Spring snowmelt peak flow （SSPF） can cause serious damage. Precipitation as rainfall directly contributes to the SSPF and influences the characteristics of the SSPF, while temperature indirectly impacts the SSPF by shaping snowmelt rate and determining the soil frozen state which partitions snowmelt water into surface runoff and soil infiltration water in spring. It is necessary to identify the important and significant paths of climatic factors influencing the SSPF and provide estimates of the magnitude and significance of hypothesized causal connections between climatic factors and the SSPF. This study used path analysis with a selection of five factors - the antecedent precipitation index （API）, spring precipitation （SP）, winter precipitation as snowfall （WS）, 〈0℃ temperature accumulation in winter （[ATNI）, and average 〉0℃temperature accumulation in spring （AT） - to analyze their influences on the SSPF in the Kaidu River in Xinjiang, China. The results show that {ATN}, AT and WS have a significant correlation with the SSPF, while API and SP do not show a significant correlation. AT and WS directly influence the SSPF, while as the influence of[ATN] on SSPF is indirect through WS and AT. The indirect influence of [ATN[ on SSPF through WS accounts for 69% of the total influence of [ATN] on SSPF. Compared to the multiple linear regression method, path analysis provides additional valuable information, including influencing paths from independent variables to the dependent variable as well as direct and indirect impacts of external variables on the internal variable. This information can help improve the description of snow melt and spring runoff in hydrologic models as well as the planning and management of water resources.
SWAT(soil and water assessment tool)水文模型是近几年国内外使用最广泛的分布式水文模型,因其良好的精度、广泛的适应度得到了众多水文工作者的欢迎。但由于其参数率定部分过高的计算量,给水文工作者在建模调参过程中造成了大量的不便。为了提高SWAT参数率定的运行效率,基于克里金插值理论,对于SWAT原有的优化算法进行改进,提出了一种基于随机过程的参数率定方法,通过一种期望优化的方式对拟合曲面进行寻优。通过Intel MKL和一个对等式并行框架加以并行实现,大大提高了SWAT参数寻优的计算效率以及参数精度,并通过天山北坡流域的模拟验证了并行模型的有效性,说明了克里金优化方法在水文模拟参数寻优的过程中能够达到良好的精度。
随着大规模水文模拟需求的不断提高,如何解决计算需求问题逐渐成为水文研究的一个热点。SWAT(soil and water assessment tool)模型在进行大规模水文模拟时有着良好的适应性与准确度,但其敏感度分析模块由于计算量过高,计算时长往往长达数月之久。为了加快SWAT敏感度分析的运行速度,针对SWAT敏感度分析模块的特点,基于MPI提出了一种高效的主—从式并行计算框架,并在此框架的基础上,通过将正演过程并行化,在敏感度分析的主—从并行框架中引入通信子空间的操作,将并行化的正演与主—从式的外层并行框架相结合,得到一种混合式的敏感度分析并行框架,大大提高了对参数集合的敏感度分析速度,将SWAT敏感度分析模块使用的处理器数量从原始的单核串行一跃提升到百核的数量级。最后通过天山北坡流域的模拟验证了此并行框架的可行性。
Evapotranspiration(ET) and its controlling mechanism over the desert riparian forests in arid regions are the important scientific basis for the water resources managements of the lower reaches of the inland rivers of China. Nearly three years of continuous measurements of surface ET, soil water content at different depths and groundwater table over a typical Tamarix spp. stand and a typical Populus euphratica stand were conducted in the lower reach of the Tarim River. The ET seasonal trends in the growing season were controlled by plant phenology, and ET in non-growing season was weak. The diurnal variations of ET resulting from the comprehensive effects of all atmospheric factors were significantly related with reference ET. The spatial pattern of ET was determined by vegetation LAI, more vegetation coverage, more ET amount. Groundwater is the water source of surface ET, and the soil water in shallow layers hardly took part in the water exchange in the groundwatersoil-plant-air system. The temporal processes of ET over the Tamarix stand and the Populus stand were similar, but the water consumption of the well-grown Populus euphratica was higher than that of the well-grown Tamarix spp. Further analysis indicates that plant transpiration accounts for most of the surface ET, with soil evaporation weak and negligible; groundwater table is a crucial factor influencing ET over the desert riparian forests, groundwater influences the processes and amounts of ET by controlling the growth and spatial distribution of desert riparian forests; quantifying the water stress of desert riparian forests using groundwater table is more appropriate, rather than soil water content. Based on the understanding of ET and water movements in the groundwater-soil-plant-air system, a generalized framework expressing the water cycling and its key controlling mechanism in the lower reaches of the inland rivers of China is described, and a simple model to estimate water requirements of the desert riparian forests is presented.
