双校正模式下的大清河流域陆气耦合洪水预报研究

基于数值大气模式WRF、三维变分数据同化WRF-3DVar、河北雨洪模型以及实时校正模型ARMA,在北方半湿润半干旱地区的大清河流域构建了陆气耦合洪水预报系统,并利用2012、2013年发生的3场降雨洪水,对系统的降雨洪水预报结果进行分析。结果表明:雷达反射率与GTS数据的同时同化,可有效改善数值大气模式对中小尺度流域的降雨预报效果,从而降低系统的洪水预报误差,ARMA模型的应用,能够进一步提升系统的洪水预报精度,随着预见期的延长,系统的预报精度下降,但系统在6h预见期内仍表现出较好的应用效果。因此,在数据同化和实时校正的"双校正"模式下,陆气耦合洪水预报系统在延长洪水预报预见期的同时,具有较高的洪水预报精度,具有一定的应用前景。     

基于多模式降水集成的陆气耦合洪水预报

针对基于单模式陆气耦合模型在洪水预报中存在的不确定性问题,利用三种数值天气预报模式MC2、GEM与T213开展基于多模式降水集成的陆气耦合洪水预报研究。通过淮河流域2007～2009年3场暴雨洪水验证表明:基于多模式降水集成的陆气耦合洪水预报可以有效地减小不确定性、提高精度和稳定性;但基于多模式降水集成的陆气耦合洪水预报对小尺度局地强降水仍存在较大的不确定性,今后需对此展开进一步的研究,以给出更加稳定、可靠的洪水预报结果。     

陆气耦合模型在实时暴雨洪水预报中的应用

采用加拿大区域性中尺度大气模式MC2(Canadian Mesoscale Compressible Community)和新安江模型单向耦合模型系统,对2005年7月4～15日发生在淮河流域的一场暴雨洪水,进行了实时预报.采用王家坝以上流域的实测降水和王家坝断面的实测洪水资料,对MC2预报降水的时空分布和陆气耦合模型预报的洪水过程进行了分析.结果表明,MC2对该场强降水过程具有很好的预报能力,陆气耦合模型有效地增长了洪水预报的预见期,具有很好的应用前景.     

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。     

基于分布式水文模型的中小河流洪水预报技术

中小河流大多位于资料短缺的山丘区,洪水具有突发性强、汇流时间快、预见期短以及分布广的特点。中小河流洪水预报首要目的和任务是预警预报,预报方式应以自动预报为主,以实现及时预警,最大程度地避免人员伤亡,减轻灾害损失。本文分析了中小河流洪水预报的特点与难点,提出了中小河流洪水预报的思路及实用预报模型与方法,开展了基于分布式水文模型的中小河流洪水预报技术在新安江上游屯溪流域预警预报中的应用研究,以期为当前所开展的全国中小河流洪水预警预报系统建设提供参考。研究表明:分布式水文模型是资料短缺地区中小河流洪水预报的有效方法,基于分布式水文模型的洪水预报技术能够满足中小河流洪水自动预警预报的生产需要。     

广西河流预报新模式及关键技术研究与应用

随着全国中小河流水文监测工程项目的实施以及国家、广西水情预警发布管理办法的出台,在洪涝频繁、中小河流站点点多面广的新形势下,如何实现广西河流预报任务由传统的预报增加至预警、预报双重要求成为近年来工作中迫切要解决的问题。基于水量平衡原理及洪水水位涨率变化规律分析研究,根据“预测-预警-预报”河流预报模式,研究开发了适应此模式的河流预报方法和技术。2015年以来,此模式及方法技术广泛应用于广西河流预报中,实现大江大河、中小河流洪水科学预测预警预报,水文预报合格率均超过90%。同时有效延长洪水预见期,大江大河洪水预见期平均提前24h,中小河流洪水预见期平均提前5h。     

2008年三峡水库入库洪水特点与预报效果评价

对2008年三峡水库入库洪水特点进行分析,并针对场次洪水和连续自动预报两种方式进行预报误差统计,两种方式预报结果都表明2008年实时联机水文预报精度达到水文情报预报规范的甲级标准,同时发现未来降雨对三峡水库入库预报精度影响较大,部分场次洪水结合(不结合)未来降雨的预报精度差别很大,且随着预见期增长,预报误差也加大。     

中小河流洪水预报研究进展

中小河流洪水预报是当前国际水文科学中亟待研究的重大科学问题之一。由于中小河流存在分布广、降水及下垫面空间异致性强、产汇流时间短、突发性强、水文资料欠缺等特点,因此预报精度较低,预见期短,预报难度大。目前国内外学者对中小河流洪水预报开展了一定的研究,主要集中在中小河流洪水形成机理、洪水预报模型、缺资料中小河流水文模型参数确定方法、洪水预报耦合降水预报等方面。对中小河流洪水预报相关研究进展和存在的问题进行了总结,并指出未来在多源信息高效融合、精细化洪水模拟、高精度降水预报等方面还应进行更深入的研究。     

TIGGE降雨信息在柴河流域洪水预报中可利用性评估

基于全球多模式集合预报(TIGGE)资料,以柴河流域为研究区域,采用TS评分、Brier评分和Talagrand分布等方法对欧洲中期天气预报中心(ECMWF)、美国国家环境中心(NCEP)和英国气象局(UKMO)三个中心集合预报的6h、12h、24h短期降雨量进行定量评估和对比,并分别以实测降雨和NCEP预报降雨驱动新安江模型模拟洪水过程,据此探讨了集合降雨预报的可利用性。得到两个重要结论:ECMWF、NCEP和UKMO对低雨量级降雨的预报效果较好,各雨量级的预报效果有随预见期增长而增加的趋势,且普遍存在空报率较高、漏报率偏低的情况;集合降雨预报信息可应用于新安江模型进行洪水预报,并能够有效的延长洪水预报的预见期。研究成果可在适当条件下推广应用至其它流域的洪水预报作业中。     

基于深度学习的雷达降雨临近预报及洪水预报

为探究深度学习的雷达降雨临近预报在流域洪水预报中的适用性,采用U-Net、嵌入注意力门的Attention-Unet和添加转换器的多级注意力TransAtt-Unet开展雷达降雨临近预报,将预报降雨作为HEC-HMS水文模型的输入,对柳林实验流域进行洪水预报。结果表明：1 h预见期时,Attention-Unet对短时强降雨预报结果较好,TransAtt-Unet预报降雨模拟的洪峰流量和径流量相对误差小于20%,各深度学习模型对量级较大的降雨和洪水预报精度较高;2 h预见期的预报降雨强度、降雨总量、洪峰流量和径流量存在显著低估,U-Net能取得相对较好的降雨预报结果。基于深度学习的1 h预见期雷达降雨临近预报及洪水预报可为流域防洪减灾提供科学依据。     

海河流域河川径流对气候变化的响应机理

利用可变下渗容量(Variable Infiltration Capacity,VIC)模型,在海河流域选取了6个典型流域来率定VIC模型的参数。通过模型参数移植技术,建立了全流域的径流模拟平台。根据假定的气候变化情景,分析了海河流域河川径流对气候变化的响应机理。结果表明:在年平均气温升高2℃时,海河流域的径流量将减少6.5%;当年降水量增加或者减少10%时,海河流域的径流量将分别增加26%和减少23%;当汛期降水占年降水量的比例分别增加或者减少10%时,全流域的径流量将会增加12%或者减少7%;在空间上,在年平均气温升高和年降水量变化的情景下,海河流域西北部的河川径流比东南部更敏感;在降水年内分配变化的情景下,海河流域东南部的河川径流比西北部更敏感。总体上,年降水量越大,径流量对降水量的敏感性越小,对平均气温的敏感性也越小,而对降水年内分配的敏感性越大。     

利用WRF3D-Var同化多普勒雷达反演风场试验研究

为了将C波段雷达风场资料更好地应用于数值预报模式中,利用两步变分法反演多普勒雷达风场资料,并处理成标准的常规探空资料,以WRF模式及其三维变分同化系统为平台,针对2013年6月19日发生在天水的一次强暴雨过程进行同化雷达反演风的试验研究.试验结果表明:同化雷达反演风场后,对降水预报的改进能维持12h,尤其同化雷达反演风场后3~9h效果非常显著;0~3h作用不是很明显;9~12h预报具有一定的正作用.另外,循环同化比同化一次效果好,但并不是同化次数越多越好.因此,同化C波段雷达反演风场后,对降水预报具有一定的正作用.     

Hydroclimatological Perspective of the Kerala Flood of 2018

Flood is among the deadliest disasters in India, and the frequency of floods and extreme precipitation events is projected to increase under the warming climate. The frequency of floods in India varies geographically as some regions are more prone to floods than the others. The Kerala flood of 2018 caused enormous economic damage, affected millions of people, and resulted in the death of more than 400 people. Here we provide a hydroclimatological perspective on the Kerala flood of 2018. Using the observations and model simulations from the Variable Infiltration Capacity (VIC) model, we show that the 2018 extreme precipitation and runoff conditions that caused flooding were unprecedented in the record of the past 66 years (1951–2017). Our results show that mean monsoon precipitation has significantly declined while air temperature has significantly increased during 1951–2017 in Kerala. The drying and warming trends during the monsoon season resulted in a declined total runoff in large part of the state in the last 66 years. Apart from the mean hydroclimatic conditions, extreme precipitation, and extreme total runoff have also declined from 1951 to 2017. However, 1 and 2-day extreme precipitation and extreme runoff conditions in August 2018 exceeded substantially from the long-term 95^th percentiles recorded during 1951–2017. Since there is no increase in mean and extreme precipitation in Kerala over the last six decades, the extreme event during August 2018 is likely to be driven by anomalous atmospheric conditions due to climate variability rather anthropogenic climate warming. The severity of the Kerala flood of 2018 and the damage caused might be affected by several factors including land use/land cover change, antecedent hydrologic conditions, reservoir storage and operations, encroachment of flood plains, and other natural factors. The impacts of key drivers (anthropogenic and natural) on flood severity need to be established to improve our understanding of floods and associated damage.     

基于灰色理论的西藏暴雨洪涝预测

以西藏38个气象站点1985-2014年实测降水量为基础数据,应用灰色预测模型和长序列趋势拟合方法预测2015-2019年的年降水量和季节降水量,经2015-2017年的实测降水数据验证,其结果良好,最后以调整后的降水距平百分率为洪涝指标,分析2018年、2019年的洪涝预测情况。结果表明,西藏2018年降水集中在东南部地区的林芝、米林、波密和察隅一带, 2019年降水主要集中在芒康地区,季节上春季降水主要集中在雅鲁藏布江大拐弯地区,夏季主要在索曲河、怒江上游地区,秋季主要在嘉黎、林芝、波密一带,冬季降水集中在聂拉木地区。2018年和2019年洪涝易发生在春季和冬季,洪涝情况都相对正常,大部分县区为无涝或正常状态,少数县区为轻微洪涝,极少数县区为中涝或特涝, 2019年洪涝情况相较于2018年整体偏多,其中在2018年的拉孜县和2019年的拉萨城关区,极有可能出现特涝灾害。     

Evaluation of real-time flash flood forecasts for Haiti during the passage of Hurricane Tomas, November 4–6, 2010

The January 2010 earthquake that devastated Haiti left its population ever more vulnerable to rainfall-induced flash floods. A flash flood guidance system has been implemented to provide real-time information on the potential of small (~70 km²) basins for flash flooding throughout Haiti. This system has components for satellite rainfall ingest and adjustment on the basis of rain gauge information, dynamic soil water deficit estimation, ingest of operational mesoscale model quantitative precipitation forecasts, and estimation of the times of channel flow at bankfull. The result of the system integration is the estimation of the flash flood guidance (FFG) for a given basin and for a given duration. FFG is the amount of rain of a given duration over a small basin that causes minor flooding in the outlet of the basin. Amounts predicted or nowcasted that are higher than the FFG indicate basins with potential for flash flooding. In preparation for Hurricane Tomas’ landfall in early November 2010, the FFG system was used to generate 36-h forecasts of flash flood occurrence based on rainfall forecasts of the nested high-resolution North American Model of the National Centers for Environmental Prediction. Assessment of the forecast flood maps and forecast precipitation indicates the utility and value of the forecasts in understanding the spatial distribution of the expected flooding for mitigation and disaster management. It also highlights the need for explicit uncertainty characterization of forecast risk products due to large uncertainties in quantitative precipitation forecasts on hydrologic basin scales.     

区域气候模式在中国西北地区气温和降水长时间序列模拟的误差分析

气温和降水微小的变化都可能引起中国西北地区脆弱生态环境空间格局剧烈变化.由于西北地区气象站点分布稀疏,站点观测结果难于完全代表西北地区复杂地貌的气象要素空间分布格局.区域气候模式模拟可以弥补现有观测资料的不足,但模式模拟存在的误差,往往制约着西北区域气候变化评估工作.本文在CN05格点数据和站点数据基础上,从空间格局、极值模拟等方面对高分辨率WRF模式数据气温和降水模拟精度进行评估,假定偏差恒定基础上,使用分类回归树模型建立偏差预测模型.结果表明:CN05数据和WRF模式多年年平均气温和多年平均年降水数据空间格局基本一致,但WRF模式结果空间格局更加细致.WRF模式结果在山区和湖泊地区降水偏高估,气温和降水距平在新疆地区相对于其他地区变化较为一致;95%分位数的降水在甘肃和新疆地区模拟结果好于宁夏、青海和陕西;5%和95%分位数气温,WRF模式大部分站点表现为暖偏差,而CN05数据表现为冷偏差;不同的地貌单元区域气温和降水的偏差时间序列存在一定的相关性.在西北五省区气温和降水的分类树模型中,由于高程和地形复杂度对偏差预测模型影响较大,无法在整个西北地区建立统一的误差预测模型.     

Flood forecasts based on multi-model ensemble precipitation forecasting using a coupled atmospheric-hydrological modeling system

The recent improvement of numerical weather prediction (NWP) models has a strong potential for extending the lead time of precipitation and subsequent flooding. However, uncertainties inherent in precipitation outputs from NWP models are propagated into hydrological forecasts and can also be magnified by the scaling process, contributing considerable uncertainties to flood forecasts. In order to address uncertainties in flood forecasting based on single-model precipitation forecasting, a coupled atmospheric-hydrological modeling system based on multi-model ensemble precipitation forecasting is implemented in a configuration for two episodes of intense precipitation affecting the Wangjiaba sub-region in Huaihe River Basin, China. The present study aimed at comparing high-resolution limited-area meteorological model Canadian regional mesoscale compressible community model (MC2) with the multiple linear regression integrated forecast (MLRF), covering short and medium range. The former is a single-model approach; while the latter one is based on NWP models [(MC2, global environmental multiscale model (GEM), T213L31 global spectral model (T213)] integrating by a multiple linear regression method. Both MC2 and MLRF are coupled with Chinese National Flood Forecasting System (NFFS), MC2-NFFS and MLRF-NFFS, to simulate the discharge of the Wangjiaba sub-basin. The evaluation of the flood forecasts is performed both from a meteorological perspective and in terms of discharge prediction. The encouraging results obtained in this study demonstrate that the coupled system based on multi-model ensemble precipitation forecasting has a promising potential of increasing discharge accuracy and modeling stability in terms of precipitation amount and timing, along with reducing uncertainties in flood forecasts and models. Moreover, the precipitation distribution of MC2 is more problematic in finer temporal and spatial scales, even for the high resolution simulation, which requests further research on storm-scale data assimilation, sub-grid-scale parameterization of clouds and other small-scale atmospheric dynamics.     

Simulation Study on Precipitation Recycling Ratio in the Tibetan Plateau from 1982 to 2005

A dynamical downscaling approach using a regional climate model WRF (Weather Research and Forecasting Model Vision 3.5) driven by a global climate model CCSM4 (The Community Climate System Model Version 4) was adopted, and the downscaling results for the historical period (1982-2005) were evaluated for annual mean precipitation rate and evaporation rate over the Tibetan Plateau (TP). Furthermore, the spatial distribution and seasonal variation characteristics of Precipitation Recycling Ratio (PRR) simulated by CCSM4 and WRF were analyzed with the QIBT (Quasi-isentropic Back-trajectory method). The results show that the historical spatial distributions of annual mean precipitation rate and evaporation rate over the TP were found to better reproduce in the dynamical downscaling modeling compared to its coarse-resolution forcing. The PRR of the TP is 32% simulated by WRF, with a higher PRR in the wet season and a lower PRR in the dry season for the river basins in the northern TP, but the opposite seasonal variation was found for the river basins in the southern TP. In addition, the different land covers over the TP are more precisely represented in the WRF model, the PRR of grassland, shrubland and sparsely vegetation is higher than that of other land cover types.     

概化的Tank模型及其在龙羊峡水库汛期旬平均入库流量预报中的应用

Tank模型（又称为水箱模型），是一种用于流域径流预报的确定性水文模型，根据龙羊峡水库入库主要产流区－－黄河上游唐乃亥水文站以上流域下垫面条件下产汇流特性，将其概化为以降雨量为输入，径流量为输出的单孔出流的线性水箱，工用于该水库汛期旬平均入库流量的预报，经对历史资料进行拟合和试验预报的 结果表明，该模型具有较高的预报精度，现已应用于黄河上游龙羊峡水库汛期旬平均入库来水量的中期预报中，取得了十分显著的经济效益。