雅砻江上游积雪面积变化与径流关系研究

发布时间：2019-03-22 20:15

【摘要】：雅砻江是长江上游金沙江最大的一级河流,是中国水能资源最富集的河流之一,也是我国水电规划建设的十三大水电基地之一。甘孜以上地区作为雅砻江的河源,其河川径流特征变化将会从长期和短期两方面影响雅砻江中下游水资源的开发和利用情况。分析和预测甘孜段径流的变化规律,不仅有助于深入了解雅砻江流域的水资源特性,为流域内水资源合理开发利用提供依据,也有助于雅砻江流域梯级水电站调度的径流预报提供参考。本文将雅砻江流域甘孜以上地区作为研究区,以2009-2014年HJ数据和2000-2014年MODIS积雪产品、甘孜水文站径流数据、研究区内的清水河、石渠和甘孜气象站的逐日降水和气温数据作为数据资料。首先,对比研究已有HJ数据的积雪识别方法的优缺点,提出了改进的HJ数据积雪识别方法,采用GF数据对其进行精度评价,并作为参考值对MODIS积雪产品进行精度评估;其次,在获取积雪面积数据的基础上,分析了研究区径流、积雪覆盖面积、气温和降水的不同阶段的年内、年际变化特征;再次,利用相关分析和多元回归分析法分析了各阶段径流及其影响因素的相关方向和相关程度,探讨了枯水期和融雪影响期径流变化对积雪、气温和降水变化的响应程度,并对径流进行了预测分析;最后,实现研究区HBV水文模型的模拟,并根据相关分析结果,增加积雪覆盖面积的输入改进水文模型,提高径流的模拟预测精度。结果表明,经同时相的GF数据精度验证,本文提出的仅用环境星CCD数据作为数据源的积雪识别算法分类精度在90%以上,可以快速、有效地提取研究区积雪面积。雅砻江甘孜以上地区径流曲线呈“单峰型”,属于雨水、融雪水混合补给类型的河流,2000-2014年间融雪影响期和汛期径流均表现为显著增加趋势。研究区积雪累积和融化幅度较大,平均每年融化43332.33 km2的新雪,枯水期平均积雪面积年际波动较大,无明显的变化规律,融雪影响期和汛期积雪面积均呈现微弱的上升趋势。枯水期降水有微弱的增加趋势,高海拔地区的气温增加趋势显著,整体气温的增加随海拔的降低而稍有减缓;融雪影响期的降水除甘孜附近无明显变化趋势,研究区其他地区均有显著增加;汛期气温相对稳定,降水呈微弱的上升趋势。研究区枯水期径流主要通过地下水补给,与气温呈中度正相关,与降水和积雪面积无显著相关性;融雪影响期径流与降水和气温表现为显著正相关,与积雪覆盖面积表现为显著负相关,降水、气温和积雪面积的弹力系数分别为0.289、0.779、0.004,且均通过90%置信度的检验,其中,降水增加对径流变化的影响程度为58.11%,冬季积雪覆盖面积和气温对径流的影响程度和为30.85%;汛期径流与降水表现为中度正相关,降水的弹力系数为0.219,对径流的影响程度为89.07%。融雪影响期径流与冬季积雪面积、上月均气温和上月均降水的回归结果以及枯水期径流与上月径流的回归结果在95%的置信度下通过了回归方程的五种检验,且拟合度R2均在0.80以上,可用于预估枯水期和融雪影响期的月均径流。将2005~2007年作为率定期,2010~2013年作为验证期,应用HBV水文模型对甘孜站日、月径流进行模拟,验证期模型效率系数为0.8153和0.8702,RE为-8.399和-8.4105,说明HBV水文模型总体上可较好地模拟径流变化。改进后的水文模型效率系数为0.8726和0.9250,RE为-4.3313和-4.3725,模型效率系数和误差范围均有所改善,可以更好预测和模拟研究区径流。
[Abstract]:The Yamen River is the largest river in the Jinsha River in the upper reaches of the Yangtze River. It is one of the most abundant rivers in China's hydropower resources. It is also one of the ten major hydropower bases in China's hydropower planning and construction. As the river source of the Yamen River, the change of the river runoff characteristics will influence the development and utilization of the water resources in the middle and lower reaches of the Yamen River from the long-term and the short-term. The analysis and prediction of the change law of the runoff in the river basin can not only help to understand the characteristics of the water resources in the Yalanjiang river basin, provide the basis for the rational development and utilization of the water resources in the basin, and also help to provide the reference for the runoff prediction of the cascade hydropower station in the Yabjiang river basin. As a research area, the daily precipitation and air temperature data of Qingshuihe, Shiqu and Ganzi meteorological station in the study area are used as data for the study area of the area of Ganzi and above in the Yalanjiang river basin. The daily precipitation and air temperature data of the Clear Water River, the stone canal and the Ganzi meteorological station in the study area are taken as data by the HJ data in 2009-2014 and the MODIS snow-covered products of 2000-2014 and the runoff data of the Ganzi hydrological station. First of all, the advantages and disadvantages of the snow cover recognition method of HJ data are compared, the improved HJ data snow-snow identification method is proposed, the precision evaluation is carried out by using the GF data, and the MODIS snow product is evaluated with the reference value as a reference value, and secondly, on the basis of acquiring the snow area data, The annual and interannual variation of runoff, snow cover area, air temperature and precipitation in the study area were analyzed, and the correlation direction and degree of the runoff and its influencing factors were analyzed by correlation analysis and multiple regression analysis. In this paper, the response degree of the runoff change on the change of snow, air temperature and precipitation during the dry season and the snow-melting period is discussed, and the runoff is predicted and analyzed. Finally, the simulation of the HBV hydrological model in the study area is realized, and the input and improved hydrological model of the snow cover area is increased according to the relevant analysis results. And the simulation prediction precision of the runoff is improved. The results show that, with the verification of the data accuracy of GF data at the same time, the classification accuracy of the snow-snow recognition algorithm only with the environmental star CCD data as the data source is above 90%, and the area of the snow in the study area can be effectively and effectively extracted. The runoff curve of the area above the Ganzi River in the Yafangjiang River is a "unimodal", which belongs to the mixed recharge type of the rainwater and the snow-melting water. In the period 2000-2014, the snow-melting-effect period and the flood-season runoff show a significant increase in the trend. The accumulated and melting range of the snow in the study area is larger, the average annual melting of the new snow is 43332.33km2, and the average snow cover area in the dry season is relatively large, there is no obvious change law, and the snow-melting effect period and the snow cover area in the flood season show a weak upward trend. There is a slight increase in the precipitation in the dry season, the increase of the temperature in the high-altitude area is significant, the increase of the whole air temperature is slightly slower with the decrease of the altitude, and the precipitation of the snow-melting-effect period has no obvious change trend in the vicinity of the Ganzi, and there is a significant increase in the other parts of the study area; The temperature of the flood season is relatively stable and the precipitation is in a weak upward trend. The runoff in the dry season of the study area is mainly supplied through the groundwater, which is moderately positive with the temperature, and has no significant correlation with the precipitation and the area of the snow. The runoff and the precipitation and the air temperature show a significant positive correlation with the snow-melting effect, and the rainfall is negatively correlated with the snow cover area. The elastic coefficient of the air temperature and the snow cover area is 0.289, 0.779, and 0.004, and all passes the test of 90% confidence, among which, the effect of the precipitation increase on the runoff change is 58.11%, and the influence degree of the snow cover area and the air temperature on the runoff is 30.85%; The runoff and precipitation in the flood season are moderate and positive, the elastic coefficient of the precipitation is 0.219, and the effect on the runoff is 89.07%. The regression results of the annual runoff and the winter snow cover area, the upper monthly temperature and the last monthly precipitation, and the regression results of the runoff in the dry season and the last-month runoff have passed the five tests of the regression equation with 95% confidence, and the fitting degree R2 is more than 0.80, Can be used for estimating the monthly average runoff of the dry season and the snow-melting influence period. During the period from 2005 to 2007, as the period of verification, the period of 2010 to 2013 was used as the verification period, and the day and monthly runoff of the Ganzi station were simulated by using the HBV hydrological model. The efficiency coefficient of the model was 0.8153 and 0.8702, and RE was-8.399 and-8.4105, which indicated that the hydrological model of the HBV could well simulate the runoff change. The improved hydrological model efficiency coefficient is 0.8726 and 0.9250, RE is-4.3313 and-4.3725, the model efficiency coefficient and the error range are all improved, and the study area run-off can be better predicted and simulated.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P426.635;P333.1

【参考文献】