干旱区灌区尺度灌溉及作物水分利用效率遥感评价方法

发布时间：2018-05-20 20:28

本文选题：河套灌区 + 遥感　；参考：《清华大学》2016年博士论文

【摘要】：干旱区灌区灌溉水利用效率与作物水分利用率的定量评价是农业水资源高效利用的基础。论文以干旱区代表性灌区——内蒙古河套灌区主要农作区为研究对象,综合遥感信息、现场试验和实地调查信息,利用灌区遥感蒸散发模型、主要作物(玉米)种植分布遥感识别模型及其产量估算模型,分析了研究区2003-2012年蒸散发的时空变化、玉米种植分布及产量分布规律,并进一步定量评价了灌区灌溉及作物水分利用效率。论文首先采用两种双源蒸散发模型(TSEB模型和HTEM模型)结合MODIS数据计算2003-2012年河套灌区主要农作区生育期(4-10月)的蒸散发过程,并从点尺度及区域尺度对模型的精度进行评价。结果表明,HTEM模型在河套灌区具有更高的精度。为评价河套灌区的灌溉效率,论文以作物主要生育期(4-10月)灌溉地的蒸散发量扣除有效降水量作为灌溉水的有效利用量。将灌溉水有效利用量分别与灌溉净引水量和灌溉总引水量的比值定义为基于净引水量和总引水量的灌溉水有效利用系数(ηN和ηI)。对河套灌区2003-2012年灌溉水有效利用系数进行了分析和评价。结果表明,基于净引水量的灌溉水有效利用系数(ηN)近年来有增加的趋势,同时ηN随降水量和净引水量的减小而增大,净引水量减少对ηN的影响要大于灌区节水改造工程的影响。为了进一步探索河套灌区主要作物的空间分布,论文以玉米为例,通过实地调查结合MODIS遥感数据获得玉米NDVI时间序列,利用非对称Logistic曲线对玉米NDVI时间序列进行拟合,得到植被指数、物候指数的特征值。建立了基于物候-植被指数特征值的玉米分布遥感识别模型。结果表明,该模型可以适用于复杂种植结构下的多年间玉米分布识别,并且有较高的识别精度。最后,根据前文的研究结果,基于NDVI时序结果确定河套灌区玉米生育期,并分析其生育期内耗水时空分布规律。利用水分生产函数Jensen模型估算研究区玉米产量,进而求得河套灌区玉米水分利用效率,并分析2003-2012年间玉米水分利用效率时空变化规律。结果表明,基于NDVI时间序列求得玉米生育期与实际观测结果相吻合,利用Jensen模型模拟得到的玉米产量具有较高的精度。以上研究结果表明遥感信息及模型为干旱区灌区尺度灌溉及作物水分利用效率的定量评价提供了一条有效可行的途径。
[Abstract]:The quantitative evaluation of irrigation water use efficiency and crop water use efficiency in arid areas is the basis for efficient utilization of agricultural water resources. This paper takes the representative irrigation area of arid area-Hetao Irrigation area of Inner Mongolia as the research object, synthesizes the information of remote sensing, field test and field investigation, and makes use of the model of remote sensing evapotranspiration of irrigation area. The remote sensing recognition model and yield estimation model of main crop (maize) planting distribution were analyzed. The temporal and spatial variation of evapotranspiration, maize planting distribution and yield distribution in the study area from 2003 to 2012 were analyzed. Furthermore, the irrigation and crop water use efficiency in irrigated area were evaluated quantitatively. In this paper, two dual-source evapotranspiration models (TSEB model and HTEM model) combined with MODIS data were used to calculate the evapotranspiration process in the main farming areas of Hetao Irrigation area from 2003 to 2012. The accuracy of the model was evaluated from point scale and regional scale. The results show that the HTEM model has higher accuracy in Hetao irrigation area. In order to evaluate the irrigation efficiency of Hetao irrigation area, the evapotranspiration of irrigated land was deducted from the effective precipitation as the effective utilization of irrigation water. The ratio of the effective utilization of irrigation water to the net irrigation water diversion and the total irrigation water diversion is defined as the effective utilization coefficient of irrigation water (畏 N and 畏 I) based on the net water diversion and the total water diversion. The effective utilization coefficient of irrigation water in Hetao irrigation area from 2003 to 2012 was analyzed and evaluated. The results show that the effective utilization coefficient of irrigation water (畏 _ N) based on net water diversion is increasing in recent years, and 畏 _ N increases with the decrease of precipitation and net water diversion, and the effect of net water diversion on 畏 _ N is greater than that of water-saving reconstruction project in irrigation area. In order to further explore the spatial distribution of main crops in Hetao Irrigation area, the NDVI time series of maize was obtained by field investigation and MODIS remote sensing data, and the NDVI time series of maize was fitted by asymmetric Logistic curve. The characteristic values of vegetation index and phenological index are obtained. A maize distributed remote sensing recognition model based on phenological and vegetation index eigenvalues was established. The results show that the model can be used to identify the distribution of maize in multiple years under complex planting structure and has high recognition accuracy. Finally, based on the NDVI time series, the growth period of maize in Hetao irrigation area was determined, and the temporal and spatial distribution of water consumption in the growth period was analyzed. Jensen model of water production function was used to estimate maize yield in the study area, and then to obtain maize water use efficiency in Hetao irrigation area, and to analyze the temporal and spatial variation of maize water use efficiency from 2003 to 2012. The results showed that the maize growth period based on NDVI time series was consistent with the observed results, and the maize yield simulated by Jensen model had high precision. The results show that remote sensing information and model provide an effective and feasible way for quantitative evaluation of irrigation and crop water use efficiency in arid area.
【学位授予单位】：清华大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S274;S127

【相似文献】