Mapping agricultural responses to water supply shocks in large irrigation systems, southern India

Irrigated agriculture experienced a water supply shock during a drought in southern India in 2002-2003. In this paper, hotspots of agricultural change were mapped and put in the context of hydrology and water management. Time series of MODIS imagery taken every eight days before (2001-2002) and during (2002-2003) the supply shock were combined with agricultural census data to document changes in cropping patterns in four large irrigation projects in the downstream sections of the Krishna and Godavari River basins (total command area 18,287 km²). The area cropped in rice in the four irrigated command areas decreased by 32% during the drought year, and rice production in the two districts that experienced the largest flow reductions fell below production levels of 1980. The irrigation project that showed the largest change in double cropped area (−90%) was upstream of the Krishna Delta. In the Krishna Delta, large areas changed from rice-rice to rice-gram double cropping. Historical water management contributed to the vulnerability of rice production to drought: the main reservoir in the system was drained to dead storage levels by the end of each growing season over 1968-2000, with little carryover storage. The land cover change maps suggested that the lower Krishna Basin has experienced a “hard landing” during basin closure, and revised management strategies that account for the new flow regime will be required to maintain agricultural production during droughts.     

MODIS土地利用/覆被多时相多光谱决策树分类

利用MODIS多时相与多光谱结合,尝试探讨低成本、高精度的北京土地利用/覆被实时获取方法。首先根据归一化植被指数（NDVI）的均值、标准差建立了研究区各地类的典型NDVI时间序列曲线,进而提取了6个可以反映区域物候模式、植被生长速率等信息的分类参数;然后对反映地表土壤信息较多的3月份多光谱影像进行主成分变换,选取第一主成分（PC1）作为辅助分类参数;最后基于分类回归树（CART）算法进行监督决策树分类。经SPOT-5影像验证,分类总体精度达到83%,Kappa系数为0.769,PC1辅助分类后总体精度提高     

近20年华北粮食主产区冬小麦种植面积遥感快速提取研究

【目的 】基于遥感数据,研究快速提取华北粮食主产区近20年（2001—2020年）冬小麦种植面积的方法,生成准确的长时间序列冬小麦面积遥感产品,为政府决策机构和科研单位的工作提供数据支持。【方法 】文章基于经过滤波重构的MODIS植被指数产品,分析了研究区不同纬度下冬小麦在整个生长季中的时序特征,考虑到不同区域冬小麦物候差异,提出了一种关键生长季时序NDVI曲线匹配的方法,在无样本的条件下,快速提取冬小麦面积。通过使用统计年鉴进行面积验证,并结合目视解译的样本和高分辨率数据哨兵2号提取的结果,计算混淆矩阵并进行精度评价。【结果 】与2001—2018年的统计年鉴数据对比,平均相对误差为16.1%;与目视解译和哨兵2号分类结果中的6 459个采样点的精度评价相比,总体精度达到87.4%,kappa系数为0.61。【结论 】根据冬小麦的物候特征,通过提取NDVI的时序特征并采用时序NDVI曲线匹配算法,可以快速准确地提取华北粮食主产区冬小麦的种植面积和分布情况。     

基于GF-1影像NDVI年度间相关分析的冬小麦面积变化监测

为实现区域冬小麦种植面积变化的快速监测,减少监测难度,提高监测效率和精度,该文提出一种基于年际NDVI相关关系的监测方法(relationship analysis of normal difference vegetation index,rNDVI)。选择河北省黄骅市、孟村县、海兴县3个县市为研究区,基于2014年4月14日、2017年4月26日两个时期的GF-1/WFV数据,基于rNDVI方法,通过将样本点两年度的NDVI值构建二维空间,采用最小二乘法拟合的方法获得不变地物点的上下包络线方程,进而得到冬小麦变化区域的监测阈值,提取冬小麦种植增加和减少区域,实现对研究区域的变化监测。结果表明,采用rNDVI算法总体精度分别为90.60%,Kappa系数为0.84,相比传统的先最大似然分类后再提取冬小麦种植变化区域的方法,总体精度与Kappa系数分别提高了6.6个百分点和16.7%。对冬小麦增加区域、冬小麦减少区域的变化监测结果进行分析,发现基于rNDVI的变化监测方法可以有效提高裸地、线状道路、破碎的冬小麦地块等区域的变化识别能力,提高监测精度。同时分别利用2014年3月1日和2017年3月12日、2014年5月17日与2017年5月20日两对GF-1/WFV数据进行基于rNDVI的冬小麦变化区域监测,结果表明3月份的监测精度较低,主要是由于3月份冬小麦长势尚不明显,5月份与4月份的总体精度相近,主要是由于5月份冬小麦NDVI已较高,易于识别。上述研究结果表明,基于rNDVI的冬小麦变化快速监测方法可以有效监测区域冬小麦种植面积的变化情况,算法简单高效,且能够在种植结构相对单一的冬小麦分布区域保持较高精度,能够满足农情遥感监测信息快速获取的需要。     

基于时序NDVI图谱库提高土地覆盖分类精度的方法

为提高MODIS土地覆盖产品的分类精度,该文以河南省为试验区,首先将MODIS土地覆盖产品(MCD12Q1)分为高精度区域和低精度区域,然后通过构建时序NDVI图谱库并利用图谱曲线相似性测定方法,改进MCD12Q1低精度区域的分类精度。结果表明:1)时序NDVI是土地覆盖的重要分类特征,二者之间具有较强的关联性。2)利用时序NDVI图谱库能够明显提高MODIS土地覆盖产品的分类精度,改进后的MCD12Q1的总体分类精度分别由72.76%(比较评价)、64.52%(样本评价)提高到83.05%和81.72%。3)不同土地覆盖类别精度提高的程度不同,林地、草地、耕地、人工地表以及水体的生产者精度分别提高35.36%、29.51%、2.98%、6.96%和6.11%。4)对于判定时序NDVI曲线相似度的2种具体方法而言,最小距离法(minimum distance,MD)总体上优于光谱角度匹配法(spectral angle mapper,SAM)。综上,保留现有土地覆盖产品中分类精度较高的部分,基于时序NDVI图谱库改进分类精度较低的部分,是提高现有土地覆盖产品分类精度的有效方法。     

基于高分辨率遥感影像的农作物灾损评估研究

【目的】2016年湖北省部分地区遭受洪涝灾害,导致该省农作物产量下降。利用遥感手段进行灾损评估研究可快速、准确获取灾后农作物的受灾程度与减产量,给农民提供实时、准确的参考数据,以减少农民的经济损失。【方法】基于2015年7月与2016年7月的灾前灾后Landsat-8影像,利用面向对象分类方法和分层抽样方法对湖北省荆州市与荆门市部分地区的农作物进行受灾面积提取,统计农作物受灾面积总量;对灾前灾后农作物的NDVI值进行对比,根据灾前灾后的NDVI差值进行受灾等级划分,以此得到农作物的受灾等级。【结果】(1)由连续暴雨引起的洪涝灾害导致研究区农作物都有不同程度的受损,采用了面向对象分类方法精度指标的Kappa系数为0.849 5,以此分类统计出的受灾农作物总面积为635.838 km~2;(2)受灾严重的区域主要是依傍河岸和地势低洼处。荆州市的弥市镇虎渡河附近的农作物受灾等级最高,受灾最为严重,几乎完全被洪水淹没;而荆州市马良镇的杨家湾因为地势低以及对农作物的灾害预防措施做得不够,也造成大量的农作物被淹没,甚至出现了绝收现象。【结论】对于农作物灾后监测,有关部门要借助遥感手段及时向农民提供数据支持,将农民的经济损失降到最低,最大程度地保障农民的生活。     

Integrating space and time: a case for phenological context in grazing studies and management

In water-limited landscapes, patterns in primary production are highly variable across space and time. Livestock grazing is a common agricultural practice worldwide and a concern is localized overuse of specific pasture resources that can exacerbate grass losses and soil erosion. On a research ranch in New Mexico with average annual rainfall of 217 mm, we demonstrate with a quantitative approach that annual seasons vary greatly and examine foraging patterns in Angus-Hereford (Bos taurus) cows. We define five seasonal stages based on MODIS NDVI: pre-greenup, greenup, peak green, drydown and dormant, and examine livestock movements in 2008. Daily distance traveled by cows was greater and foraging area expanded during periods with higher precipitation. A regression model including minimum NDVI, rainfall and their interaction explained 81% of the seasonal variation in distance traveled by cows (P<0.01). Cows explored about 81 ha·d⁻¹ while foraging, but tended to explore smaller areas as the pasture became greener (greenup and peak green stages). Cows foraged an average of 9.7 h daily and spent more time foraging with more concentrated search patterns as pastures became greener. Our findings suggest that phenological context can expand the capacity to compare and integrate findings, and facilitate meta-analyses of grazing studies conducted at different locations and times of year.     

Impact of short‐term climate effects and sea ice coverage variation on Japanese scallop aquaculture in Saroma Lake,Japan

Saroma Lake is the southernmost lagoon exhibiting seasonal sea ice coverage in the Northern hemisphere and one of the most important aquaculture areas for Japanese scallops (Mizuhopecten yessoensis). Under conditions of adequate food and appropriate temperature, scallops grow well from the time of ice melting in April to harvesting starting in July. However, over the past decade, Saroma Lake frequently has not been completely covered by ice in winter, and the production of Japanese scallops has shown significant changes. Therefore, this study integrated data from satellite remote sensing, buoys, and in situ observations with climatic events [the winter East Asian Monsoon (EAM) and El Niño/La Niña–Southern Oscillation events] to investigate the impact of ice coverage variations on scallop growth in Saroma Lake between 2007 and 2015. Daily ice conditions were detected using Moderate Resolution Imaging Spectroradiometer false colour images and an unsupervised classification method. The results indicated that EAM strongly influenced the ice coverage variation in Saroma Lake through their effects on temperature during winter. Ice coverage variations show a strong correlation with water temperature and spring phytoplankton blooms, which are the two most important environmental factors for scallop growth. In addition, extreme climate events could cause water temperature anomalies (as in 2015) which are unfavourable for scallop growth. Monitoring ice conditions should be considered when developing plans and management strategies for scallop aquaculture in Saroma Lake.     

基于遥感影像的县域冬小麦种植信息提取

冬小麦是陕西省关中地区重要的粮食作物。本研究以陕西省武功县为例,使用Landsat8 OLI影像作为数据源,综合使用NDVI阈值分割和决策树分类技术提取了武功县2017年冬小麦种植区域,并对各乡镇冬小麦种植面积和空间分布信息进行统计分析。结果表明：（1）在本研究区,冬小麦越冬期结合油菜开花期是利用遥感影像提取冬小麦种植信息的最佳时期组合;（2）运用NDVI阈值分割法结合决策树分类从OLI影像提取冬小麦种植面积的精度达到93%以上;（3）武功县2017年冬小麦种植面积为21 181.97 hm²,下辖各乡镇中长宁和贞元两镇冬小麦种植面积较大,分别占总面积的24%和19%;在县域尺度上,使用遥感技术提取的冬小麦种植面积和空间分布信息是准确可靠的。     

基于MODIS数据的山东省农作物旱情遥感监测报告

干旱胁迫能影响作物生长发育,给农业生产带来巨大危害.为了及时准确地掌握农作物旱情,本研究以山东省夏季农作物为研究对象,采用高等时间分辨率MODIS影像,建立了普适性较高的归一化差值植被指数,并通过观测分析其时间与空间维的变化情况判断农作物长势及旱情,从而建立了一种简便、高效的利用遥感技术监测旱情的方法.该方法适用于营养生长阶段的农作物,原理简单,操作方便,易于进行技术推广.