南方丘陵地区水稻种植面积遥感信息提取的可行性分析

以浙江省为试验区，在地理信息系统支持下综合利用多种地理信息，探讨丘陵地区大面积提取水稻种植面积信息的可行性。开展了分类识别方法的比较试验及训练样点相对稳定性试验。针对丘陵地区的复杂地形，在数字化地形图的基础上，建立数字地形模型（ＤＴＭ），并衍生出地面坡度等地貌因子的数字化图像，结合ＮＯＡＡ／ＡＶＨＲＲ数据，进行分类。试验结果表明，传统的分类识别方法中，最大似然法的分类精度可满足业务化运行的要求；建立在混合像元分解基础上的模糊监督分类，有较高的分类精度和较好的稳定性，具有较强的适应性；地貌因子参与遥感影像的分类，不仅可以有效地提高丘陵地区水稻种植面积信息的提取精度，而且还可以使面积信息提取精度保持一定的稳定性，提高空间精度；为探讨丘陵地区水稻种植面积信息遥感提取的可靠性和客观性，在训练样点保持相对稳定的前提下，对１９９６年和１９９７年浙江省水稻种植面积进行测算，两年的数量精度均在９２％以上。     

基于面向对象分类的南方水稻种植面积提取方法

南方丘陵地区水稻种植具有分散、地块小、形状多样等特点,利用中低分辨率遥感数据提取水稻种植面积,难以满足精度要求。以SPOT5遥感影像为数据源,应用面向对象的分类方法提取了广西玉林市辖区晚稻种植面积。针对试验区不同稻作区的种植特点,选择其适合的尺度及参数进行多尺度影像分割,建立影像对象的层次结构,计算对象的光谱、几何及拓扑关系等特征,形成分类规则对不同稻作区进行信息提取。采用野外实地调查数据对分类结果进行类别和面积一致性检验,总体精度96.31%,Kappa系数0.9226,面积一致性精度99.92%。
     

利用HJ星遥感进行水稻抽穗期长势分级监测研究

对水稻长势进行遥感分级监测,制作能够直观反映水稻长势等级的遥感专题图,便于农业技术人员及时制定有效的田间管理措施,达到增产的目的。以江苏省泰兴市为例,利用HJ-A/B卫星遥感影像,提取水稻的种植面积并分析抽穗期水稻的长势情况。在利用GPS实地取样调查和建立解译标志的基础上,进行HJ-A/B卫星影像校正,人机交互式判读解译等操作,并将GPS样点数据校验贯穿到整个分类过程中,面积信息解译精度在90%以上。最后,利用归一化植被指数(NDVI)反演叶面积指数(LAI)数据信息,依据LAI数据进行水稻长势分级,制作了泰兴市水稻抽穗期长势分级遥感监测专题图。     

利用面向对象的分类方法提取水稻种植面积

结合广西水稻面积的遥感解译工作,应用SPOT4遥感数据和遥感处理软件ENVI,利用面向对象的遥感分类的方法提取早稻种植面积。分类结果表明,利用面向对象的分类方法有效解决逐像素分类结果的"椒盐"效应,获得比传统的像素级分类方法更高的分类精度,为广西水稻种植面积的自动提取提供了广阔的前景。     

基于多时相HJ卫星的冬小麦面积提取

我国环境与灾害监测预报小卫星HJ-1A/B具有较高的时间和空间分辨率,在作物种植面积提取和长势监测等方面具有较大优势。本文以江苏省姜堰市为研究区,根据冬小麦的物候规律和季相节律的差异性,选取返青期和拔节期两个生育期的HJ卫星影像,借鉴分层信息提取法原理,综合利用监督分类和非监督分类法,结合人机交互目视解译和实地定位调查等资料提取了姜堰市的冬小麦种植面积,总体面积提取精度达到90.22%,样点空间匹配精度为81.25%,实验基地空间匹配精度为80.34%。结果表明：HJ卫星能够用于提取南方地区冬小麦种植面积和长势监测,满足农情监测的需要,且利用多时相遥感影像能有效地增加信息量,实现信息互补,有助于提高监测精度。     

基于多时相NDVI及特征波段的作物分类研究

时相和光谱特征信息在农作物种植分类提取方面具有十分重要的应用价值。以黑龙江大型农场--友谊农场为研究区域,利用4景不同时相的TM和SPOT卫星遥感影像,提取相应时相的NDVI时间谱图像数据作为新波段信息,在分析地物目标在相应影像各波段上光谱和时间特征的基础上,设计了决策树分类算法,通过对待分类影像进行系列阈值分割和掩膜处理,成功提取黑龙江友谊农场的大豆、玉米和水稻的种植信息,分类总体精度达到98.67%。     

ASTER数据在棉花信息提取中的应用--以兵团农一师十六团为例

以2001、2003、2004三年秋季的ASTER遥感影像为数据源，结合相应地面实测数据，使用ERDAS软件，在几何精校正的基础上用最大似然监督分类法对兵团农一师16团农耕地进行棉花种植面积提取和棉花品种（长绒棉和陆地棉）分类，结果表明：棉花种植面积提取精度达98．21％，品种分类结果表明：长绒棉实际种植面积2964hm^2，分类面积2768．72hm^2，分类精度达93．97％；陆地棉实际种植面积2375．5hm^2，分类面积2556．74hm^2，分类精度达92．91％；上述分类结果为大面积棉花遥感估产打下了基础。     

基于GIS的水稻遥感估产模型研究

以NOAA/AVHRR资料为主,利用GIS技术提取水稻可能种植区域,在此基础上计算各区和各县的比值植被指数和规一化植被指数,提出的水稻遥感估产比值模型和回归模型,预报浙江省的水稻总产,1998年的拟合精度和1999年的预报精度都达到95%以上.     

应用高分辨率遥感影像提取作物种植面积

利用中低分辨率遥感影像提取作物分类种植面积的精度,往往难以满足农业遥感估产的需要。随着新型传感器的不断出现,应用高分辨率遥感影像高精度地提取作物分类面积日益成为发展趋势。由于高分辨率遥感影像提供的地物纹理、色调与形状等信息更加丰富,当前基于对象的地物识别分类方法仍不成熟,处理操作中人为干预过多,而且较为复杂,因此尝试以地面调查信息为辅助参量,采用常规基于像元的最大似然法监督分类方法,依据多尺度遥感影像信息提取的原理,分阶段地逐步提取作物种植面积,以此为农业遥感估产服务。     

遥感数据分类结果的精度分析

遥感数据分类结果的精度包括位兰精度、类型精度及数童精度三种形式。着重讨论了后两种精度的分析过程，并比较了这两种精度分析得出的数值之间的关来。同时还对基于像元分解的分类结果精度分析方法进行了探索，并在上海市水稻种植面积提取结果的精度分析中进行了应用。     

基于HJ-1B卫星遥感数据的水稻识别技术研究

为快速、准确地在遥感图像上识别水稻作物的信息,满足县级尺度水稻遥感监测的需要,以野外实地调查资料、1∶5万地形图数据为辅助,通过光谱分析法,分析研究HJ-1B星CCD数据的水稻作物的光谱反射特性,建立水稻作物遥感信息识别模型。采用决策树分类方法提取水稻作物信息,并将该技术方法应用于广西宾阳县水稻作物信息提取研究。采用实测样地数据,利用混淆矩阵进行精度评价验证,总精度为94.9%,Kappa系数为0.8533。研究表明,该水稻作物的识别技术,可以为了解我国水稻种植情况,进行水稻长势监测和产量估测提供技术参考。     

运用NOAA AVHRR和Landsat TM数据估算多年水稻种植面积

介绍了综合运用NOAA AVHRR和Landsat TM数据进行多年水稻种植面积监测的一种方法,以湖北省为例,首先运用Landsat TM数据计算了该省1992年的水稻种植面积;接着运用1992年和1994年的NOAA AVHRR数据分别计算这两年的水稻像元数,以这两年水稻像元数的变化来反映水稻种植面积的变化;最后运用线性模型,估算1994年的水稻种植面积。所得的1994年水稻种植面积与湖北省农调队资料相比精度为84.5%。运用同样的方法估算1995年该省的水稻种植面积,精度达90%以上。     

Use of RADARSAT‐1 data and a digital elevation model to assess flood damage and improve rice production in the lower part of the Chi River Basin,Thailand

In Thailand, flooding due to seasonal monsoon conditions frequently destroys a substantial amount of rice production, the most important agricultural activity of the country. Taking the 2001 monsoon flooding that hit the Lower Chi River Basin as an example, we developed a new method for accurately assessing damage to flood‐affected paddies. A RADARSAT‐1 image acquired during peak flooding was combined with a 30‐m digital elevation model (DEM) to develop a ‘flood‐level‐determination’ algorithm for estimating floodwater depth. Based on the elongation capability of the rice varieties, a water depth of 80 cm was used to separate ‘non‐damaged’ from ‘damaged’ paddy areas, indicating that about 60% of the paddy fields in the flooded areas were non‐damaged paddies. To minimize the loss of rice and maximize farmers' incomes, a map of rice varieties appropriate for the damaged paddy areas was produced, combining the flood‐affected paddy map with the flood frequency map. Our results demonstrate the potential of using single‐date RADARSAT‐1 data and a DEM to provide accurate and economic means of assessing flood damage to rice fields that can be used to improve rice production.     

Combining satellite multispectral image data and a digital elevation model for mapping debris-covered glaciers

Automated glacier mapping from satellite multispectral image data is hampered by debris cover on glacier surfaces. Supraglacial debris exhibits the same spectral properties as lateral and terminal moraines, fluvioglacial deposits, and bedrock outside the glacier margin, and is thus not detectable by means of multispectral classification alone. Based on the observation of low slope angles for debris-covered glacier tongues, we developed a multisource method for mapping supraglacial debris. The method combines the advantages of automated multispectral classification for clean glacier ice and vegetation with slope information derived from a digital elevation model (DEM). Neighbourhood analysis and change detection is applied for further improvement of the resulting glacier/debris map. A significant percentage of the processing can be done automatically. In order to test the sensitivity of our method against different DEM qualities, it was also applied to a DEM obtained from ASTER stereo data. Additionally, we compared our multisource approach to an artificial neural network (ANN) classification of debris, using only multispectral data. While the combination with an ASTER-derived DEM revealed promising results, the ANN classification without DEM data does not.     

Assessing two large area burnt area products across Australian Southern Forests

Burnt area is a critical parameter for estimating emissions of greenhouse gases associated with biomass burning. Several burnt area products (BAPs) derived from Earth Observation satellites/sensors have been released; these are based on different spatial resolutions and derived using different methodologies so that accuracies can vary amongst them. This study validates a global (MODIS) and a national (AVHRR) BAP across Australian southern forests using two reference datasets: state fire histories (SFHs) from 2000 to 2013 and a forest cover map derived through high resolution air photo interpretation (API). The spatial and temporal agreement between fires in the BAPs and reference SFH were analysed based on 2610 sample points representative of Australian southern forest types (successful detection was evaluated according to fire type: planned burn vs. wildfire, size of fire, and land tenure). Results show that both BAPs were most successful when identifying large wildfires (>5000 ha). Overall accuracy for AVHRR and MODIS was 73.9% and 62.5%, respectively. When compared to the API derived forest cover map as reference dataset, both products achieved higher overall accuracies (94.1% for AVHRR and 87.1% for MODIS); an expected result given that the fires detected in this dataset are known to be observable using Earth observation data. But regardless of reference dataset, the AVHRR BAP which is tailored to Australian conditions achieved better results than the MODIS global BAP. Also, the AVHRR archive in Australia goes back to 1988, which is an important consideration for calculating wildfire history for greenhouse gas accounting.     

Efficiency Research of Planted Rice Area Estimation From NOAA一AVHRR Data in Taihu Domain

运用NOAA-AVHRR资料估算水稻种植面积，是遥感应用领域中一个新的研究方向，结合国家“八五”攻关项目“太湖地区遥感话产”的要求，在太湖地区进行了初步的尝试:(1)根据估产精度要求和NOAA一AVHRR资料校正精度，探讨了运用NOAA一AVHRR资料估产所需的最小区域范围。(2)针对太湖地区的具休地理环境设计了提取水稻种植曲积的技术方案，并在试验区取得了初步成果。     

Change detection with heterogeneous data using ecoregional stratification, statistical summaries and a land allocation algorithm

A ca. 1980 national-scale land-cover classification based on aerial photo interpretation was combined with 2000 AVHRR satellite imagery to derive land cover and land-cover change information for forest, urban, and agriculture categories over a seven-state region in the U.S. To derive useful land-cover change data using a heterogeneous dataset and to validate our results, we a) stratified the classification using predefined ecoregions, b) developed statistical relationships by ecoregion between land-cover proportions derived from the 1980 national-level classification and aggregate statistical data that were available in time series for all regions in the U.S., c) classified multi-temporal AVHRR data using a process that constrained the results to the estimated proportions of land covers in ecoregions within a multi-objective land allocation (MOLA) procedure, d) interpreted land cover from a sample of aerial photographs from 2000, following the protocols used to produce the 1980 classification for use in accuracy assessment of land cover and land-cover change data, and e) compared land cover and land-cover change results for the MOLA method with an unsupervised classification alone. Overall accuracies for the 2000 MOLA and unsupervised land-cover classifications were 85% and 82%, respectively. On average, the 1980-2000 land-cover change RMSEs were one order of magnitude lower using the MOLA method compared with those based on the unsupervised data.     

Regional scale mapping of fractional rice cropping change using a phenology-based temporal mixture analysis

Monitoring changes of paddy rice is challenging due to its diverse cropping patterns and spectral variation. To investigate the spatio-temporal changes of rice cropping, we used the 10-day composited Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) time series data with a spatial resolution of 250 m to map the sub-pixel rice spatial distributions in the Hunan Province, the top one region in rice planting area in southern of China. A method of improved phenology-based temporal mixture analysis (PTMA) was presented to identify early, middle, and late rice cropping patterns. The results show that the PTMA is effective to extract rice cropping. The nine rice cropping patterns were classified as early, middle, and late rice cropping, and fractional rice cropping within 250 m pixels was obtained to analyse the internal changes. Both the local planting conditions and different forms of rice cultivation were compared with statistical data. Overall, MODIS-estimated fractional rice agreed well with field samples at the pixel level and statistical data at the county level, which demonstrates the effectiveness of the PTMA method for mapping rice in these hilly regions with small-size paddy rice field. The changes show that single-cropping rice and double-cropping rice have been frequently transferred in space, which could be important information to support agricultural decision-making.