Land Evaluation in Semi-arid Region Using Remotely Sensed Derived Vegetation Index under Extreme Drought Event
To avoid encountering 504 error when downloading big files,please take CAJ download or PDF download by chapter as your first priority.
【Author in Chinese】 孙福军；
【Author's Information】 沈阳农业大学， 土地利用与信息技术， 2017， 博士【Subtitle】以辽宁省朝阳市为例
【Abstract in Chinese】 土地评价是科学合理利用土地资源的重要前提,也是土地管理的一项基础性工作。目前多是通过选择影响土地生产力的诸要素,根据各要素的特点确定评价标准,经综合计算来评定土地等级。表面上客观、精确,但各种因素对作物生长的影响程度不同,因素之间的相互关系也不十分明确,评价结果的准确性并不高。而且评价需要考虑的因素众多,调查工作量巨大,应用起来十分困难。不同区域自然条件不同,因此影响和制约土地生产能力的关键因素也有所不同,在半干旱地区,光照、热量、土壤等条件一般相对较好,影响土地生产能力的关键限制性因素主要是水分条件。作物的生长状况是其适应土地生产力的结果,是土地特性的综合反映和最直接的表现。极端干旱条件下,半干旱地区的作物长势也由于水分供应过少而呈现出极端状态,直接体现了区域土地生产能力。遥感技术具有快速、准确、可周期性观测和覆盖范围大等特点,已经成为开展大范围作物生长状态监测行之有效的技术方法,遥感植被指数是植被生长状态的最佳指示因子,在评估植物生物量等诸多方面应用广泛。因此,本研究以半干旱地区的朝阳市为例,利用高时间分辨率的遥感影像,通过极端干旱条件下的遥感植被指数表征作物的生长状态,根据作物长势差异进行土地等级评定;利用标准化降雨指数SPI分析了研究区近60年的干湿变化情况,分别以多年(2001～2015)数据、典型气候代表性年份(2008～2010)数据和极端干旱年份(2009)数据为基础,通过植被指数的相关计算及变化率的分析,获取朝阳地区作物长势及其空间分布情况的多种结果;开展野外实地调查,对基于遥感植被指数的几种评价结果、农用地分等定级结果和耕地地力评价结果进行验证和比较,判断基于极端干旱条件下遥感植被指数评价结果相对于其他结果的准确性和效果;探索建立一种准确、快速、高效的土地评价方法。结果如下:(1)通过现有资料分析表明,朝阳地处暖温带半干旱地区,影响土地生产能力的自然因素中,光照、温度和土壤条件相对都较好,而水分条件相对较差,有十年九旱之说。年平均降水量仅为500 mm左右,降水时空分布极不均匀,是该地区频发干旱的主要原因之一,春秋两季由于多风更容易发生干旱。因此,影响朝阳地区土地生产能力的主导限制因素是水分条件。(2)通过野外实地验证比较,以极端干旱年份(2009)数据和多年(2001～2015)数据获取的评价成果准确度相对较高,以典型气候代表性年份(2008～2010)数据获取的评价结果次之。所有基于遥感植被指数的评价结果准确度均优于农用地分等定级和耕地地力评价的结果。经综合比较,以极端干旱年份(2009)归一化植被指数(NDVI)作为依据直接进行等级划分的方法,最简单有效。(3)以极端干旱条件下归一化植被指数(NDVI)作为依据,以平均值和正、负一个标准差为间隔将其划分了4个等级(即种植玉米的适宜性等级),获取了以栅格形式表达的朝阳市土地评价成果图,将这种基于遥感植被指数的土地评价方法简称为VIBLE(Vegetation index based land evaluation)。(4)在半干旱地区应用该方法开展土地评价的一般程序是:根据气候资料判断极端干旱时期,选择作物生长关键时段(对于北方玉米是秋季),确定可供利用的遥感影像(MODIS影像时间分辨率高,基本能保证有数据可用,但空间分辨率低;高空间分辨率的影像可能在关键时期无可用数据),提取植遥感植被指数(可选择NDVI),进行作物长势差异等级划分。据此,可评定研究区域内的土地生产力等级。本研究所用土地评价方法简单,省时省力,快速准确,所得结果(土地的生产力高低及其空间分布情况)是用栅格形式表达的,可以客观反映出土地质量在空间上的渐变性,对土地利用、管理和规划等具有更现实的指导意义。该方法可为其他地区开展土地评价工作提供借鉴和参考。
【Abstract】 Land evaluation is an important prerequisite for scientific and rational use of land resources and also a basis of land management.Factors affecting land productivity were currently evaluated to classify land productivity according to the evaluation criteria derived from their characteristics.However,the effects of various factors on crop growth are different,and relationships between factors are not very clear.Therefore,previous land evaluation results were not accurate enough.In addition,the current land evaluation with many influencing factors under a huge workload survey is very difficult to apply.Different regions have different natural conditions,so key factors affecting and restricting the land production capacity are also different.In the semi-arid region,conditions including illumination,heat,and soils are generally good,whereas water mainly restrict the land productivity.The growth status of crops is the result of adaptation to land productivity,and is a comprehensive reflection and the most direct performance of the land characteristics.Under extreme drought conditions,crop growth in semi-arid areas also showed extreme state due to the lack of water supply,which indicates regional land productivity.Remote sensing technique has characteristics including quick,accuracy,periodic observation,and large coverage.It has become an effective method to monitor crop growth conditions.Remote sensing vegetation index is the best indicator of vegetation growth,and is widely used in the evaluation of plant biomass.An example of Chaoyang in semi-arid area is studied.The vegetation index of crop growth status in extreme drought conditions,from remote sensing images with high temporal resolution,was used to classify land productivity based on crop growth difference.Analysis of the vegetation index and its change rate on data of several continuous years(2001-2015),typical years(2008-2010),and extreme drought year 2009 were conducted to address results of crop growth and its spatial distribution in Chaoyang area.Comparisons of evaluation results between using the vegetation index based on land evaluation(VIBLE),agricultural land classification and gradation,and arable land productivity evaluation were done following a field survey to address the accuracy and validity of VIBLE.Finally,a quick,accurate,and efficient land evaluation method for land evaluation is explored and proposed.Some meaningful results were listed as follows:(1)Previous data of Chaoyang was analyzed.The results showed that Chaoyang is located in the warm temperate semi-arid region.Natural factors of illumination,heat,and soils are generally optimal,whereas water mainly restricts the land productivity.The average annual precipitation is only about 500 mm.The spatial and temporal distribution of precipitation is very uneven,which is one of main causes for frequent drought in the region.In addition,spring and autumn wind is more prone to cause drought.Therefore,water is the main restricting factor for land productivity.(2)The comparison between land evaluation results of VIBLE and observational points was conducted.The land evaluation accuracy of extreme drought year 2009 data and several continuous years(2001-2015),was higher than that using typical years(2008-2010)data.The accuracy of VIBLE land evaluation results was higher than that using agricultural land classification and gradation and arable land fertility evaluation.The method using normalized difference index(NDVI)of extreme drought year 2009 data to classify land was finally demonstrated to be simple and effective.(3)According to the difference of NDVI under extreme drought,the productive capacity of arable land in the study area was evaluated.Four levels of evaluation results were divided(e.g.suitability level for planting corn)using average values and standard deviation as intervals,and the grid land evaluation map was finally obtained.The land evaluation method based on vegetation index extracting from remote sensing data thus was proposed.(4)General processes of VIBLE were proposed and listed as follows.Extreme drought period according to climatic data was firstly identified;crop growth key period(autumn for the northern corn)was selected;available remote sensing images were determined(high temporal resolution MODIS images are available,but with low spatial resolution;high spatial resolution images may not be available at critical times);the vegetation remote sensing vegetation index was extracted to classify the crop growth;finally,land productivity in the study area was classified.The land evaluation method used in this study is simple,time-saving,fast,and accurate.The results(the productivity of land and its spatial distribution)are expressed in the form of grid,which can objectively reflect the spatial variability of land quality,and has a realistic guiding significance for land use planning and land management.This research will provide a valuable reference for land evaluation in other areas.