基于多时空分辨率降水数据的黄河源区径流模拟研究

发布时间：2018-05-22 15:19

本文选题：TRMM + 黄河源区　；参考：《中国地质大学(北京)》2017年硕士论文

【摘要】：地处青藏高原东部的黄河源区属于典型的寒旱区,近年来的观测数据表明,在全球气候变暖的条件下,该区域的降水特征产生了一定变化,因此研究降水对其径流过程产生的影响对于揭示水文循环以及对气候的响应机制具有重要的科学意义。本文选用TRMM遥感降水数据,采用相关系数、相对误差和均方根误差三个指标,从时间尺度和空间尺度分析了TRMM数据在黄河源区及周边地区的适用性,并利用TRMM数据分析了源区降水的时空变化特征;在数据相关性分析的基础上,选用了与TRMM降水数据具有较强相关性的MODIS NDVI植被数据,通过降尺度方法将空间分辨率为0.25°的TRMM数据提高到1 km,同时利用比例指数法将降尺度的TRMM年数据转化为月降水数据;在以上工作的的基础上,分别将降尺度的TRMM数据和分辨率为0.5°的格点降水数据作为输入变量,运用SRM融雪径流模型对源区2010年的径流进行模拟,分析两种不同降水数据对模型精度的影响;在模拟中,还运用源区周边站点降水、气温数据改进了模型参数临界温度。研究结果表明,TRMM降水数据整体高于气象站点实测值,在温度较高的4-10月的数据精度优于其它低温月份;随着月、季、年时间尺度的增大,TRMM数据的精度也随之提高;源区降水的空间分布特征表现为从东南向西北递减,降水受局部地形的影响并不随高程升高而线性增大;研究区内NDVI和TRMM在年尺度的相关系数为0.81,且在分辨率为0.75°时具有最好的相关性;降尺度后TRMM降水数据经检验与气象站观测数据的相关系数略低于原始TRMM数据,相对误差和均方根误差变化不大;利用比例指数法得到的降尺度的TRMM月数据经检验在玛多、达日等多数站点的精度有一定程度的增大;以气象站降水数据为输入变量的SRM模拟结果优于0.5°格点降水数据和降尺度TRMM数据;模型在融雪期的模拟精度高于非融雪期,在融雪期内径流峰值的模拟精度较低;运用改进后的临界温度对模拟流量略有降低,说明将临界温度调整后积雪转化的径流量变小,导致模型精度降低。
[Abstract]:The source region of the Yellow River located in the eastern part of the Qinghai-Xizang Plateau is a typical cold and arid region. The observed data in recent years show that under the condition of global warming, the precipitation characteristics of the region have changed to a certain extent. Therefore, it is of great scientific significance to study the influence of precipitation on the runoff process in order to reveal the hydrological cycle and the response mechanism to climate. In this paper, TRMM remote sensing precipitation data are used to analyze the applicability of TRMM data in the source region of the Yellow River and its surrounding areas from the time scale and the spatial scale, using the correlation coefficient, relative error and root mean square error. On the basis of data correlation analysis, MODIS NDVI vegetation data with strong correlation with TRMM precipitation data are selected, and the characteristics of temporal and spatial variation of precipitation in source region are analyzed by using TRMM data, and MODIS NDVI vegetation data with strong correlation with TRMM precipitation data are selected based on the data correlation analysis. The TRMM data with a spatial resolution of 0.25 掳are raised to 1 km by downscaling method, and the annual downscaling TRMM data are converted into monthly precipitation data by using the proportional index method. The downscale TRMM data and the grid precipitation data with a resolution of 0.5 掳are used as input variables, and the runoff of the source region in 2010 is simulated by using the SRM snowmelt runoff model. The effects of two different precipitation data on the accuracy of the model are analyzed. The critical temperature of the model parameters is improved by using the precipitation and temperature data around the source area. The results show that the precipitation data of TRMM are higher than the measured values of meteorological stations, and the precision of the data in April to October is better than that in other low temperature months, and the accuracy of TRMM data increases with the increase of monthly, seasonal, annual time scale, and so on. The spatial distribution of precipitation in the source region is decreasing from southeast to northwest, and the precipitation does not increase linearly with elevation. The correlation coefficient between NDVI and TRMM in the study area is 0.81at the annual scale and has the best correlation at the resolution of 0.75 掳, and the correlation coefficient between the TRMM precipitation data and the observational data of meteorological stations after downscaling is slightly lower than that of the original TRMM data. The relative error and root mean square error have little change, and the monthly data of downscaling TRMM obtained by the proportional index method have some degree of increase in the accuracy of most stations, such as Mador, Dadi and so on. The SRM simulation results using precipitation data of meteorological station as input variables are better than that of 0.5 掳lattice precipitation data and downscale TRMM data, and the simulation accuracy of the model in the snowmelt period is higher than that in the non-snowmelt period, and the simulation accuracy of the peak runoff value in the snowmelt period is lower. The modified critical temperature is used to reduce the simulated flow slightly, which shows that the runoff of snow conversion after adjusting the critical temperature is smaller, which leads to the decrease of model precision.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TV121

【相似文献】