基于多时序特征和卷积神经网络的农作物分类

近年来,以卷积神经网络为主的深度学习模型在各种遥感应用中都显示出巨大的潜力。以加州帝国郡为研究区,以Landsat 8 OLI年内时序遥感影像计算时序植被指数NDVI、EVI、RVI以及TVI,组合后输入到构建的一维卷积神经网络 模型,以实现作物的高精度精细分类。为了验证卷积模型的优越性,另搭建了基于递归神经网络及其变体的深度学习模型。结果表明：①引入其他时序特征后,能够有效地提高卷积神经网络的分类精度。NDVI+EVI+TVI+RVI组合特征总体精度和Kappa系数最高,分别是89.667 4%和0.856 0,对比NDVI时序特征总体精度和Kappa系数提高了近4%和0.6。②在与其他深度学习模型的对比中,一维卷积神经网络分类精度最高,能够从时序数据中较为准确捕捉作物时序特征信息,尽管递归神经网络被广泛应用于序列数据的研究,但分类结果要略差于卷积神经网络。实验表明在NDVI的基础上引入其他植被指数辅助,能够有效地提高分类精度。基于一维卷积神经网络的深度学习框架为长时间序列分类任务提供了一种有效且高效的方法。     

Waterbody mapping: a comparison of remotely sensed and GIS open data sources

Surface water maps are essential for many environmental applications. Waterbody delineation from satellite images remains a challenging task due to sensor limitations, the presence of clouds, the low albedo surfaces in urban areas, topographic, and atmospheric conditions. In this paper, a model based on the Supported Vector Machine (SVM) classifier was adopted for waterbody extraction from Sentinel-2, Landsat 8 Operational Land Imager (OLI) and RapidEye satellite images. As well, the accuracy of two other sources (OpenStreetMapping (OSM) and Military Geographic Institute (MGI)) was tested. The free images from Sentinel-2 and Landsat 8 OLI were more accurate (Kappa (KHAT):0.89, 0.88) data sources than commercial RapidEye images (KHAT: 0.79). Regarding the performance between Sentinel-2 and Landsat 8 OLI, Sentinel-2 obtained the most accurate results (overall accuracy 94.49 vs. 94.17, commission error 1.34 vs. 1.87). Due to the variable spatial resolution of OSM and MGI data, it was not possible to detect small waterbodies with these sources, and therefore high values of omission error and a strong underestimation of the area of surface water were obtained. This study demonstrates the suitability of free images for mapping and monitoring of surface waterbodies, including small water bodies.     

融合不精准先验知识的Landsat 8 OLI影像深度学习分类方法

遥感影像解译是获得土地利用和土地覆盖（LULC）信息最为重要的途径之一,而自动化分类是提高LULC信息获取效率的关键。实际场景中包含大量不精准的先验知识,提取并融合其中的可用知识能进一步提高影像分类方法的精度、自动化率和规模应用能力。基于上述情况,提出了一种融合不精准先验知识的Landsat 8 OLI影像深度学习分类方法。该方法可自动规避先验知识中的不精准单元,在图斑约束空间内实现了分类样本的自动化区域选择和特征提取,并获得了高置信度知识,然后利用这些分类样本训练深度残差网络,从而实现大区域影像的精确分类。以常州市新北区为例进行实验,选用该区域2009年土地利用现状数据作为先验数据,2014年Landsat 8 OLI影像作为待分类影像。实验结果表明,所提方法可融合不精准先验知识,对大面积连片LULC信息分类精确,主要地类图斑界限准确,全图分类图斑精度达到了88.7%,Kappa系数为0.842。该方法能配合深度学习方法实现高精度Landsat 8 OLI遥感影像分类。     

Study on Crops Remote Sensing Classification based onMulti-temporal Landsat 8 OLI Images

It is crucial for agricultural production to know crop planting situation.Temporal remote sensing images and subtle spectral characteristics of ground features play an important role in extracting crops distribution.At this point,multi-temporal Landsat 8 OLI images were used to extracting the distribution of main crops in the east of Xinrong district of Datong city by using Spectral Angle Mapper(SAM) combined with the decision tree classification,and the extracting result was compared with the result that maximum likelihood extracted.The results show that：① The planting area of spring corn,grain,soybean and potato is decreased and mosaic distribution in order.② The overall accuracy obtained by SAM combined with the decision tree classification is 85.34% and the Kappa coefficient is 0.76,which is outperformed the results of maximum likelihood with the increase of 22.51% and 0.31,respectively,the classification results was more consistent with the actual distribution of main crops.③ The classification accuracy of main crops used the multi-temporal remote sensing images was obviously higher than that of single-temporal image,and the difference between ground features and spectra in middle or high resolution images can effectively weaken by analyzing multi-temporal data from the perspective of difference of spectral angle.The results not only confirmed the positive effect of multi-temporal remote sensing images on crops classification,but also developed the SAM combined with decision tree classification in crops classification of medium-high resolution remote sensing images,which has a certain application prospect.     

Assessment of trophic state and water quality of coastal-inland lakes based on Fuzzy Inference System

Accurate assessment and monitoring of coastal and inland water quality by satellite optical remote sensing is challenging due to improper atmospheric correction algorithm, inaccurate quantification of in-water constituents' concentration and a lack of efficient models to predict the water quality status. The present study aims to address the latter two parts in conjugation with an appropriate atmospheric correction algorithm to assess trophic status and water quality conditions of two coastal lagoons using Landsat-8 OLI data. Three vital underwater light attenuating factors, directly related to water quality, are considered namely, turbidity, chlorophyll and colored dissolved organic matter (a_CDOM). These water quality parameters are quantified based on certain sensitive normalised water-leaving radiance band ratios and threshold values. To assess the accuracy of the derived products, these algorithms were applied to independent in-situ data and statistical evaluation of the results showed good agreement between the estimated and measured values with the errors within desirable limits. Being a primary nutrient indicator, the chlorophyll concentration was used to evaluate Trophic State Index. The Water Quality Index was derived from three parameters namely, chlorophyll concentration, turbidity and a_CDOM(443) which were expressed in terms of Trophic State Index, Turbidity Index and Humic-Fulvic Index, respectively. The Water Quality Index maps, derived using a Fuzzy Inference System based on the Centre of Gravity method, provided insights into spatial structures and temporal variability of water quality conditions of the coastal lagoons which are influenced by anthropogenic factors, hydrographic changes and land-ocean-atmospheric interaction processes.     

A thermodynamic analysis of the SO2/H2SO4 system in SO2-depolarized electrolysis

The hybrid sulfur thermochemical cycle has been proposed as a means to produce efficiently massive quantities of clean hydrogen using a high-temperature heat source like nuclear or solar. The cycle consists of two steps, one of which is electrolytic. The reversible cell potential for this step and, hence, the resulting operating potential will depend on the concentrations of dissolved SO₂ and sulfuric acid at the electrode. To understand better how these are related as functions of temperature and pressure, an Aspen Plus phase equilibrium model using the OLI Mixed Solvent Electrolyte physical properties method was employed to determine the activities of the species present in the system. These activities were used in conjunction with the Nernst equation to determine the reversible cell potential as a function of sulfuric acid concentration, temperature and pressure. A significant difference between the reversible and actual cell potentials was found, suggesting that there may be considerable room for reducing the operating potential.     

Ground-based Automatic Radiometric Calibration of Land Observation Satellite Optical Sensors and Cross Validation Analysis

Due to the low cost-effectiveness and large uncertainty of single calibration for traditional ground-based radiometric calibration methods，it is difficult to meet the requirement ofhigh-precision radiometric calibration of satellite payloads.Aroutinely-operated ground-based automatic radiometric calibration method was developed and applied on radiometric calibration and cross validation analysis of Landsat-8/OLI opticalsensor based on the “National Calibration and Validation Site for High Resolution Remote Sensors” (hereinafter referred to as the “Baotou Site”) in Baotou.The comparison of 11 observation results (from May 2016 to April 2017) between on-board calibration and ground-based calibrationare in good agreement：for the four bands of blue，green，red and near infrared，the average relative deviation between ground-based calibration and on-board calibration was 0.83%，-0.21%，-0.20%，and -1.37%，respectively，while the standard deviation was 2.78%，2.89%，2.94%，and 2.20%，respectively.Further，quantitative analyses on the sources of errors in the process of ground-based automatic radiometric calibration was conducted.The results showed that the final uncertainties of ground-based automatic radiometric calibration in the four bands of blue，green，red，and near infrared were 5.06%，4.65%，4.80%，and 4.98%，respectively.Good consistency between ground-based calibration and on-board calibration proved the reliability of this method，which can dramatically promote the frequency and timeliness of satellite radiometric calibration.     

锂辉石水热钾碱分解制取碳酸锂相平衡模拟与优化试验

在综合分析锂辉石精矿物相组成基础上,采用OLI Analyzer软件对LiAlSi_2O_6-K_2CO_3-H_2O体系锂辉石水热分解反应多相平衡进行模拟。结果表明,β-锂辉石在热力学上极易与K_2CO_3反应生成麦钾沸石和碳酸锂,当K_2CO_3加入量为K~+/Li~+≥1.4时,即可完全转化。水热试验获得的优化条件为:K_2CO_3加入量为K~+/Li~+=1.5、反应温度240℃、反应时间120min、水固质量比4,此条件下锂辉石中锂的浸出率为96.43%,最终反应产物主要为麦钾沸石与碳酸锂,与模拟结果相一致。     

基于Landsat 8 OLI的特征变量优化提取及森林生物量反演

基于江苏省常熟市虞山地区Landsat 8OLI影像和55块调查样地数据,利用多元逐步回归法建立森林生物量模型,并讨论了预测结果及其精确性。选择包括各波段灰度值、不同波段灰度值之间的线性和非线性组合(包括18种植被指数)、纹理信息以及主成分分析、最小噪声分离变换等在内的53个特征变量。通过分析53个特征变量与森林地上、地下生物量的Pearson相关性,进行特征变量的优化提取。结果表明:所有样地无区分分析时,地上和地下生物量的模型精度均达到0.4以上,基于3种森林类型(针叶林、阔叶林和混交林)进行地上和地下生物量建模时精度有明显提高,达到0.67以上,地上生物量和地下生物量的估测结果均为混交林优于阔叶林,阔叶林优于针叶林。     

基于多时相Landsat8 OLI影像的作物种植结构提取

针对基于多时相遥感影像、多种特征量提取多种作物种植结构在我国研究较少的现状,利用多时相Landsat8OLI影像数据,根据温宿县不同作物的农事历,通过分析主要地物的光谱特征和归一化植被指数的时间变化信息,构建不同作物种植结构提取的决策树模型,实现了对温宿县多种作物种植结构信息的提取。结果表明:1水稻的最佳识别依据是5月20日影像的近红外波段和7月23日影像的NDVI值;棉花和春玉米的最佳识别依据是5月20日~9月9日影像的NDVI变化值;冬小麦—夏玉米和林果的最佳识别依据是5月20日~7月23日影像的NDVI变化值;2与单时相监督分类相比,多时相决策树法对多种作物种植结构的提取效果更理想,总体精度提高了7.90%,Kappa系数提高了0.10;3Landsat8OLI影像数据分辨率高、成本低、获取方便,是农作物遥感的良好数据源。