In order to draw IVL contours for Goa coastal region, local model was set up and nearshore waves were simulated for the period May 2004–May 2005. Based on the nearshore SWH distribution, IVL contours have been fixed for the Mormugao port and Panaji coastal regions.     

Optimal interpolation of buoy data into a deterministic wind–wave model

Third-generation wave models have been evolved in 1980s with the state-of-the-art physics of wave generation. Using these models, the real time wave estimation is made possible but, in general, it is found to be underpredicted. This is mainly due to the smoothened wind vectors from the atmospheric model. An accurate prediction of wind is thus necessary to improve the wave prediction further. A better way of overcoming the discrepancies in the wind is by the way of wave data assimilation. In the present study, an operationally efficient yet a versatile assimilation model, optimal interpolation (OI), has been presented. The weighting matrix, so-called gain matrix, has been formulated according to the model physics by which the wind generates waves. The efficiency of the assimilative model using real time buoy observations at the Arabian Sea has been evaluated and described in this article. The root mean square error reduction of wave height is found to be of the order of 30–50% at the validation stations.     

Applicability of artificial neural networks for obtaining velocity models from synthetic seismic data

Seismic velocity analysis is a crucial part of seismic data processing and interpretation which has been practiced using different methods. In contrast to time consuming and complicated numerical methods, artificial neural networks (ANNs) are found to be of potential applicability. ANN ability to establish a relationship between an input and output space is considered to be appropriate for mapping seismic velocity corresponding to travel times picked from seismograms. Accordingly a preliminary attempt is made to evaluate the applicability of ANNs to determine velocity and dips of dipping layered earth models corresponding to travel time data. The study is based on synthetic data generated using inverse modeling approach for three earth models. The models include a three-layer structure with same dips and same directions, a three-layer model with different dips and same directions, as well as a two-layer model with different dips and directions. An ANN structure is designed in three layers, namely, input, output, and hidden ones. The training and testing process of the ANN is successfully accomplished using the synthetic data. The evaluation of the applicability of the trained ANN to unknown data sets indicates that the ANN can satisfactorily compute velocity and dips corresponding to travel times. The error intervals between the desired and calculated velocity and dips are shown to be acceptably small in all cases. The applicability of the trained ANN in extrapolating is also evaluated using a number of data outside of the range already known to ANN. The results indicate that the trained ANN acceptably approximates the velocity and dips. Furthermore, the trained ANN is also evaluated in terms of capability of handling deficiency in input data where acceptable results were also achieved in velocity and dip calculations. Generally, this study shows that velocity analysis using ANNs can promisingly tackle the challenge of retrieving an initial velocity model from the travel time hyperbolas of seismic data.     

Interaction of monsoonal wave, current and tide near Gopalpur, east coast of India, and their impact on beach profile: a case study

A 97-day-long record on waves and currents was obtained using wave rider buoy and current meter moored at 2.5 km off Gopalpur from 19 May to 23 August 2008 representing southwest monsoon months. A Valeport tide gauge was used to record water level at Gopalpur port. Simultaneously, beach profiles at 4 transects were monitored using real-time kinematic (RTK) global positioning system (GPS). A total of 636,167 waves were analyzed for the period; a range of 3,200–9,700 waves approach the coast in an individual day. During the study, unusual characteristics of wave were observed on July 29, 2008, with a magnitude of significant wave height, H_s = 2.85 m, maximum wave height, H_max = 5.22 m, and peak wave period, T_p = 10.2 s, and on August 11, 2008, with H_s = 2.28 m, H_max = 5.37 m, and T_p = 11.1 s. Significant beach loss was noticed during these periods, and severe erosion was recorded on August 1, 2008. Beach profile data indicates that 18–58 cu. m/m sediment was lost during the study period. The paper provides an overview of the statistical analysis of wave heights, periods, direction, and spectral energy density and explains the cause of coastal erosion and loss of sediment.     

Sea State Hindcast with ECMWF Data Using a Spectral Wave Model for Typical Monsoon Months

A third generation wave prediction model (WAM) suitably modified for the Indianseas was used to simulate the sea state during typical monsoon months of July 1987,88 and 89. During July 1987, small-scale disturbances over the Arabian Sea and theBay of Bengal as reported from earlier works were noticed in the wave hindcast. Thegridded monthly averaged wave heights for the above periods were compared with theGEOSAT altimeter data. A comparison between model computed wave parameters andavailable ship observations was also carried out for July 1987, 88 and 89.The modelwas run with ECMWF analyzed wind field for a period of one month. The comparisonshowed that the wave heights from the model matches with those from GEOSATAltimeter and ships of opportunity.     

径流长期预报的人工神经网络方法

提出径流长期预报的人工神经网络方法。运用神经网络的一典型模型──“反向传播”模型,以大伙房水库在补水期的径流状况作为研究对象,尝试了神经网络方法的效果。结果表明,该方法预报成功率较高,容错能力较强,可望成为径流长期预报的有效的辅助手段。