Identification of physical processes inherent in artificial neural network rainfall runoff models

The emergence of artificial neural network (ANN) technology has provided many promising results in the field of hydrology and water resources simulation. However, one of the major criticisms of ANN hydrologic models is that they do not consider/explain the underlying physical processes in a watershed, resulting in them being labelled as black‐box models. This paper discusses a research study conducted in order to examine whether or not the physical processes in a watershed are inherent in a trained ANN rainfall‐runoff model. The investigation is based on analysing definite statistical measures of strength of relationship between the disintegrated hidden neuron responses of an ANN model and its input variables, as well as various deterministic components of a conceptual rainfall‐runoff model. The approach is illustrated by presenting a case study for the Kentucky River watershed. The results suggest that the distributed structure of the ANN is able to capture certain physical behaviour of the rainfall‐runoff process. The results demonstrate that the hidden neurons in the ANN rainfall‐runoff model approximate various components of the hydrologic system, such as infiltration, base flow, and delayed and quick surface flow, etc., and represent the rising limb and different portions of the falling limb of a flow hydrograph. Copyright © 2004 John Wiley & Sons, Ltd.     

Long-term abundance patterns of macroalgae in relation to environmental variables in the Tagus Estuary (Portugal)

Seasonal and inter-annual patterns of macroalgal abundance in a Tagus Estuary oyster reef are described. Macroalgal abundance was estimated as canopy percent cover by three permanent point intercept transects over a 7-year period. Four categories were defined, corresponding to bare substrate and three different macroalgal functional-form groups: (1) ULVA, foliose group, included Ulva spp.; (2) GRACIL, terete corticated macrophyte group, included only Gracilaria gracilis; and (3) FILAM, small (<10 cm) filamentous group, including eight species. A canonical correspondence analysis (CCA) showed that: (1) ULVA were associated with long and hot days, being usually dominant during spring and especially summer; (2) FILAM were associated with mild temperatures and relatively long days, abundant in spring but showed frequent peaks in summer; and (3) GRACIL were also favoured by spring season, although associated to lower temperature and less daylight hours than FILAM. GRACIL and FILAM were present throughout the year. On the contrary, ULVA were absent or with low cover during colder periods. A negative correlation between GRACIL and FILAM seems to indicate competition between the two categories. The applied models explained 23.3% of the temporal variance in category abundance. Rainfall negatively affected macroalgal cover, as indicated by the positive correlation between rainfall and bare substrate. Our conclusions are in agreement with previous studies that consider algae as excellent environmental integrators, even on a small scale, due to a strong link between the macroalgal communities and relevant environmental variables. It is also relevant that this study used open-access databases of environmental variables, which open up new possibilities for mining existing data resources in new ways. Due to large inter-annual variability, long-term studies are essential to understand population dynamics in estuarine phytobenthic communities.     

Using δN values in algal tissue to map locations and potential sources of anthropogenic nutrient inputs on the island of Maui, Hawai‘i, USA

Macroalgal blooms of Hypnea musciformis and Ulvafasciata in coastal waters of Maui only occur in areas of substantial anthropogenic nutrient input, sources of which include wastewater effluent via injection wells, leaking cesspools and agricultural fertilizers. Algal δ¹⁵N signatures were used to map anthropogenic nitrogen through coastal surveys (island-wide and fine-scale) and algal deployments along nearshore and offshore gradients. Algal δ¹⁵N values of 9.8‰ and 2.0-3.5‰ in Waiehu and across the north-central coast, respectively, suggest that cesspool and agricultural nitrogen reached the respective adjacent coastlines. Effluent was detected in areas proximal to the Wastewater Reclamation Facilities (WWRF) operating Class V injection wells in Lahaina, Kihei and Kahului through elevated algal δ¹⁵N values (17.8-50.1‰). From 1997 to 2008, the three WWRFs injected an estimated total volume of 193 million cubic meters (51 billion gallons) of effluent with a nitrogen mass of 1.74 million kilograms (3.84 million pounds).     

人工掏挖坡面侵蚀微地貌演化及其水力学特性分析

人工掏挖是黄土高原地区夏闲地翻耕时广为应用的耕作方式,为探明其坡面侵蚀过程,利用三维激光扫描仪及ArcGIS软件,阐明其在人工模拟间歇降雨下坡面微地貌、侵蚀产沙及水力学参数演变规律。结果表明:①坡面侵蚀历经溅蚀-片蚀、断续细沟、连续细沟3个阶段;降雨截止时细沟平面密度、平均沟深、最大沟长和最大沟深分别增至初始的1.42倍、2.24倍、15.5倍和2.43倍。②地表糙度随降雨历时推移从1.706呈近似线性趋势减小至1.488;累积降雨量达80 mm之前,径流量、含沙量随地表糙度的减小增加缓慢,但之后随地表糙度减小增加剧烈。③降雨过程中流速呈现波动增加趋势,水流由层流快速过渡至紊流,但始终保持为缓流,水流阻力呈波动下降,且形态阻力一直居于主导地位。人工掏挖耕作坡面在雨强1.5 mm/min、降雨量80 mm以内可起到蓄水保土作用,但在连续强降雨下也更易引起细沟侵蚀。     

基于WorldView影像实现暗礁识别定位

通过对WorldView遥感影像的纠正、融合、直方图拉伸、采集等处理,获得了海域内暗礁的空间位置及其数字正射影像,并根据相关资料及外业调绘数据,录入暗礁属性信息,建立暗礁的地理信息数据库,实现暗礁的识别定位。     

Structural damage detection using the optimal weights of the approximating artificial neural networks

This work presents a novel neural network‐based approach to detect structural damage. The proposed approach comprises two steps. The first step, system identification, involves using neural system identification networks (NSINs) to identify the undamaged and damaged states of a structural system. The partial derivatives of the outputs with respect to the inputs of the NSIN, which identifies the system in a certain undamaged or damaged state, have a negligible variation with different system errors. This loosely defined unique property enables these partial derivatives to quantitatively indicate system damage from the model parameters. The second step, structural damage detection, involves using the neural damage detection network (NDDN) to detect the location and extent of the structural damage. The input to the NDDN is taken as the aforementioned partial derivatives of NSIN, and the output of the NDDN identifies the damage level for each member in the structure. Moreover, SDOF and MDOF examples are presented to demonstrate the feasibility of using the proposed method for damage detection of linear structures. Copyright © 2001 John Wiley & Sons, Ltd.     

雷电定位系统与人工观测雷暴日数统计比较

为了利用雷电定位系统 (lightning location system,LLS) 资料统计人工观测雷暴日数,采用湖北省2007—2012年LLS监测资料,选取25个气象站为圆心,统计其不同监测半径 (r) 圆区域内LLS监测的雷电日数,并与人工观测雷暴日数进行比较。结果表明:r≤7 km时,LLS监测平均年雷电日数小于人工观测平均年雷暴日数;r≥8 km时, LLS监测平均年雷电日数大于人工观测平均年雷暴日数;r=22 km圆区域内年平均雷电日数可替代最大年雷暴日数。根据r=7 km,r=8 km圆区域内LLS监测的年雷电日数、年平均地闪密度资料,分别采用直接替代法、地闪密度法和该文提出的二元法计算年雷暴日数,结果显示:二元法效果最好。二元法计算的2007—2012年25个站平均年雷暴日数与人工观测相等,平均差异为7.4%;二元法计算的2013年年雷暴日数与人工观测相差0.8 d,平均差异为12.3%。     

基于生态化人工海岸建设的理论研究与应用 ——以天津永定新河口综合整治修复工程为例

分析生态文明背景下围填海区域人工海岸建设存在的问题,结合淤泥质人工海岸空间特点,总结提出了生态岸线整治修复技术、仿自然化处理设计、景观构建适宜性、人工海岸生态系统等生态化建设基础理论,并以天津永定新河口综合整治修复工程为例,重点阐述了人工海岸生态化建设过程中生态系统和景观构建的方法要点和具体思路.结果表明:在人工海岸生...     

德州市深层地下水人工回灌试验浅探

德州市德城区由于长期开采深层地下水,引起了水位区域性大幅度下降,形成了大面积地下水降落漏斗,并产生了地面沉降等不良地质环境问题。为保护德州市地质环境和深层地下水资源,于2002年开展了深层地下水人工回灌试验。根据回灌试验成果和深层地下水目前开采强度推算,利用德城区现有部分(65%)深机井在丰水季节进行人工回灌,单井回灌量保持1200m3/d,每年保持90d的回灌时间,回灌量可达到1080×104m3/a,可增可采资源量675×104m3/a。因此开展人工回灌可以有效地补给地下水源,解决深层地下水资源枯竭问题。     

"催化-供给"云降水形成机理的数值模拟研究

利用含有详细微物理过程的一维层状云模式模拟,研究了2002年4月5日冷锋降水性层状云云系中"催化-供给"云的微物理结构、降水粒子形成的环节和微物理过程,并从降水形成的环节和云的结构分析人工增雨的条件.结果说明,"催化-供给"云具有显著的分层结构:云内高层是冰晶,下层是雪,接下来是霰和过冷云水组成的冰水混合层,最下方是云中暖区的液水层.作为催化云层的冰水层对降水的贡献约25.5%,冰水混合层为31.3%,液水层为43.1%,亦即供给云对降水的贡献约74.4%.具有"催化-供给"云结构的层状云降水形成的主要环节是:冰晶通过凝华增长转化成雪,雪撞冻过冷云水、收集冰晶和凝华增长转化形成霰,霰靠撞冻过程、收集雪过程长大,从而形成可以降落到云的暖区融化形成雨水的粒子,它对降水的贡献较大.凝华和撞冻增长过程是冰粒子增长的主要物理过程,也是雨水产生的重要过程."催化-供给"云体系是重要的人工增雨条件,云中水汽对雨水形成的贡献与过冷云水几乎相当,与过冷云水一样,水汽也是人工增雨的重要条件.