地理空间情报预测分析的一种方法

地理空间情报是对影像和地理空间信息的发掘和分析。其中,地理空间情报预测分析是决策的基础。传统的地理空间情报预测分析大多是基于逻辑思维的定性研究,此处提出了一个地理空间情报预测分析的模型——因果关联图,同时给出了计算预测事件可信度的方法。该方法能够定量、简捷、快速、有效地实现地理空间情报预测分析。     

基础理论与预报实践

总结了自2009年秋开始的全国预报员轮训的教学工作,强调基础知识、天气实践和长期积累的重要性.具体总结了T-logp图和静止卫星云图分析应用及准地转理论教育中的基础性问题,并以暴雨预报思路为例讨论了如何将复杂问题简单化.最后就Q矢量、螺旋度和等熵位涡三个问题讨论了把简单问题复杂化所引起的后果.     

Selecting the appropriate hydraulic model structure using low-resolution satellite imagery

This study aims at investigating the value added by an explicit representation of floodplain processes in hydraulic models, by comparing the results of a pure 1D model to the ones of an integrated 1D-2D model. The analysis is performed on a 96 km reach of the River Po (Italy). For this test site, two different model structures (1D and 1D-2D) are implemented and compared. The models are first calibrated using high water marks of a high magnitude event and then validated by means of a coarse resolution space-borne flood extent map of a low-magnitude event. The explicit inclusion of 2D features strongly improves the performance of the model, both in terms of accuracy and robustness. The possible causes of this behaviour are investigated and some statements are drawn based on the analysis of the features of the two flood events. Some general conclusions on the effect of the choice between 1D and 2D models arise from the analysis of the calibration-validation procedure. In this context a new measure of fit is proposed, to be used if binary (wet/dry) observations are available and is proven to be able to better distinguish among different sets of model parameters. Lastly, the study also shows the ability of timely low resolution satellite imagery to assist the selection of the most appropriate hydraulic model structure.     

Climate change impacts on pricing long-term flood insurance: A comprehensive study for the Netherlands

Recently long-term flood insurance contracts with a duration of 5, 10 or 15 years have been proposed as a solution for covering flood risk and mitigating increasing flood losses. Establishing a long-term relation between the policyholder and the insurer can provide better incentives to reduce risk through undertaking damage mitigation measures. However, the uncertainty about the development of future flood risk in the face of climate and socio-economic change may complicate insurers’ rate-setting of long-term contracts. This issue has been examined in this study by estimating the effects of these changes on flood risk and pricing flood insurance premiums of short- and long-term flood insurance contracts in all (53) dike-ring areas in the Netherlands. A broad range of simulations with hydrological and flood damage models are used to estimate the future development of flood risk and premiums. In addition, the long-term development of insurance funds is estimated with a spatial “Climate Risk Insurance Model (CRIM)” for a private insurance arrangement and for a ‘three-layered’ public-private insurance program. The estimation of flood insurance premiums of long-term insurance contracts reveals fundamental problems. One is that there is an incentive for either the consumer or the insurer to prefer short-term rather than long-term contracts in the face of climate-related uncertainty. Therefore, it seems advisable to examine the introduction of one-year flood insurance contracts in the Netherlands, at least until the large uncertainties with climate and socio-economic change on flood risk have been resolved. The estimations performed with the Climate Risk Insurance Model indicate that a private insurance fund could have difficulties with building up enough financial reserves to pay for flood damage, while the layered public-private insurance scheme is more robust.     

基于配料法的“08.6.3”河南强对流天气分析和短时预报

采用基于构成要素的预报方法(配料法),使用业务中能够及时获取的常规观测资料、静止气象卫星云图和多普勒天气雷达资料,对2008年6月3日下午发生在河南大部的以雷暴大风为主的强对流天气进行分析.针对2008年6月3日08时500、700、850 hPa和地面图上温、湿、风等的配置,结合探空、水汽云图和多普勒天气雷达回波图,对构成强对流天气的几个主要因素:层结稳定度、水汽、触发机制、垂直风切变、对流层干层等进行分析,结果表明:河南大部处于较强的条件不稳定、强烈的深层(地面到500 hPa)垂直风切变和明显的高空干平流的环境中,这种环境有利于高度组织性的超级单体风暴的产生和发展,导致雷暴大风和冰雹天气.而地面冷锋、辐合线和高空次天气尺度涡旋伴随的正涡度平流共同作用产生的上升气流导致雷暴在山西中南部生成,然后向东南方向移下太行山.进入河南有利的环境下获得猛烈发展.在充分利用"配料法"的主要思路和现有观测资料条件下,结合预报经验,2008年6月3日强对流过程可以提前数小时进行正确的短时预报.     

有害藻华的预测技术和防灾减灾对策研究进展

有害藻华给海洋生物、海水养殖业、海洋环境和公共健康安全造成了极大的威胁,已经成为世界性的海洋环境灾害之一.近几十年,我国近海有害藻华的发生频率和发生规模不断扩大,给国家经济发展和公共医疗卫生造成了极大的经济损失和不利影响.主要介绍了国内外常用的有害藻华预测技术,并参考国内外有效的有害藻华防治措施,为我国有害藻华的防灾减灾工作提出建议.我国亟需逐步建立有害藻华综合预警报业务体系和完备的应急减灾体系,才能进一步提升有害藻华等海洋环境灾害的应对能力.     

新安江模型参数多目标优化研究

水文模型的参数优化率定一直以来是水文预报领域的重要研究内容,当水文模型的结构确定后,水文模型参数的选择对水文模型整体性能和水文预报结果的好坏有着至关重要的影响.针对传统水文模型参数优选采用单一目标不能充分全面挖掘水文观测资料中蕴含的水文特征信息的缺陷,本文以新安江三水源模型为例,尝试采用多目标优化算法优化率定水文模型,算例应用分析表明,通过合理的选择目标函数的种类和数目,采用多目标进化算法优化率定模型参数,可以获得相对于单目标率定模型参数更优的结果.进一步,研究工作针对模型参数优化的结果进行分析,可以明显看出模型参数优化中存在“异参同效”现象,为后续模型参数不确定性分析等相关研究工作的开展做好了铺垫.     

基于神经网络的规则提取及其渔业应用研究

为了解决远洋渔业中过度依赖经验而产生的盲目捕捞问题,结合海洋环境数据和历史产量数据对渔场进行有效分析,提出了一种基于径向基函数神经网络(Radial basis function neural network,RBFNN)的栖息地指数(HSI)预测方法,并将其应用于印度洋海域大眼金枪鱼(Thunnus obesus)栖息地指数的预测。在RBFNN训练过程中使用模糊C均值(Fuzzy c-means,FCM)聚类算法,在基于神经网络的规则提取过程中首次采用了和声搜索(Harmony search,HS)算法。实验研究表明,利用FCM改进后的RBFNN,均方误差(Mean square error,MSE)达到0.021 6。和声搜索由于算法简单,易于实现,能够应用于训练后的FCM-RBFNN提取分类规则,提取出的规则能够反映该渔业现状。     

基于20—30d振荡的长江下游地区夏季低频降水延伸期预报方法研究

用长江下游降水低频分量和环流低频主成分,构造多变量时滞回归模型(MLR)和主成分复数自回归模型(PC-CAR)的混合预报模型(MLR/PC-CAR),对长江下游降水低频分量进行延伸期逐日变化预报,延长预报时效。通过2011年6—8月预测试验表明,20—30 d时间尺度的长江下游低频降水预测时效可达50 d左右,采用南半球中高纬度地区850 hPa低频经向风的主成分作为预测因子的模型的预测精度明显高于东亚地区低频经向风作为预测因子的模型。这表明在20—30 d时间尺度上,长江下游降水与南半球中纬度绕球遥相关(SCGT)型有关的主分量的时滞相关更加密切。进一步对于较强20—30 d振荡的多年资料构建的MLR/PC-CAR混合模型预测试验表明,SCGT是预测夏季长江下游低频降水未来50 d变化的显著信号。基于SCGT的发展和演变,对于把握类似长江下游地区2011年6月初旱涝急转和7月中旬持续降水和强降水过程异常变化过程很有帮助,SCGT可以作为夏季长江下游20—30 d低频降水和强降水过程进行延伸期预报的主要可预报性来源之一。     

相似集合预报技术在粤西风场预报中的应用

利用ERA5再分析资料与常规站点观测资料,采用相似集合法(AnEn)对粤西地面风的日平均风速与日最大风速释用进行订正.结果表明:日平均风速与日最大风速订正后的准确性得到明显提高,且在地形复杂的山区也有较好的订正效果.日平均风速的逐日均方根误差总体小于日最大风速,且更为稳定.对比相似成员集合性能,日平均风速各成员也比日最...