Transient Rainfall-Runoff Loadings to a Partial Exfiltration System: Implications for Urban Water Quantity and Quality

Modification of rainfall-runoff processes by urban infrastructure and anthropogenic activities impacts receiving waters and the surrounding terrestrial environment. Infiltration–exfiltration systems such as a partial exfiltration reactor (PER) when loaded by transient sheet flow have the potential to attenuate the impact of both the quantity and quality of urban runoff. These in situ systems are subject to highly variable water quality and quantity while functioning under variably saturated flow conditions. To improve the understanding of field-scale PER performance as a rainfall-runoff unit operation and process, a two-dimensional (2D) numerical model was used to simulate the effluent hydrograph and water content profiles under transient hydraulic loadings. Richard’s equation was applied in the 2D model using parameters estimated from laboratory experiments and hydrographs measured for an in situ PER. The temporal dynamics of the water content illustrated the ability of the PER to lower peak flow, redistribute volume, and attenuate temporal aspects of the inflow hydrograph. Results demonstrated the role of the PER to attenuate runoff water quantity, while also providing water quality improvements, as illustrated for suspended solids and dissolved Cu. Simulation of historical events for different surrounding soils illustrated the critical role of surrounding soil conditions on PER performance. While the PER demonstrated water quantity attenuation benefits for design storms (1, 2, 5?year return periods), results also illustrate how a given PER design for clayey soils conditions can be limiting for intense events. Evaporation was a dominant mechanism for the drying process in the PER upper layer; with a residual moisture content in the porous pavement layer achieved in less than 2?days in summer for Cincinnati, Ohio.     

Pollution Buildup on Road Surfaces

Samples of sediment found on an urban road in Aberdeen (Scotland) were collected by washing designated surfaces. This method, called the “wet” method, was capable of collecting sediment of the smallest particle size range that is normally left behind by traditional sampling techniques using dry vacuuming. Over 17 months (mainly on a weekly basis), 66 samples were collected and analyzed for sediment loading, particle size distribution, concentration of heavy metals (Zn, Cu, Pb, Cd) in four sediment size fractions, and several dissolved pollutants in the effluent collected from washing (NO3?,?SO42?,?PO43?,?Cl?,?F?,?NH4+, total organic carbon, total carbon). Standard statistical methods, including multiple regressions, were used to determine relationships amongst different sediment characteristics. It was found that sediment loading, as well as concentrations of Cl? and SO42?, were highest in the winter months, especially when snow was present on the road surface. It was observed that 66% of total road sediment loading was found within a 0.5 m strip next to the curb. The average sediment particle sizes found were smaller than those previously recorded in the literature. As expected, the concentrations of heavy metals were highest in the smallest particle size fraction analyzed (<63?μm), and this occurred during the summer months when less sediment was available on the street surface. The antecedent dry weather period had a very weak and negative influence on the loading rate of the smallest particle size fraction next to the curb.     

流域内森林植被对径流的影响

分析了大兴安岭地区呼玛河上游森林植被覆盖率变化对流域径流的影响。     

基于多尺度下AR(P)耦合预测模型的应用研究

基于多尺度分析理论，运用Mallat算法和Daubeclfies小波，把时间序列分解为比原始序列更单一的细节和慨貌部分，并利用AR(P)模型能反映时间序列中邻近时刻间联系的特性，对序列分解后的部分进行拟合与预测，然后再由多尺度分析中的重构方法进行序列重构，由此建立耦合的预测模型。通过黄河青铜峡270多年(1724～1997)年径流时间序列的建模及验证，表明拟建的耦合模型与传统单一模型的预测精度相比，由50％提高到90％，可用于实际需要。     

鸡西市地表水资源量计算与分析

对鸡西市降水量、蒸发量、径流量的多年平均值及不同频率的降水量、径流量进行了分析计算。     

基于BP神经网络马尔科夫模型的径流量预测

讨论了马尔科夫链状态划分的黄金分割率法和"马氏性"检验法,并针对BP神经网络预测和马尔科夫预测的优缺点,提出了BP神经网络与马尔科夫相耦合的BP神经网络马尔科夫模型,以石泉水库年入库径流量为例,验证了该方法的可行性.     

不同土地利用方式坡面产流输沙过程模拟研究

针对目前多数坡面侵蚀预测模型难以模拟土地利用变化对坡面产流输沙过程影响的问题,基于运动波方程和泥沙输移公式,给出了一种能够反映不同土地利用方式下坡面产流输沙时空变化的简化模型,并利用裸地、草地、坡耕地、灌木地及幼林地的试验及自然降雨条件下坡面产流输沙实测资料对模型进行验证,符合较好.利用该模型对黄土高原不同降雨条件下不同土地利用场景坡面产流输沙特征进行了模拟,结果表明:典型降雨条件下,5 a树龄以上的林地或覆盖度在60％以上的草地坡面标准小区基本不再发生侵蚀;坡度小于15°的裸地坡面中下部发生轻度侵蚀;十年一遇降雨条件下,10 a树龄以上的林地坡面土壤侵蚀基本得到控制,而草地调控坡面水沙的能力有限,裸地更差.对低强度短历时降雨,草地调控水沙的见效较快;对高强度长历时降雨,林地调控水沙的作用较大,两者宜配合进行;对不同坡度裸地,要严格控制坡面坡长.研究结果为优化土地利用方式提供了理论支撑.     

丹江口水库径流长期预报研究

应用时间序列分析方法，建立了丹江口水库径流序列的长期预报分解模型；采用多种方法对序列特性（如平稳、趋势、周期等）进行了检验，深入讨论了序列周期的稳定性和外推性。计算成果表明所提模型是合理的．     

双超模型在半湿润地区的应用

中国北方多属半干旱、半湿润地区,其降雨径流的形成过程较为复杂,既有蓄满产流,又有超渗产流,针对这种产流现象,分析其物理机制,在前人研究的基础上,探讨了适用于超渗产生地表径流、超持产生壤中流的＂双超＂产流模型和汇流模型,并将其应用于文峪河流域,进行了洪水模拟应用,研究结果表明,洪水过程的模拟预测结果与实测结果基本吻合;对于不同年份降雨的洪水模拟,除少数敏感的模型参数外,其它模型参数几乎不变,说明该模型弹性较好,因此该模型在这类地区具有较强的适用性和实用性。