Perspective of Chinese GF-1 high-resolution satellite data in agricultural remote sensing monitoring

High-resolution satellite data have been playing an important role in agricultural remote sensing monitoring. However, the major data sources of high-resolution images are not owned by China. The cost of large scale use of high resolution imagery data becomes prohibitive. In pace of the launch of the Chinese ＂High Resolution Earth Observation Systems＂, China is able to receive superb high-resolution remotely sensed images （GF series） that equalizes or even surpasses foreign similar satellites in respect of spatial resolution, scanning width and revisit period. This paper provides a perspective of using high resolution remote sensing data from satellite GF-1 for agriculture monitoring. It also assesses the applicability of GF-1 data for agricultural monitoring, and identifies potential applications from regional to national scales. GF-1＇s high resolution （i.e., 2 m/8 m）, high revisit cycle （i.e., 4 days）, and its visible and near-infrared （VNIR） spectral bands enable a continuous, efficient and effective agricultural dynamics monitoring. Thus, it has gradually substituted the foreign data sources for mapping crop planting areas, monitoring crop growth, estimating crop yield, monitoring natural disasters, and supporting precision and facility agriculture in China agricultural remote sensing monitoring system （CHARMS）. However, it is still at the initial stage of GF-1 data application in agricultural remote sensing monitoring. Advanced algorithms for estimating agronomic parameters and soil quality with GF-1 data need to be further investigated, especially for improving the performance of remote sensing monitoring in the fragmented landscapes. In addition, the thematic product series in terms of land cover, crop allocation, crop growth and production are required to be developed in association with other data sources at multiple spatial scales. Despite the advantages, the issues such as low spectrum resolution and image distortion associated with high spatial resolution and wide swath width, might pose challenges for GF-1 data applications and need to be addressed in future agricultural monitoring.

The first quantitative remote sensing of ocean internal waves by Chinese GF-3 SAR satellite

Quantitative analysis and retrieval is given by the State Key Laboratory of Satellite Ocean Environment Dynamics（SOED）,Second Institute of Oceanography（SIO）,State Oceanic Administration（SOA）,China,from the first batch of GF-3 synthetic aperture radar（SAR）data with ocean internal wave features in the Yellow Sea.

Progress in the application of ocean color remote sensing in China

After many years＇ endeavor of research and application practice, the ocean color remote sensing in China has been growing into a new technique with valuable practicality from its initiate stage of trial research. With the aim of operational service, several kinds of ocean color remote sensing application systems have been developed and realized the long-term marine environmental monitoring utilizing the real-time or near real-time satellite and airborne remote sensing data. New progresses in the technology and application of ocean color remote sensing in China are described, including the research of key techniques and the development of various application systems. Meanwhile, according to the application status and demand, the prospective development of Chinese ocean color remote sensing is brought forward. With Chinese second ocean color satellite （ HY-1 B） orbiting on 11 April 2007 and the development of airborne ocean color remote sensing system for Chinese surveillance planes, great strides will take place in Chinese ocean color remote sensing application with the unique function in marine monitoring, resources management and national security, etc.

Advances on Technology and Application of Ocean Color Remote Sensing in China (2004 - 2006)

China has great progress in the technology and application of ocean color remote sensing during 2004-2006. In this report, firstly, four major technical advances are displaying, including (1) the vector radiative transfer numerical model of coupled ocean-atmosphere system; (2) the atmospheric correction algorithm specialized on Chinese high turbid water; (3) systematical research of hyper-spectrum ocean color remote sensing; (4) local algorithms of oceanic parameters, like ocean color components, ocean primary productivity, water transparency, water quality parameters, etc. On the foundation of technical advances, ocean color remote sensing takes effect on ocean environment monitoring, with four major kinds of application systems, namely, (1) the automatic multi-satellites data receiving, processing and application system; (2) the shipboard satellite data receiving and processing system for fishery ground information; (3) Coastal water quality monitoring system, integrating satellite and airborne remote sensing technology and ship measurement; (4) the preliminary system of airborne ocean color remote sensing application system. Finally, the prospective development of Chinese ocean color remote sensing is brought forward. With Chinese second ocean color satellite (HY-1B) orbiting, great strides will take place in Chinese ocean color information accumulation and application.

Extracting Lotus Fields Using the Spectral Characteristics of GF-1 Satellite Data

The lotus(Nelumbo nucifera Gaertn.)is an aquatic plant that grows in shallow water and has long been cultivated in South China.It can improve the incomes of farmers and plays an important role in alleviating poverty in rural China.However,a modern method is required to accurately estimate the area of lotus fields.Lotus has spectral characteristics similar to those of rice,grassland,and shrubs.The features surrounding areas where it is grown are complex,small,and fragmented.Few studies have examined the remote sensing extraction of lotus fields,and automatic extraction and mapping are still challenging methods.Here,we compared the spectral characteristics of lotus fields and other ground objects and devised a remote sensing method for the rapid extraction of lotus fields.Using this method,the extraction accuracy of lotus was 96.3%.The Kappa coefficient was 0.926,which is higher than those of the unsupervised K-means classification,Mahalanobis distance,and support vector machine supervised classification,and demonstrates the potential of this method for extracting and mapping lotus fields by remote sensing.

Prediction of chlorophyll a concentration using HJ-1 satellite imagery for Xiangxi Bay in Three Gorges Reservoir

Since the impoundment of the Three Gorges Reservoir in 2003, algal blooms have frequently been observed in it. The chlorophyll a concentration is an important parameter for evaluating algal blooms. In this study, the chlorophyll a concentration in Xiangxi Bay, in the Three Gorges Reservoir, was predicted using HJ-1 satellite imagery. Several models were established based on a correlation analysis between in situ measurements of the chlorophyll a concentration and the values obtained from satellite images of the study area from January 2010 to December 2011. Chlorophyll a concentrations in Xiangxi Bay were predicted based on the established models. The results show that the maximum correlation is between the reflectance of the band combination of B4/（B2＋B3） and in situ measurements of chlorophyll a concentration. The root mean square errors of the predicted values using the linear and quadratic models are 18.49 mg/m3 and 18.52 mg/m3, respectively, and the average relative errors are 37.79% and 36.79%, respectively. The results provide a reference for water bloom prediction in typical tributaries of the Three Gorges Reservoir and contribute to large-scale remote sensing monitoring and water quality management.

基于GF-1遥感图像和ResNet50网络的PM_(2.5)浓度预测研究

基于GF-1卫星图像、结合Merra-2气象数据作为辅助预测变量,构建了长三角地区基于ResNet50网络的PM_(2.5)预测模型。其中气象参数可以为模型提供较为准确的PM_(2.5)浓度基准,而GF-1图像能帮助模型更合理准确地预测PM_(2.5)浓度的空间变化。利用十折交叉验证和测试集验证对模型进行检验,结果显示:模型的皮尔森相关系数R为0.948,预测PM_(2.5)的RMSE为6.6μg/m~3。反演得到分辨率为500 m的PM_(2.5)浓度分布图合理稳健。GF-1遥感图像和ResNet50网络适用于PM_(2.5)浓度预测,可以作为辅助监测手段,为长三角地区PM_(2.5)热点识别、后续流行病学研究提供数据支撑。

Improvement of Mono-window Algorithm for Retrieving Land Surface Temperature from HJ-1B Satellite Data

The thermal infrared channel (IRS4) of HJ-1B satellite obtains view zenith angles (VZA) up to ±33°. The view angle should be taken into account when retrieving land surface temperature (LST) from IRS4 data. This study aims at improving the mono-window algorithm for retrieving LST from IRS4 data. Based on atmospheric radiative transfer simulations,a model for correcting the VZA effects on atmospheric transmittance is proposed. In addition,a generalized model for calculating the effective mean atmospheric temperature is developed. Validation with the simulated dataset based on standard atmospheric profiles reveals that the improved mono-window algorithm for IRS4 obtains high accuracy for LST retrieval,with the mean absolute error (MAE) and root mean square error (RMSE) being 1.0 K and 1.1 K,respectively. Numerical experiment with the radiosonde profile acquired in Beijing in winter demonstrates that the improved mono-window algorithm exhibits excellent ability for LST retrieval,with MAE and RMSE being 0.6 K and 0.6 K,respectively. Further application in Qinghai Lake and comparison with the Moderate-Resolution Imaging Spectroradiometer (MODIS) LST product suggest that the improved mono-window algorithm is applicable and feasible in actual conditions.

Ocean Observation from Haiyang Satellites: 2012–2014

During 2012 and 2014, China has two Haiyang(which means ocean in Chinese, referred to as HY) satellites operating normally in space which are HY-1B and HY-2A. HY-1B is an ocean color environment satellite which was launched in April 2007 to observe global ocean color and sea surface temperature, and HY-2A is an ocean dynamic environment satellite which was launched in August 2011 to obtain global marine dynamic environment parameters including sea surface height,significant wave height, ocean wind vectors, etc. Ocean observation data provided by HY-1B and HY-2A have been widely used by both domestic and international users in extensive areas such as ocean environment protection, ocean disaster prevention and reduction, marine environment forecast,ocean resource development and management, ocean investigations and scientific researches, etc.

Remote sensing application system for water environments developed for Environment Satellite 1

Remote sensing data collected by the Environment Satellite I are characterized by high temporal resolution,high spectral resolution and mid-high spatial resolution.We designed the Remote Sensing Application System for Water Environments(RSASWE) to create an integrated platform for remote sensing data processing,parameter information extraction and thematic mapping using both remote sensing and GIS technologies.This system provides support for regional water environmental monitoring,and prediction and warning of water pollution.Developed to process and apply data collected by Environment Satellite I,this system has automated procedures including clipping,observation geometry computation,radiometric calibration,6S atmospheric correction and water quality parameter inversion.RSASWE consists of six subsystems:remote sensing image processing,basic parameter inversion,water environment remote sensing thematic outputs,application outputs,automated water environment outputs and a non-point source pollution monitoring subsystem.At present RSASWE plays an important role in operations at the Satellite Environment Center.

基于GF-1卫星数据的农作物种植面积遥感抽样调查方法

GF-1号卫星是中国2013年4月26日发射的一颗高分辨率遥感卫星,为解决该新型卫星数据在农作物对地抽样遥感调查中的应用技术方法问题,该文针对GF-1号卫星数据的特点,研究了基于GF-1号卫星16m WFV传感器和2m/8m PMS传感器卫星数据的农作物种植面积遥感抽样调查方法。根据研究区物候历,选择农作物识别关键期的16m WFV传感器数据进行多时相农作物种植面积的中分辨率遥感提取;在中分辨率农作物面积遥感分类图基础上,计算研究区域的MORAN I指数,确定格网抽样单元的大小,进行多目标农作物的MPPS(multivariate probability proportional to size)抽样;对抽样单元采用2m/8 m PMS传感器卫星数据进行高分辨率农作物面积制图;最后根据MPPS抽样方法进行总体农作物种植面积的推断,并计算CV值,评价抽样精度。以江苏省东台市为研究区对GF-1号卫星数据进行了应用研究。研究结果表明,GF-1号卫星数据完全可以应用于县级农作物种植面积的提取,农作物种植面积提取精度优于90%。

基于6S模型的GF-1卫星影像大气校正及效果

高分一号(GF-1)卫星是中国高分系列卫星的首发星,自2013年4月成功发射以来,在中国农业遥感业务工作中得到了广泛应用,已成为中国大宗农作物种植面积遥感监测的主要数据源。该文基于6S(second simulation of a satellite signal in the solar spectrum)辐射传输模型原理,设计并实现了适合于GF-1卫星数据大气校正算法与程序。算法以GF-1卫星1级数据、元数据及传感器公开参数为输入数据,不需要其他外源辅助数据,经过辐射定标,计算各波段平均太阳辐射值、表观反射率,通过选择大气模式,驱动6S模型获取表观反射率转换为地表反射率的参数,逐像元计算影像地表反射率。在算法研制的基础上,应用Fortran和IDL语言编写了大气校正批处理程序,实现了大气校正过程的批处理。该文采用2014年4月3日、6月28日、11月2日,以及2015年1月19日4个时相北京地区GF1卫星WFV(wide field view)数据,分别代表春夏秋冬4个季节,通过与ENVI软件的FLAASH(fast line-of-sight atmospheric analysis of spectral hypercubes)大气校正结果对比进行评估。2种方法 4个时相各波段全年相对偏差为3.26%,蓝光波段偏差最大为11.21%,其次是红、近红和绿光波段,分别为1.19%、0.73%和0.24%。作物覆盖区平均相对误差为12.99%,冬季最高为17.40%,秋季和春季分别为15.02%和14.15%,夏季相对差异最小为8.31%。各波段地表反射率的整体校正情况并未有太大差异,但6S校正后各波段反射率普遍比FLAASH校正结果略微偏高。2种校正结果计算的NDVI也基本一致,相对偏差0.64%;除水体外,绝对值差值的平均值均在0.0548以内。从计算效率来分析,6S模块实现了商用软件FLAASH模块中未提供的批量计算,在相同硬件环境下计算效率提高了75.0%以上。研究结果表明了该文开发的大气校正程序能够稳定批量处理GF-1卫星数据,可以作为农业遥感监测业务流程的组成部分。

基于GF-1卫星遥感数据识别京津冀冬小麦面积

省级尺度冬小麦面积的精准获取技术是农作物面积遥感监测研究的主要内容之一。为了获取省级尺度的冬小麦种植面积,该文以北京市(京)、天津市(津)和河北省(冀)3个省域范围为例,以国家标准地形图分幅为分类的图幅单元,利用国产GF-1/WFV数据,构建冬小麦面积指数,实现了省级尺度冬小麦面积的识别。本文以冬小麦全部9个月生育期的984景影像作为数据源,依次经过数据预处理、标准图幅单元的NDVI合成、样本点选择、冬小麦面积指数构建、冬小麦作物类型确认、省域范围制图及精度验证等步骤完成研究区域内冬小麦面积的提取。采用区域网平差和6S大气校正算法对数据源预处理,以中国1︰10万标准地形图分幅为分类图幅单元构建冬小麦面积指数,将冬小麦面积指数按照1%的比例等分,并将面积指数从0到100%分割为101个提取节点,将提取节点的NDVI值依次与类型确认样本比较,精度最高的则确认为冬小麦面积提取阈值,同时将该阈值应用于图幅单元内冬小麦面积指数影像,获取冬小麦种植分布。最后冬小麦面积识别的精度表明,以标准地图分幅作为计算单元,在GF-1影像基础上,利用冬小麦面积指数能够显著提高冬小麦与其他地物类型的波谱差异,且冬小麦的总体识别精度达到89.6%,用户精度达到89.8%,制图精度96.5%,Kappa系数0.72。在典型区域,本文算法与监督分类算法精度结果较为一致,除制图精度相差4.77%外,总体精度与用户精度差都在1.00%以内,说明本文算法具有精度高、运行效率高、分类单元识别结果一致性强的特点,能够满足省级尺度农情遥感业务监测的需要。

基于有理多项式模型区域网平差的GF-1影像几何校正

2013年4月成功发射的GF-1卫星是中国高分系列卫星的首发星,影像在中国农情遥感监测业务中得到了广泛应用,已成为大宗农作物种植面积遥感监测的主要数据源之一。高精度几何位置的配准是卫星农情定量化应用的基础与前提,该文提出了一种基于区域网平差方法修正GF-1卫星WFV(wide field view,WFV)影像RPC(rational polynomial coefficients,RPC)参数,获取更高几何定位精度的校正方法,形成了模式化的业务处理流程,为该影像在农情遥感监测中的应用奠定了基础。算法流程包括2个部分,首先是基于像面间仿射变换关系及有理多项式RFM(rational function model,RFM)模型构建轨道间的区域网平差数学模型,其次是根据影像连接点及少量控制点输入求解所有参与平差的卫星影像定向参数,获取亚像元级的校正结果。平差参数的解算是通过两步求解完成的,初始平差参数是根据连接点及对应的DEM高程值进行平差迭代至收敛,结果平差参数是将初始平差参数作为初始值代入区域网平差模型,并以逐点消元方式约化法方程,解算出各影像的仿射变换参数。该文在求解平差参数过程中,直接使用DEM(digital elevation model)上获取的高程值作为约束条件,消除了平面坐标与高程的相关性,保证了区域网平差模型能够解算。混合地形、平原、山区3种情况下区域网平差结果表明,全连接点平差结果具有较高的相对定位精度,其行方向的中误差分别为0.3046、0.4674、0.3365像元,列方向的中误差分别为0.3677、0.2849、0.2889像元;而结合少量控制点的区域网平差则同时具有很高的绝对定位精度,其行方向的中误差分别为0.3648、0.5041、0.3605像元,列方向的中误差分别为0.4954、0.4039、0.6323像元,整体达到了亚像素级。最后,在农业应用基础控制底图的支持下,分别对原始影像、RPC校正影像、区域网平差后的影像进行几何配准,分析不同输入影像条件下的几何校正精度,仅有区域网平差后的影像达到了亚像元的校正精度,混合地形、平原、山区3种情况下行方向的中误差分别为0.6857、0.6664、1.0646像元,列方向的均方差分别为0.4342、0.4696、0.5609像元,但与几何校正前精度相比没有明显改善,说明本文提出的研究方法可以实现少量控制点条件下的几何精校正。不同DEM校正结果表明,对于山区,更高分辨率的DEM可以获得更好的定位精度。上述研究充分表明,该方法对GF-1/WFV数据的处理有效且可行,并在农业部中国农情遥感业务工作中得到了初步应用。

GF-1图像在中印边境楚鲁松杰村地质灾害调查中的应用

为了解高分一号卫星(GF-1)图像的应用性能和查明中印边境楚鲁松杰村一带的地质灾害发育规律,利用DEM和GF-1图像对该地区的岩性、构造和地质灾害进行详细的遥感解译。研究结果表明,GF-1数据的融合图像对灾害体的微地貌形态反映较明显,能够用于识别地质灾害体的位置、边界范围等特征信息;区内地质灾害类型主要为滑坡和寒冻风化碎屑流等,滑坡、崩塌、泥石流多发生在[25°,35°]坡度段内,而寒冻风化碎屑流的发生则与坡度的增大成正相关;地质灾害运移方向与主构造线方向、基岩层面倾向基本一致。研究还揭示出该区内分布多级河流阶地,较大的垂直落差及丰富的第四系松散堆积物是地质灾害发育的重要因素。

甘肃红山地区重要控矿地质单元GF-1数据遥感解译与应用

为了解高分一号(GF-1)卫星图像在地质找矿领域的应用性能,查明甘肃红山地区的成矿地质环境,利用GF-1图像对红山地区的控矿地层、构造和岩体进行遥感解译。在收集、梳理前人研究成果和野外验证的基础上,综合分析红山地区的成矿地质特征和规律,总结区内与多金属矿密切相关的地质体,建立该区重要控矿地层、控矿构造和控矿岩体的GF-1图像解译标志;通过分析重要控矿地质单元的成矿地质特征和成矿规律,解析和挖掘典型矿床的控矿地质要素特征,整合典型矿床所处地质环境和控矿及赋矿要素信息,综合剖析典型多金属矿床的多源异常特征,构建找矿模型;开展找矿应用,圈定遥感找矿有利地段。经野外查证,新发现铜、钼、铁、锌等多金属矿化线索,均位于GF-1图像解译的重要控矿地质单元区,取得了较好的找矿成果,表明国产卫星数据能够很好地应用于地质矿产勘查领域。

基于GF-1 WFV影像和随机森林算法的总氮反演研究

基于高分一号卫星影像遥感数据和水库水质实测数据,利用随机森林回归算法建立遥感反射率与总氮浓度的定量反演模型,以获取东风水库总氮浓度的时空分布情况,进而对水库总氮污染情况进行分析。结果表明,总氮反演模型精度较高,决定系数R2为0.879,均方根误差为0.169 mg/L,但仍有提升空间。将模型运用于2016-2018年GF-1 WFV影像遥感数据,反演得到东风水库总氮浓度时空分布情况。结果表明,2016年水库总氮浓度呈波动性变化,2017-2018年除个别月份外,水库总氮浓度整体呈下降趋势。从水质标准来看,水库总氮浓度介于Ⅲ~Ⅴ类水质标准,且存在富营养化风险。从不同季节变化情况来看,春冬季总氮浓度较高,夏秋季浓度较低。从空间差异性来看,区域变化规律性不太明显。

基于GF-1卫星影像的中国冬小麦制图研究

旨在实现冬小麦面积自动化提取,提出一种基于冬小麦NDVI加权指数(WNDVI)的分类算法。将影像分割成标准的小区域,从而构建冬小麦分类单元,基于WNDVI,采用自适应方法确定冬小麦种植区和非种植区WNDVI分割阈值。以2013年10月至2014年5月期间的GF-1卫星WFV影像为例,构建了全国14个省(市、区)的1180个分类单元的WNDVI,实现了2014年全国主产区冬小麦种植区的空间分布图;采用14233个样本对结果进行验证精度,总体精度达到了90.6%。本方法自动化程度高,结果稳定,适合大范围冬小麦面积监测业务化运行。

Multi-temporal urban semantic understanding based on GF-2 remote sensing imagery:from tri-temporal datasets to multi-task mapping

High resolution satellite images are becoming increasingly available for urban multi-temporal semantic understanding.However,few datasets can be used for land-use/land-cover(LULC)classification,binary change detection(BCD)and semantic change detection(SCD)simultaneously because classification datasets always have one time phase and BCD datasets focus only on the changed location,ignoring the changed classes.Public SCD datasets are rare but much needed.To solve the above problems,a tri-temporal SCD dataset made up of Gaofen-2(GF-2)remote sensing imagery(with 11 LULC classes and 60 change directions)was built in this study,namely,the Wuhan Urban Semantic Understanding(WUSU)dataset.Popular deep learning based methods for LULC classification,BCD and SCD are tested to verify the reliability of WUSU.A Siamese-based multi-task joint framework with a multi-task joint loss(MJ loss)named ChangeMJ is proposed to restore the object boundaries and obtains the best results in LULC classification,BCD and SCD,compared to the state-of-the-art(SOTA)methods.Finally,a large spatial-scale mapping for Wuhan central urban area is carried out to verify that the WUsU dataset and the ChangeMJ framework have good application values.