Modeling elements and geographical factors of formation of the monthly runoff as exemplified by the Kuda river

We have formulated the main principles of constructing the model for a seasonal cycle of the runoff with a temporal resolution of one month. We worked out the elements of technology for the first stages of model making. Based on analyzing the local data on monthly characteristics of the runoff, precipitation and temperature, we formulated and verified a number of hypotheses about the physicalgeographical patterns of formation of seasonal and interannual variability in water flow rates for three river discharge sections within the Kuda river basin. The most important factors have been determined, which govern the observed values of monthly mean flow rates.     

The impacts of climate change on hydrology in a typical glacier region—A case study in Hailuo Creek watershed of Mt. Gongga in China

The glaciers of the Hengduan Mountains play an important role in the hydrology processes of this region. In this study, the HBV Light model, which relies on a degree-day model to simulate glacier melting, was employed to simulate both glacier runoff and total runoff. The daily temperature and precipitation at the Hailuo Creek No. 1 Glacier from 1952 to 2009 were obtained from daily meteorological observed data at the glacier and from six national meteorological stations near the Hailuo Creek Basin. The daily air temperature, precipitation, runoff depth, and monthly potential evaporation in 1995, 1996, and 2002 were used to obtain a set of optimal parameters, and the annual total runoff and glacier runoff of the Hailuo Creek Glacier(1952–2009) were calculated using the HBV Light model. Results showed the average annual runoff in the Hailuo Creek Basin was 2,114 mm from 1952 to 2009, of which glacial melting accounted for about 1,078 mm. The river runoff in the Hailuo Creek catchment increased as a result of increased glacier runoff. Glacier runoff accounted for 51.1% of the Hailuo Creek stream flow in 1994 and increased to 72.6% in 2006. About 95% of the increased stream flow derived from the increased glacier runoff.     

Modeling tropical river runoff: A time dependent approach

Forecasting of rainfall and subsequent river runoff is important for many operational problems and applications related to hydrology. Modeling river runoff often requires rigorous mathematical analysis of vast historical data to arrive at reasonable conclusions. In this paper we have applied the stochastic method to characterize and predict river runoffofthe perennial Kulfo River in southem Ethiopia. The time series analysis based auto regressive integrated moving average （ARIMA） approach is applied to mean monthly runoff data with 10 and 20 years spans. The varying length of the input runoff data is shown to influence the forecasting efficiency of the stochastic process. Preprocessing of the runoff time series data indicated that the data do not follow a seasonal pattern. Our forecasts were made using parsimonious non seasonal ARIMA models and the results were compared to actual 10-year and 20-year mean monthly runoff data of the Kulfo River. Our results indicate that river runoff forecasts based upon the 10-year data are more accurate and efficient than the model based on the 20-year time series.     

Sediment concentration changes in runoff pathways from a forest road network and the resultant spatial pattern of catchment connectivity

Hydrological connectivity is a term often used to describe the internal linkages between runoff and sediment generation in upper parts of catchments and the receiving waters. In this paper, we identify two types of connectivity: direct connectivity via new channels or gullies, and diffuse connectivity as surface runoff reaches the stream network via overland flow pathways. Using a forest road network as an example of a landscape element with a high runoff source strength, we demonstrate the spatial distribution of these two types of linkages in a 57 km² catchment in southeastern Australia. Field surveys and empirical modelling indicate that direct connectivity occurs primarily due to gully development at road culverts, where the average sediment transport distance is 89 m below the road outlet. The majority of road outlets were characterised by dispersive flow pathways where the maximum potential sediment transport distance is measured as the available hillslope length below the road outlet. This length has a mean value of 120 m for this catchment. Reductions in sediment concentration in runoff plumes from both pathways are modelled using an exponential decay function and data derived from large rainfall simulator experiments in the catchment. The concept of the volume to breakthrough is used to model the potential delivery of runoff from dispersive pathways. Of the surveyed road drains (n=218), only 11 are predicted to deliver runoff to a stream and the greatest contributor of runoff occurs at a stream crossing where a road segment discharges directly into the stream. The methodology described here can be used to assess the spatial distribution and likely impact of dispersive and gullied pathways on in-stream water quality.     

渔子溪流域水文过程影响因素的特征时间尺度分析

采用渔子溪流域三期遥感影像(1986、1994和2002年)获取该区域的土地利用/覆被资料,并结合实际调查数据、采用斑块驱动模型模拟得到该流域1986~2004年间逐年土地利用变化数据;然后根据该模拟数据以及渔子溪流域不同时间尺度的气象和径流资料,采用逐步线性回归和多元非线性回归方法,对渔子溪流域不同时间尺度下降雨-土地利用-径流系统及其内部各元素之间的关系进行了统计分析,结果表明:(1)不同时间尺度下径流量与降雨量的相关关系具有明显差别,土地利用变化对降雨-径流关系有重要影响,在年尺度上,降雨量对径流量的影响要大于气温对径流的影响,而在月尺度上则相反。(2)不同时间尺度下非线性回归的模拟误差要比线性模拟误差下降1.18%~23.8%,且不同影响因子在不同时间尺度上的非线性程度各不相同。     

挠力河流域河流径流量对气候变化的敏感性分析

流域径流量对气候变化的敏感性分析是理解气候变化对流域水资源影响的重要手段。本文利用非更新式人工神经网络(ANN)模型,以年平均降雨、年最低气温和最高气温为输入参数,年平均径流量为输出变量,构建了三江平原挠力河流域的径流量预测ANN模型;并根据IPCC第四次报告的气候变化模式,设定了9种不同的气候变化情景,利用构建的ANN模型分析了流域径流量对气候变化的敏感性。结果表明:构建的人工神经网络模型能够较好的模拟径流量,可用于气候变化的敏感性分析;挠力河流域上游径流量对气候变化的敏感性要大于中游区域的,降水变化对径流量的影响大于气温对其产生的影响。     

湖泊集水域地表—地下径流联合模拟

研制了流域尺度的地表- 地下径流联合模拟的分布式水文模型。模型考虑了地表径流、土 壤水、地下水之间的相互作用和水量交换, 更真实地模拟径流系统。特别是, 考虑了湖泊- 流域系 统的特点, 例如, 多条入湖河流、直接入湖的坡面水流和地下入湖径流等, 使模型比现有水文模型 更适合于湖泊集水域径流系统的模拟。模型在云南抚仙湖集水域作了初步应用研究, 模拟结果与 河道径流、土壤含水量和地下水位等观测数据的比较显示, 模型模拟效果理想。此外, 模拟结果与 SCS 模型结果的对比分析, 进一步验证了模型的有效性。该模型可用于研究湖泊与流域的相互作 用、模拟流域水文过程对自然条件改变或人类活动的响应、探究地表径流- 地下水- 湖泊之间的相 互作用。模型也可用作湖泊- 流域系统水量平衡分析和水资源管理的有效计算工具。     

长江河口区平均海面季节性变化的分析

长江河口区的月平均海面变化主要受径流及其它海洋水文气象因子的作用,而且具有明显的季节性变化。本文为了探讨月平均海面时空变化的规律,运用了经验特征函数,对月平均海面的时空变化进行分解。以三个相互正交模型,估算了水文气象因子在月平均海面变化中所起的作用。     

基于地块汇流网络的小流域水沙运移模拟方法研究

全流域逐地块水土流失计算,是目前水土保持定量评价的重要手段,其实现过程既需要有考虑地貌因子和上下游关系的土壤侵蚀模型,也需要建立全流域地块汇流网络,并与侵蚀模型有机连接。本文针对黄土高原特殊的地理条件和水土流失规律,对传统的基于栅格的小流域汇流技术进行了改进,考虑地块间上下游汇流关系,建立了基于地块的水沙汇流网络模型,模拟水沙在流域复杂下垫面的汇流过程,提取出流域各地块间的水沙汇流网络,并计算出3个重要参数:流域地块间水沙汇流的顺序、流域地块间水沙汇流数目的空间分配、水沙流经各地块的坡长。将土壤侵蚀模型按地貌部位与特征分为坡面模型、沟坡模型和沟道输移模型,与地块汇流网络有机集成,实现了水沙运移的全流域按地块沿程计算。     

Comparison of different models for annual hydrological cycle and calculated estimates of the monthly runoff from the catchment

A comparison is made of the performance of different schemes (models) for calculating the monthly mean water runoff from the catchment. To maximize objectivity of results and for ease of the modeling process, a special-purpose software was developed in order to construct the models compared using a unified technology.     

新疆主要河流水文极值变化趋势

水文极值是工程水文计算的基础资料,实测样本资料的变化直接影响水文设计值。基于1956-2006年的实测洪峰、洪量、年最小流量、最枯月径流量等资料,分析了各水文要素的变化及其趋势。结果表明:20世纪80年代中期以来超标准洪峰、洪量的频次增加,大多数河流洪水峰、量都呈增大变化趋势,说明新疆洪水又进入活跃期,在工程水文设计洪水计算中要应尽可能搜集和利用近期暴雨和洪水资料,在洪水系列中增加更多大洪水信息,提高设计洪水计算成果的稳定性。年最小流量和最枯月径流量是维持河流生命和河道两岸自然生态的基础流量,基流量增大有利于对生态补给、水利发电,水库蓄水,径流年内分配趋于均匀有利于水量年内调节。但在枯水径流分析计算中应充分重视不同年份最枯水量出现时间有明显推迟的迹象。     

陇中黄土丘陵沟壑区人工草地土壤水蚀预测模型

针对陇中黄土丘陵沟壑区土壤水蚀过程复杂且难以有效预测的问题,以定西市安家沟水土保持试验站2005—2016年1~12月人工草地径流场试验数据为主要来源,将流域月降雨量、月侵蚀性降雨量、月径流量、月降雨强度、径流场面积、径流场坡度、土壤砂粒含量、土壤粘粒含量8个因子作为输入因子,月土壤水蚀量作为输出,运用偏最小二乘法(Partial Least-Squares Regression,PLSR)和长短期记忆(Long Short-Term Memory,LSTM)循环神经网络建立人工草地土壤水蚀预测模型,并利用BP(Back Propagation)、RNN(Recurrent Neural Network)、LSTM常见神经网络模型,对模型的有效性进行评估。结果表明：PLSR将模型8个输入因子减少为4个,从而有效解决LSTM神经网络模型对样本数量要求过高的问题; PLSR和LSTM神经网络模型的结合可以有效提高模型对人工草地土壤水蚀过程的预测精度和收敛速度,预测结果的平均相对误差小于4%,相关系数高于其他3种神经网络模型,而迭代次数、均方根误差和平均绝对误差均低于其他3种模型;研究发现坡度对人工草地土壤水蚀过程影响较为明显,降雨量小于25 mm时,人工草地土壤水蚀量不会随坡度增加而明显增长,但当降雨量超过25 mm时,人工草地土壤水蚀量会随坡度明显增加。 PLSR LSTM神经网络土壤水蚀预测模型可以准确预测陇中黄土丘陵沟壑区人工草地土壤水蚀量,为该地区水土流失的准确预报提供新的思路和方法。     

A catchment-scale model of mountain stream channel morphologies in southeast Australia

The position of mountain streams high in the channel network and their proportional dominance mean that channel modifications and adjustments within these systems will have important implications for downstream processes and linkages. This study develops an analysis framework for examining the catchment-scale distribution of reach morphologies, and the relationship among reach type, catchment lithology and flow competence in southeast Australian mountain streams. The analysis framework is applied to three catchments which have contrasting proportions of the two dominant lithologies of the region, Devonian granites and Ordovician metasediments.The model successfully delineated 68% of reach types, and the resultant spatial maps allowed the effects of stream network position and catchment specific controls on channel morphology to be evaluated. Maximum lengths of the majority of reach morphology types were in second-order streams and the maximum number of morphology types (six) was present in third-order streams, with dramatic reductions in reach type variability as the network expands. The position of catchment lithology within the channel network structure was recognized as more important than the aerial extent of a particular lithology on the distribution and abundance of reach morphologies. The model provides an important tool in the management of channel networks for the protection or restoration of ecological diversity, by identifying river segments and tributaries with high morphological diversity.     

基于PCA和BP神经网络的径流预测

径流预测为流域水资源的合理开发利用与统筹配置提供依据。运用多元线性回归、主成分回归、BP神经网络及主成分分析和BP神经网络相结合的方法,对新疆呼图壁河流域石门水文站2009-2011年各月径流量进行预测,并采用相关系数、确定性系数及均方根误差对各模型预测精度进行比较。结果表明：（1）神经网络等智能算法具有高速寻优的能力,对短时间尺度的月径流量的预测结果较好;（2）主成分回归等常规算法能充分反映出某地区径流的年际的稳定性,对全年径流总量的模拟精度较高;（3）主成分分析和BP神经网络相结合的方法,提高了神经网络的收敛速度,同时降低了局部极值的影响,优于简单的BP神经网络,适用于呼图壁河月径流量预测。     

气候变化情景下青海湟水流域径流变化的HIMS模拟分析

基于国产HIMS(Hydro-Informatic Modeling System)模型,以青海湟水流域为研究区域,利用1986-2000年33个雨量站和8个气象站的逐日降水和气温数据,对其径流变化进行模拟;选取流域内6个水文站同期的实测径流数据,进行参数率定及验证。结果表明:HIMS模型日、月率定及验证结果良好,在湟水流域具有良好的适用性。在此基础之上,分析了湟水流域1961-2010年降水及气温的变化趋势,并对不同气候变化情景下的水文响应(径流量)进行模拟分析。结果显示气候变化对湟水流域径流量变化趋势影响显著,随气温升高和降水量的减少,径流量呈明显的减少趋势,反之,呈增加趋势。     

东线南水北调对黄浦江水质的影响

本文对长江枯季(12—3月)径流的特点、变化规律、河口盐水入侵以及对黄浦江水质的影响等进行了初步分析。长江水量虽然丰富,但枯季径流并不大,枯季径流一般占年径流总量的15%左右。频率50%的平水年各月平均流量都在15000秒立米以下,频率75%的一般少水年枯季1、2月份的月平均流量不足10000秒立米。河口盐度变化随上游来水多少而异,大通站流量在10000秒立米时河口盐度变化敏感,流量再减少河口盐度将急剧增加,便会严重污染河口和黄浦江水质,影响上海市和河口地区的工农业和城市生活用水。因此,在规划长江流域用水时,应使大通站流量在15000秒立米以上,若有困难,至少不能低于10000秒立米。     

1951-2002年长江流域降水特征

The monthly, seasonal, and annual precipitation trends in the Yangtze river catchment have been detected through analysis of 51 meteorological stations‘ data between 1950-2002 provided by National Meteorological Administration. Results reveal that: 1) Summer precipitation in the Yangtze river catchment shows significant increasing tendency. The Poyanghu lake basin, Dongtinghu lake basin and Taihu lake basin in the middle and lower reaches are the places showing significant positive trends. Summer precipitation in the middle and lower reaches experienced an abrupt change in the year 1992; 2) The monthly precipitation in months just adjoining to summer shows decreasing tendency in the Yangtze river catchment. The upper and middle reaches in Jialingjiang river basin and Hanshui river basin are the places showing significant negative trends; 3) Extreme precipitation events show an increasing tendency in most places, especially in the middle and lower reaches of the Yangtze river catchment.     

Observed precipitation trends in the Yangtze river catchment from 1951 to 2002

The monthly, seasonal, and annual precipitation trends in the Yangtze river catchment have been detected through analysis of 51 meteorological stations' data between 1950-2002 provided by National Meteorological Administration. Results reveal that: 1) Summer precipitation in the Yangtze river catchment shows significant increasing tendency. The Poyanghu lake basin, Dongtinghu lake basin and Taihu lake basin in the middle and lower reaches are the places showing significant positive trends. Summer precipitation in the middle and lower reaches experienced an abrupt change in the year 1992; 2) The monthly precipitation in months just adjoining to summer shows decreasing tendency in the Yangtze river catchment. The upper and middle reaches in Jialingjiang river basin and Hanshui river basin are the places showing significant negative trends; 3) Extreme precipitation events show an increasing tendency in most places, especially in the middle and lower reaches of the Yangtze river catchment.     

汉江流域安康站日径流预测的LSTM模型初步研究

论文基于2003—2014年水文资料,采用长短期记忆神经网络(Long-Short Term Memory,LSTM),构建了汉江上游安康站日径流预测模型,评价了不同输入条件下日径流预测的精度。结果表明：当预见期为1 d时,在仅以安康站前期日径流量作为输入的条件下,LSTM模型在训练期和检验期的效率系数分别达到0.68和0.74;如再将流域前期面雨量和上游石泉站前期日径流量加入LSTM网络作为输入变量,安康站日径流量预测效果将更好,训练期和检验期的效率系数最高可达到0.83和0.84,均方根误差也有显著削减,且对主要洪峰流量的预测能力也有一定提高。此外,LSTM可以有效避免过拟合等问题,具有较好的泛化性能。但当预见期从1 d延长至2、3 d时,LSTM的预测精度显著降低。     

How does alluvial sedimentation at range fronts modify the erosional dynamics of mountain catchments?

At the geological time scale, the way in which the erosion of drainage catchments responds to tectonic uplift and climate changes depends on boundary conditions. In particular, sediment accumulation and erosion occurring at the edge of mountain ranges should influence the base level of mountain catchments, as well as sediment and water discharges. In this paper, we use a landform evolution model (LEM) to investigate how the presence of alluvial sedimentation at range fronts affects catchment responses to climatic or tectonic changes. This approach is applied to a 25 km × 50 km domain, in which the central part is uplifted progressively to simulate the growth of a small mountain range. The LEM includes different slope and river processes that can compete with each other. This competition leads to ‘transport‐limited’, ‘detachment‐limited’ or ‘mixed’ transport conditions in mountains at dynamic equilibrium. In addition, two end‐member algorithms (the channellized‐flow and the sheet‐flow regimes) have been included for the alluvial fan‐flow regime. The three transport conditions and the two flow algorithms represent six different models for which the responses to increase of rock uplift rate and/or cyclic variation of the precipitation rate are investigated. Our results indicate that addition of an alluvial apron increases the long‐term mountain denudation. In response to uplift, mountain rivers adapt their profile in two successive stages; first by propagation of an erosion wave and then by slowly increasing their channel gradients. During the second stage, the erosion rate is almost uniform across the catchment area at any one time, which suggests that dynamic equilibrium has been reached, although the balance between erosion and rock uplift rates has not yet been achieved. This second stage is initiated by the uplift of the mountain river outlets because of sedimentation aggradation at the mountain front. The response time depends on the type of water flow imposed on the alluvial fans domains (× by 1.5 for channelized flow regime and by 10 for the sheet flow one). Cyclic variations of precipitation rate generate cyclic incisions in the alluvial apron. These incision pulses create knick‐points in the river profile in the case of ‘detachment‐limited’ and ‘mixed’ river conditions, which could be mistaken for tectonically induced knick‐points. ‘Transport‐limited’ conditions do not create such knick‐points, but nevertheless trigger erosion in catchments. The feedbacks linked to sedimentation and erosion at range front can therefore control catchment incision or aggradation. In addition, random river captures in the range front trigger auto‐cyclic erosion pulses in the catchment, capable of generating incision–aggradation cycles.