基于SWAT模型的区域农业干旱模拟研究

发布时间：2018-11-24 19:54

【摘要】：在气候变暖日益加剧,全球范围极端干旱事件频发的大背景下,干旱越来越受到水文学家、气象学家、农业科学家的广泛关注。贾鲁河流域位于河南省中东部,是河南省粮棉生产基地之一,也是重度和中度干旱易发区,干旱灾害已成为当地粮食安全的主要威胁之一。本文以淮河支流贾鲁河控制流域为研究区域,在分析研究区农业条件和农业干旱成因的基础上,选取土壤相对湿度为农业干旱评价指标,评价了贾鲁河流域1992-2008年的农业干旱时空演变特征和周期演变规律;同时基于DEM数据、土地利用数据、土壤类型数据及水文气象数据构建了SWAT模型,模拟了贾鲁河流域2009-2014年农业干旱过程;通过农业干旱模拟结果与历史干旱事件的对比,验证了SWAT模型对贾鲁河流域农业干旱的模拟效果。主要研究成果如下:(1)贾鲁河流域内农业干旱发生频率高,季节性特征明显,流域内发生春旱和夏伏旱的频率较高,秋旱也是主要的干旱类型之一,其发生频率低于春旱和夏伏旱;降水与10-20cm土层土壤相对湿度在0.05置信水平上显著相关,土壤相对湿度既能够较全面的反映降水在干旱形成过程中的影响,又能较直观的表现作物受旱情况,是评估贾鲁河流域农业旱情的较合适的干旱指标。(2)贾鲁河流域初春、春末夏初、伏期和秋季土壤相对湿度年际变化均表现出9年左右的周期变化规律,流域中上游周期信号稳定,周期变化尺度多样,下游周期变化尺度单一。流域内初春旱和春末夏初旱的频次较伏旱和秋旱高,上游和下游春末夏初旱的发生频次高于初春旱,中游发生初春旱的频率较高,中上游旱情比下游旱情严重;1992年、1995年和2000年流域中上游发生了持续时间长、旱情严重的春夏连旱,中游旱情延续至秋季。(3)贾鲁河流域上游地区发生春末夏初旱的频率最高,其次为初春旱,秋旱发生频率低于初春旱发生频率,伏旱发生频率最低;中游地区发生初春旱的频率最高,其次为春末夏初旱,伏旱发生频率低于春末夏初旱发生频率,秋旱发生频率最低;下游地区发生春末夏初旱的频率最高,其次为初春旱,伏旱和秋旱发生频率相当,低于初春旱发生频率。(4)基于SWAT模型的区域农业干旱模拟对2009-2014年的农业干旱事件模拟效果良好,农业旱情评估结果与历史农业干旱事件基本吻合;同时,典型代表年模拟与实测的土壤相对湿度空间分布基本一致,能够更细致地反映农业干旱的空间分布特征,可进一步应用于研究区农业干旱的预测研究。
[Abstract]:Against the background of increasing global warming and frequent extreme drought events, drought has attracted more and more attention from hydrologists, meteorologists and agricultural scientists. The Jaru River Basin is located in the central and eastern part of Henan Province. It is one of the grain and cotton production bases in Henan Province and is also a severe and moderate drought-prone area. Drought disaster has become one of the main threats to local food security. Based on the analysis of the agricultural conditions and the causes of agricultural drought, the relative humidity of soil is selected as the evaluation index of agricultural drought in this paper, which is based on the study area of the Jialu River control basin, a tributary of the Huaihe River. The temporal and spatial evolution characteristics and periodic evolution of agricultural drought in the Jaru River Basin from 1992 to 2008 were evaluated. At the same time, based on DEM data, land use data, soil type data and hydrometeorological data, a SWAT model was constructed to simulate the agricultural drought process in the Jaru River Basin from 2009 to 2014. By comparing the results of agricultural drought simulation with historical drought events, the simulation effect of SWAT model on agricultural drought in the Jaru River Basin was verified. The main research results are as follows: (1) the frequency of agricultural drought in the Jaru River basin is high, the seasonal characteristics are obvious, the frequency of spring drought and summer drought is high, and autumn drought is one of the main drought types. The frequency of occurrence of drought is lower than that of spring drought and summer drought. There was a significant correlation between precipitation and soil relative humidity in 10-20cm soil layer at the confidence level of 0. 05. The soil relative humidity could not only reflect the effect of precipitation on the formation of drought, but also show the drought situation of crops intuitively. (2) the interannual variation of soil relative humidity in early spring, late spring, early spring and early summer and autumn showed a regular cycle of about 9 years. The period signal of the middle and upper reaches of the river basin is stable, the scale of periodic variation is various, and the scale of downstream cycle variation is single. The frequency of early spring drought and early spring drought in late spring and early summer is higher than that in late spring and autumn. The frequency of drought in upper and lower spring is higher than that in early spring. The frequency of early spring drought in middle reaches is higher than that in lower reaches. The drought in upper and upper reaches is more serious than that in lower reaches. In 1992, 1995 and 2000, a long duration of drought occurred in the upper and middle reaches of the basin, with severe drought in spring and summer and drought in the middle reaches into autumn. (3) the frequency of drought in the upper reaches of the Jaru River Basin was the highest in late spring and early summer, followed by the first spring drought. The frequency of autumn drought is lower than that of early spring drought, and the frequency of summer drought is the lowest. In the middle reaches, the frequency of early spring drought is the highest, followed by late spring and early summer drought, the frequency of drought occurrence is lower than that of late spring and early summer, and the frequency of autumn drought is the lowest. The frequency of drought in the lower reaches is the highest in late spring and early summer, followed by the first spring drought, and the frequency of drought in autumn is the same as that in autumn. (4) Regional agricultural drought simulation based on SWAT model has a good effect on simulating agricultural drought events in 2009-2014, and the results of agricultural drought assessment are basically consistent with historical agricultural drought events. At the same time, the spatial distribution of soil relative humidity in typical representative years is basically consistent with the measured soil relative humidity, which can reflect the spatial distribution characteristics of agricultural drought more carefully, and can be further applied to the prediction of agricultural drought in the study area.
【学位授予单位】：华北水利水电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S423

【参考文献】