土壤侵蚀研究中降雨过程随机模拟综述

土壤侵蚀是中国严峻的环境问题之一。土壤侵蚀模型是诊断和防治土壤侵蚀问题的有力工具。降雨随机模拟模型可以弥补观测资料在时空尺度上的不足,满足土壤侵蚀模型对降雨过程数据的需求。论文总结了降雨过程基本特征、随机模拟的研究进展以及未来的发展方向。主要结论有：① 最小降雨间歇(Minimum Inter-event Time, MIT)为分割次降雨的最小时间间隔,当干期小于该临界值,则合并为1次降雨;否则分割为2次独立的次降雨事件。采用指数方法计算得到中国中东部最小降雨间歇变化于7.6~16.6 h 之间,平均值为10.7 h;当MIT值较小时,次降雨过程参数如次雨量、历时、平均雨强与峰值雨强等对MIT的变化敏感。② 降雨的时程分配特征反映次降雨量在降雨过程中的分配,是降雨过程随机模拟的重要方面。采用Huff雨型分析方法得到中国降雨以峰值出现在前期的降雨为主,峰值出现在前1/2时段的降雨占65.1%;峰值出现在前期的降雨事件,与峰值出现在后期的降雨事件相比,历时较短、雨强较大。③ 美国农业部水蚀预报项目WEPP中自带的天气随机模型CLIGEN能较好地模拟日雨量,整体上低估次降雨历时,高估峰值雨强;且对于不同大小的日雨量等级,次降雨历时和峰值雨强的偏差方向和程度不一致。④ CLIGEN的参数输入需要降雨过程观测资料,难以获取,小时降雨观测资料相对更易获取,基于此,发展了使用小时降雨数据计算CLIGEN中2个与降雨过程相关的参数TimePk和MX.5P的方法。未来需加强随机模型对极值的模拟能力,以及建立多站点、多气象要素相关的天气条件约束型随机模型。     

不同概率分布函数降雨极值的适用性分析

极端降雨极值发生的重现期是流域与城市防洪设施规划设计标准需要参考的最重要参数之一。利用常用的5种水文统计学分布函数,选取中国十大流域内10个站点不同时段的最大降雨极值序列进行拟合,并检验筛选不同站点的适用性分布函数。结果表明：10个站点拟合优度检验拟合效果较好,曲线差异度较小的分布依次为广义极值分布、对数正态分布、皮尔逊III分布;不同站点适宜性曲线的差异程度不同。研究结果可为区域降雨极值序列的拟合提供参考,即不同的区域、不同的季节、不同时长的降雨极值序列都应寻找其较适宜的分布函数并采用多种检验方法来拟合,以降低不确定性。     

不同概率分布函数降雨极值的适用性分析北大核心CSCD

极端降雨极值发生的重现期是流域与城市防洪设施规划设计标准需要参考的最重要参数之一。利用常用的5种水文统计学分布函数,选取中国十大流域内10个站点不同时段的最大降雨极值序列进行拟合,并检验筛选不同站点的适用性分布函数。结果表明:10个站点拟合优度检验拟合效果较好,曲线差异度较小的分布依次为广义极值分布、对数正态分布、皮尔逊III分布;不同站点适宜性曲线的差异程度不同。研究结果可为区域降雨极值序列的拟合提供参考,即不同的区域、不同的季节、不同时长的降雨极值序列都应寻找其较适宜的分布函数并采用多种检验方法来拟合,以降低不确定性。     

基于站点降雨量最优拟合函数的SPI指数计算

标准化降水指数(SPI)能较好地反映不同时间尺度和区域的旱涝状况,时空适应性较强,成为科学研究和实际应用最广泛的描述多时间尺度的气象干旱指标,采用不同的概率分布函数拟合降雨量系列,将会得到不同的SPI值。利用淮河流域30个气象站点1951-2013年间约60 a的逐月降雨资料,分别采用7种不同分布函数(Gamma分布、皮尔逊III型分布、正态分布、对数正态分布、广义Pareto分布、广义Logistic分布和Weibull分布)对各站点不同时间尺度(1、3、6、12和24 mon)各月份降雨系列进行拟合,以K-S方法检验了各分布函数的拟合效果,并以均方根误差(RMSE)最小为准则选取的站点各月份最优分布函数OPT作为参考,分析比较上述7种分布函数模拟的不同时间尺度下的SPI时间序列。结果表明:在各不同时间尺度上,采用皮尔逊III型分布作为计算SPI指数的分布函数误差最小,其模拟的SPI值各站点总偏差平均值均在5%以下,Weibull分布和对数正态分布计算得到的SPI时间系列结果次之,它们模拟的SPI值各站点总偏差平均值介于5%~10%之间;当前我国气象部门统一采用的Gamma分布函数,并非计算SPI指数的最佳选择,各不同时间尺度下,采用Gamma分布计算的SPI值各站点总偏差平均值均在10%左右。其他函数,如正态分布、广义Pareto分布和广义Logistic分布,其误差介于10%~30%之间,均超过采用Gamma分布计算的结果。     

广东省主要站点降雨侵蚀力时间分布规律研究

利用广东省6个站点47 a的日雨量资料,计算并分析了各站降雨侵蚀力的时间分布规律。结果表明,根据降雨侵蚀力峰值在年内出现的次数,可以分为单峰型和双峰型。降雨侵蚀力的年内分布非常集中,连续3个月的降雨侵蚀力平均占全年侵蚀力的一半以上,越靠近沿海的站点侵蚀力的集中度越高。降雨量与降雨侵蚀力集中度的分布规律一致,但与降雨量相比,降雨侵蚀力的分布更加集中。各站降雨量与降雨侵蚀力的倾向率均大于零,说明自1954年以来,雨量和降雨侵蚀力均有不同程度的增加趋势,降雨侵蚀力的增幅远远大于降雨量的增幅。但降雨侵蚀力倾向率的年内分布却是正负皆有。     

用小时降雨资料估算降雨侵蚀力的方法

降雨侵蚀力是进行土壤流失量预报的基本因子,EI₃₀是迄今得到广泛应用的定量指标,但它的计算需要降雨过程资料,使其推广应用受到很大限制。用自动气象观测提供的高精度等间隔降雨资料,代替降雨过程资料,是估算降雨侵蚀力指标的首选。国际上对此已有研究,发现有很大的地区差异性。为此,本文在水蚀严重的中国东部季风区选择5个代表站点,共456次降雨过程资料,建立了用60 min等间隔雨量资料估算次降雨侵蚀力的计算方法。研究结果表明:直接用60 min等间隔资料计算的降雨侵蚀力指标值,与用降雨过程资料计算的结果相比,降雨动能E差异较小,主要差异体现在最大30 min雨强I₃₀上,由此导致降雨侵蚀力指标EI₃₀的差异也十分明显。利用自动气象观测小时雨量资料计算降雨侵蚀力指标值,通过公式(EI₃₀)_bp=1.730(EI₃₀)₆₀转换,可以较精确地估算全国降雨侵蚀力,与用日、月和年降雨量资料的估算值相比,能够提高土壤侵蚀预报精度。     

站点密度对泥石流当日雨量和前期有效雨量计算的影响

研究气象站点密度在当日雨量和前期有效雨量因子内插中的影响程度,对提高对这2个因子的内插精度有重要的意义。本文针对我国东南地区87个泥石流灾害点和207个站点的降水资料,建立站点密度在5%~100%间变化的空间采样方案,采用反距离加权法(IDW)计算不同站点密度下泥石流灾害点的当日雨量和前期有效雨量。利用分位数方法对站点密度与当日雨量和前期有效雨量的关系进行分析,结果表明:诱发泥石流灾害的当日雨量和前期有效雨量的内插值,均随着气象站点密度的降低呈减少趋势;前期有效雨量的内插值相对于当日雨量的内插要稳定一些。在站点密度大于5%的情况下内插的前期有效雨量可以保持60%以上的雨量值,而当日雨量只大于30%。     

纵向岭谷区北回归线一带年降水区域分异特征

利用纵向岭谷区北回归线沿线及南北两侧的24 个站点1961-2007 时段月降水资料,基于全部站点的年降水主要特征统计,以及北回归线沿线的9 个站点的年降水变化的趋势性和阶段性检验、9 个站点间的年降水和5-10 月各月降水序列的相关性检验,分析研究区的年降水及变化分异特征。结果表明：研究区年降水主要特征及其年际变化,呈现出西部、中部、东部的空间分异,即三个区域各自内部自相似度高,而三个区域之间分异明显。北回归线沿线9 个站点的年降水序列都未表现出明显的“突变趋势性-阶段性”变化。研究区中部“岭-谷”相间的地形格局,对中部区域降水空间分异产生一定的影响,但相邻站点间多年际降水序列呈高度相关,意味着控制中部区域降水的决定性因素并不存在明显分异,间隔分布的“岭-谷”地形会带来局地性降水量的分异。     

降雨的空间不均性对模拟产流量和产沙量不确定的影响

在传统的水文/水质模型中,降雨被认为是在空间上均匀分布并且对模型输出的不确定性不产生影响。本文的目的是评价由于降雨的空间分布不均匀性对模型输出-产流量和产沙量-不确定性的影响。本文选取卢氏流域为研究区域,使用SWAT模型和流域内24个雨量站的降雨作为模型的输入。基于降雨空间分布均匀的假定下,每次用一个雨量站的点雨量来作为流域的面平均降雨量,模拟的产流量和产沙量的不确定性来自于降水的不均匀性。模拟的产流量和产沙量的不确定性大于降雨的不确定性,结果表明,在运用水文/水质模型时,为了准确的模拟、预测产流量和产沙量必须掌握降雨的空间分布特性并将其应用于水文、水质模拟之中。     

中国天气发生器的降水模拟

天气发生器是气候影响评价研究的重要工具,在气候变化、地球生态系统及极端气候事件发生的风险分析等方面有着广泛的应用。为了建立一个适用于中国广大地区的天气发生器,需要对各种模拟模型及其参数的估计进行深入的研究。其中降水的模拟及其参数的估计是关键,因为气温、辐射等其他气候要素的模拟依赖于降水的发生。本文重点介绍了常用的随机降水模拟模型：两状态一阶马尔科夫链和两参数GAMMA分布。根据中国672个气象站点1961~2000年的逐日降水资料,计算了降水转移概率P (WD)、P (WW) 及GAMMA分布参数ALPHA和BETA,并分析了4个参数在中国各地的空间分布特征与不同地区各参数的季节分布特征。最后根据各地不同月份计算的四个降水模拟参数对中国各地的逐日降水进行模拟,并利用1971~2000年的实测数据对30年模拟结果在统计意义上进行了检验,模拟结果较好。     

中国北方中西部沙尘暴气候特征

利用1954—2005年西北五省（区）及内蒙古沙尘暴观测资料,分别对沙尘暴的年变化、日变化和持续时间频率分布进行分型和分析,初步探讨了不同类型沙尘暴时间变化特征及其成因。结果表明,沙尘暴发生日数的年变化趋势可划分为4大类型,即春季型、冬春夏型、春夏型和冬春型,它们分别占总数的29.7%、29.7%、27.1%和13.5%,除冬春型冬季沙尘暴发生最多外,其他类型均为春季最多;沙尘暴日变化可划分为3种类型,各型高发时段和峰值时间均有明显差异;持续时间分布规律也划分为3种类型,短时型、中间型和持续型。上述类型划分及分析结果表明,沙尘暴的时间变化特征也存在明显的地域差异,这一差异是某一时期内强风、沙尘源和不稳定层结三者在不同地区配置情况差异的综合反映。     

Analysis of the severe group dust storms in eastern part of Northwest China

Based on the available original dust storm records from 60 meteorological stations, we discussed the identification standard of severe dust storms at a single station and constructed a quite complete time series of severe group dust storms in the eastern part of Northwest China in 1954–2001. The result shows that there were 99 severe group dust storms in this region in recent 48 years. The spatial distribution indicates that the Alax Plateau, most parts of the Ordos Plateau and most parts of the Hexi Corridor are the main areas influenced by severe group dust storms. In addition, the season and the month with the most frequent severe group dust storms are spring and April, accounting for 78.8% and 41.4% of the total events respectively. During the past 48 years the lowest rate of severe group dust storms occurred in the 1990s. Compared with the other 4 decades, on the average, the duration and the affected area of severe group dust storms are relatively short and small during the 1990s. In 2000 and 2001, there were separately 4 severe group dust storms as the higher value after 1983 in the eastern part of Northwest China.     

Analysis of the severe group dust storms in eastern part of Northwest China

Based on the available original dust storm records from 60 meteorological stations, we discussed the identification standard of severe dust storms at a single station and constructed a quite complete time series of severe group dust storms in the eastern part of Northwest China in 1954-2001. The result shows that there were 99 severe group dust storms in this region in recent 48 years. The spatial distribution indicates that the Alax Plateau, most parts of the Ordos Plateau and most parts of the Hexi Corridor are the main areas influenced by severe group dust storms. In addition, the season and the month with the most frequent severe group dust storms are spring and April, accounting for 78.8% and 41.4% of the total events respectively. During the past 48 years the lowest rate of severe group dust storms occurred in the 1990s. Compared with the other 4 decades, on the average, the duration and the affected area of severe group dust storms are relatively short and small during the 1990s. In 2000 and 2001, there were separately 4 severe group dust storms as the higher value after 1983 in the eastern part of Northwest China.     

中国沙尘天气的区域特征

利用筛选的1954~2000年中国338个站沙尘天气资料及相关气候资料,从沙尘天气区划方面着重分析研究了我国沙尘天气的区域特征。结果表明： 1) 我国沙尘天气多发区分别位于以民丰至和田为中心的南疆盆地和以民勤至吉兰泰为中心的河西地区。不同类型沙尘天气的空间分布范围不尽相同,其中沙尘暴主要发生在与北方沙漠及沙漠化土地相联系的极干旱、干旱和半干旱区内。扬沙和浮尘天气除了在沙尘暴发生区的绝大部分地区出现外,还向其它邻近地区扩展,如扬沙可向东北地区和东南的黄淮海平原及以南地区扩展;而浮尘天气则主要向东南方向扩展,可涉及整个黄淮海平原和长江中下游地区。相比之下,上风方向的中高纬地区,如北疆和东北北部地区,浮尘天气发生甚少。2) 全国沙尘暴天气易发区可划分为北疆、南疆、河西、柴达木盆地、河套、东北和青藏等7个亚区。沙尘暴和浮尘在南疆区发生日数最多,而扬沙在河西区发生日数最多。     

我国西北干旱区区域性沙尘暴特征及成因研究

利用1954-2001年131个国家基本基准气象站的逐日观测资料, 运用小波分析等统计方法分析了中国西北地区区域性沙尘暴天气过程特征及变化趋势, 并对其成因进行了初步探讨。结果表明: 1954-2001年, 中国西北地区共出现了1974次区域沙尘暴天气过程, 平均每年42次, 共出现了247次强沙尘暴天气过程, 平均每年5次。沙尘暴和强沙尘暴天气过程的地理分布比较一致, 有两个高发区: 一个位于南疆盆地; 另一个位于西北地区东部。沙尘暴天气过程主要发生在3~7月, 强沙尘暴天气过程集中发生在3~5月。近48a来区域沙尘暴和强沙尘暴天气过程都呈波动下降趋势, 且强沙尘暴天气过程比沙尘暴天气过程下降趋势更为剧烈。1956-1987年是沙尘暴和强沙尘暴天气过程的多发时期, 1988年后无论沙尘暴或强沙尘暴天气过程均大幅减少。2月的北极涛动指数对当年沙尘暴天气过程的发生多少具有预报意义。     

河西走廊东部沙尘暴预报方法研究

利用我国沙尘暴多发区, 甘肃省河西走廊东部民勤、凉州区等五站近50a的气象资料, 详细分析了河西东部沙尘暴频繁发生的气象因素, 应用近20a的天气资料, 结合近50a的典型个例做出沙尘暴的长期、中期、短期和短时预报预警系统。研究结果表明: 长期预报取决于冬春季气温、降水量和大风日数; 中期依靠使用国内外数值预报产品; 短期与大气环流条件、分型指标有关; 短时临近预报与高空大风形势、地面上游有无大风沙尘暴天气有关。     

中国沙尘事件的区域特征

AsevereduststormoccurredoverEastAsiainthemiddletendaysofApril1998.Afteroneweek,itsdustplumeaffectedthecontinentofNorthAmerica.Thisattractedtheattentionofmanyscientistsintherelatedcountriestoduststorms.InrecentyearsthenumberofduststormshasbeenincreasinginChina.Itisnotonlyanimportantsymptomofdesertificationfromslowerchangetofasterchange,butalsoareflectionofthesynthesisofenvironmentalstates,includingclimate,ecosystemandsoil,whicharealldeterioratingtosomedegrees.AccordingtothedefinitionofMeteo…     

Regional characteristics of dust events in China

The regional characteristics of dust events in China has been mainly studied by using the data of dust storm, wind-blown sand and floating dust from 338 observation stations through China from 1954 to 2000. The results of this study are as follows: (1) In China, there are two high frequent areas of dust events, one is located in the area of Minfeng and Hotan in the South Xinjiang Basin, the other is situated in the area of Minqin and Jilantai in the Hexi Region. Furthermore, the spatial distributions of the various types of dust events are different. The dust storms mainly occur in the arid and semiarid areas covering the deserts and the areas undergoing desertification in northern China. Wind-blown sand and floating-dust not only occur in the areas where dust storms occur, but also extend to the neighboring areas. The range of wind-blown sand extends northeastward and southeastward, but floating-dust mainly extends southeastward to the low-latitude region such as the East China Plain and the area of the middle and lower reaches of the Yangtze River. Compared with wind-blown sand, the floating-dust seldom occurs in the high latitude areas such as North Xinjiang and Northeast China. (2) The affected areas of dust storms can be divided into seven sub-regions, that is, North Xinjiang Region, South Xinjiang Region, Hexi Region, Qaidam Basin Region, Hetao Region, Northeastern China Region and Qinghai-Xizang (Tibet) Region. The area of the most frequent occurrence of dust storms and floating-dust is in South Xinjiang Region, and of wind-blown sand in the Hexi Region. In general, the frequency of dust events in all the seven regions shows a decreasing tendency from 1954 to 2000, but there are certain differences between various dust events in different regions. The maximum interannual change and variance of dust events during this time happened in South Xinjiang Region and Hexi Region. The dust events generally occur most frequently in April in most parts of China. The spring occurred days of dust events occupied 60-70% of the whole year in Hetao Region and Northeastern China Region. However, in South Xinjiang Region and North Xinjiang Region, which was less affected by monsoon climate, dust events may occur at any time of the year, less than 50% of the events in this region occur during spring. In the remaining three regions 50-60% of the dust events occur in spring of a year.     

SYNOPTIC-SCALE VARIABILITY IN THE PROBABILITY OF PRECIPITATION FROM THUNDERSTORMS IN THE UNITED STATES

A synoptic-scale climatology of precipitation amounts from thunderstorms was developed by analysis of amounts from individual storms for 220 stations in the conterminous United States for the period 1948-1977. The probability of having a thunderstorm without rainfall was assessed for each station. For storms which did produce precipitation, the probability distribution of amounts was found to be well summarized by the incomplete gamma distribution. Sets of seasonal maps of the probability of receiving any measurable amount, less than 4 mm, and more than 20 mm of precipitation are presented. Consistent spatial patterns are found. Thunder without precipitation is most likely in the west. The greatest probability of heavy precipitation occurs along the Gulf Coast, extending in the summer throughout the mid-section of the nation. Topographic effects are apparent, with mountainous areas generally having less intense precipitation than surrounding regions. [Key World: thunderstorms, precipitation probabilities, gamma distribution.]     

Exploring the relationship between vegetation and dust-storm intensity (DSI) in China

It is difficult to estimate the effects of vegetation on dust-storm intensity (DSI) since land surface data are often recorded aerially while DSI is recorded as point data by weather stations. Based on combining both types of data, this paper analyzed the relationship between vegetation and DSI, using a panel data-analysis method that examined six years of data from 186 observation stations in China. The multiple regression results showed that the relationship between changes in vegetation and variance in DSI became weaker from the sub-humid temperate zone (SHTZ) to dry temperate zone (DTZ), as the average normalized difference vegetation index decreased in the four zones in the study area. In the SHTZ and DTZ zones, the regression model could account for approximately 24.9% and 8.6% of the DSI variance, respectively. Lastly, this study provides some policy implications for combating dust storms.