紫茎泽兰是目前对我国危害最为严重的外来入侵植物之一，对我国农业、畜牧业等产业都造成巨大损失，并对人类及牲畜的健康有一定威胁。紫茎泽兰空间分布的有效监测，可为有关部门在当地对紫茎泽兰的治理与预防提供重要科学依据。传统的监测多采用实地面积统计等方法，费时费力且经济成本高，而用GIS的空间分析方法可得出紫茎泽兰的适生区；用遥感技术的方法可高效、大尺度地对紫茎泽兰的分布进行动态监测；但对紫茎泽兰空间分布的研究一般多分别单独采用这两种技术方法，而本文将GIS与遥感技术结合起来对紫茎泽兰的空间分布进行监测。在预测研究方面，本文选取了目前预测效果最好的两个模型—GARP模型与Maxent模型，同样以云南省为研究区，对紫茎泽兰在云南省的适生性及潜在地理分布进行预测。本文的研究结果为后续研究分析打下基础，主要研究结论如下： 1. 紫茎泽兰的光谱在波长674nm附近有一个吸收谷，通过对紫茎泽兰在该波长附近的光谱特征参数研究发现，紫茎泽兰有其独特的光谱特征：紫茎泽兰在波长674nm附近吸收深度为0.2618~0.4559，吸收宽度为87~138nm，左面积为53.6349~172.1831，右面积为12.2522~55.6687，吸收对称性为-0.7410~ -0.4904；而紫茎泽兰周围典型地物在波长674nm 附近的光谱吸收特征，其中土壤、岩石等背景目标在波长668nm-676nm范围内不存在有明显的吸收谷，而周围健康植被在该波段范围内有一个明显的吸收谷，且该吸收谷的吸收深度为2.5590~8.1515，吸收宽度为86~116nm，左面积为550.9244~1069.075，右面积为140.4064~253.3453，吸收对称性为-0.7429~ -0.5327，表明紫茎泽兰与其周围典型地物光谱特征参数有一定差异性； 2. 通过对紫茎泽兰及其周围典型地物的光谱曲线的微分特征对比分析可知，开白花的紫茎泽兰与其它植被存在很大差异：紫茎泽兰光谱微分曲线在波长400~450nm存在一个明显的波峰，并在波长420nm处达到峰值，其范围大约为0.05~0.35；而其它植被在该区域的光谱微分曲线近似为一直线，且在波长420nm处得值大约为0.002~0.03； 3. 基于紫茎泽兰实地光谱测量数据，分析得到基于遥感影像提取紫茎泽兰空间分布的最佳植被指数组合，是取 与 的交集。本文选取云南省为研究区，建立紫茎泽兰生境因子层，结合遥感影像紫茎泽兰的提取信息，得到紫茎泽兰确定区、紫茎泽兰疑似区、紫茎泽兰适生区以及紫茎泽兰非适生区。通过与2008年紫茎泽兰在云南省入侵面积统计的比较可知，基于GIS与遥感技术结合的方法监测紫茎泽兰分布与实际情况比较吻合； 4. 通过GARP模型与Maxent模型对紫茎泽兰在云南省的潜在分布进行了预测，并与GIS与遥感技术结合的监测结果比较可知，预测结果与监测结果比较吻合，且预测模型之间、预测结果与监测结果都可相互参照，经由实际验证得到更为精确的结果。有关部门对紫茎泽兰的确定区与严重威胁区应加大防除力度，对于紫茎泽兰的疑似区与适生区应加强预防机制，而对其非适生区控制也不能放松警惕，以防止紫茎泽兰经适生性变化而适宜该区或者入侵后再向其它相邻适生区扩散。

Eupatorium adenophorum Spreng(EAS) is a alien plant currently one of the most serious harm to our country. EAS brought massive losses to agriculture and animal husbandry industry in our country, and threated human and animal health. Monitoring the distribution of EAS may support our country the important scientific basis for related department, who implement governance and prevention to EAS. Generally, the traditional monitoring adopt the statistics method, which is time-consuming and costly. We mapped the suitable areas of EAS based on the spatial analysis method of GIS. We also might implement dynamic monitoring to EAS efficiently on large-scale with Remote Sensing technology. Usually people use these two methods respectively, but in our study we combined GIS and Remote Sensing technology together to monitoring the distribution of EAS. In prediction research, we selected two models—GARP and Maxent, which have the best prediction results currently, to predict the potential geographical distribution of EAS and choose its suitable index. This may lay a foundation for following-up research. This main research conclusions of our study are as follows: 1. EAS has a absorption spectrum at the wavelength of 674nm, we found that EAS has its unique spectral characteristic through the study of spectral characteristic parameters of EAS: the apsorption depth is 0.2618~0.4559, the absorption width is 87~138nm, the left area is 53.6349~172.1831, the right area is 12.2522~55.6687, the absorption asymmetry is -0.7410~-0,4904; spectral characteristic parameter of the typical features surrounding EAS at the wavelengh of 674nm has a great difference from EAS: soil and rock hasn’t a obvious absorption spectrum at this band range, whereas the spectral characteristic parameters of health vegetations around EAS are: the absorption depth is 2.5590~8.1515, the absorption width is 86~116nm, the left area is 550.9244~1069.075, the right area is 140.4064~253.3453, and the absorption asymmetry is -0.7429~ -0.5327; 2. Compared the spectral characteristic between EAS and the typical features surrounding EAS, white flower made the spectral differential of EAS is different from other plants: there is a obvious spectrum peak at wavelength range of 420nm of the spectral differential of EAS, and reach the peak at wavelength of 420nm, the value of this peak is about 0.05~0.35; whereas the spectral differential of other plants is approximation for a straight line and the value of the peak is about 0.002~0.03; 3. Through the analysis of EAS based on spectrum measurement data, we get the best vegetation index combination to extra the information of the space distribution of EAS: the intersection of and . We selected Yunnan Province as the study area, established the habitat factors layer of EAS, combined with extract result from remote sensing images, finally mapped EAS sure area, EAS suspected area, EAS suitable area and EAS non-suitable area. Compared with invasion area statistic of EAS in yunnan province year 2008, the monitoring method combined with GIS and remote sensing technology is consistent with the actual situation. 4. We predicted the potential geographical distribution of EAS in Yunnan Province using Maxent model and GARP model, compared with monitoring results, we found the predict result is consistent with the monitoring result. Relevant departments should launch with different levels of management and prevention work according to EAS sure area, EAS suspected area, EAS suitable area and EAS non-suitable area.