Statistical data fusion of multi-sensor AOD over the Continental United States

This article illustrates two techniques for merging daily aerosol optical depth (AOD) measurements from satellite and ground-based data sources to achieve optimal data quality and spatial coverage. The first technique is a traditional Universal Kriging (UK) approach employed to predict AOD from multi-sensor aerosol products that are aggregated on a reference grid with AERONET as ground truth. The second technique is spatial statistical data fusion (SSDF); a method designed for massive satellite data interpolation. Traditional kriging has computational complexity O(N³), making it impractical for large datasets. Our version of UK accommodates massive data inputs by performing kriging locally, while SSDF accommodates massive data inputs by modelling their covariance structure with a low-rank linear model. In this study, we use aerosol data products from two satellite instruments: the moderate resolution imaging spectrometer and the geostationary operational environmental satellite, covering the Continental United States.     

暗目标法的Himawari-8静止卫星数据气溶胶反演

Himawari-8(H8)是由日本气象厅发射的新一代静止气象卫星,可实现10 min/次的高频次对地观测,搭载的AHI(Advanced Himawari Imager)传感器设置有与MODIS暗目标气溶胶反演算法所需的类似波段。本文参考暗目标算法构建了针对该卫星传感器的陆地气溶胶反演算法:首先,通过基于地基站点观测数据的精确大气校正,统计得到短波红外与可见光波段的地表反射率比值关系,将此作为先验知识用于地—气解耦时的反射率估计;然后,初步假设大陆型气溶胶类型,利用辐射传输模型建立查找表;最后,通过模拟与卫星观测的表观反射率误差最小实现气溶胶光学厚度反演解算。选取2016年5月覆盖京津冀地区的观测数据进行测试,将反演结果与对应时间的MODIS气溶胶光学厚度产品进行对比验证,空间分布趋势一致、相关性较高,相关系数R达到0.852;通过与地基观测网AERONET站点实测数据对比验证,所有站点的相关系数R~2均大于0.88,精度较高。利用反演的高时间分辨率产品,分析了京津冀地区的大气空间分布和日变化情况,结果表明:采用暗目标法对H8静止卫星陆地气溶胶光学厚度反演具有一定的潜力和可行性,能反映气溶胶的高时间变化信息,有望成为大气环境污染变化监测新的重要手段。     

查找表方法确定气溶胶类型

针对传统气溶胶类型确定方法的局限性以及当前气溶胶类型确定存在的困难,提出一种使用多波段气溶胶光学厚度数据确定气溶胶类型的方法。基于大气颗粒物的散射与吸收特性分析,通过构建查找表的方法实现气溶胶类型的确定。该方法利用Mie散射理论通过正向模拟不同类型气溶胶粒子数量与多波段光学厚度之间的关系来构建查找表,基于该查找表,使用440 nm、670 nm、870 nm及1020 nm 4个波段的气溶胶光学厚度确定气溶胶类型。使用模拟的多波段气溶胶光学厚度数据开展了气溶胶类型的确定实验,分析了不同波段气溶胶光学厚度误差对气溶胶类型确定结果的影响。结果表明,该方法可根据4个波段的气溶胶光学厚度以较高的精度确定出沙尘性、水溶性和煤烟3种气溶胶粒子的数量,从而确定气溶胶类型。     

基于HJ1A数据的沈阳市雾霾时空变化模拟研究

基于HJ1A-CCD数据,采用6S大气辐射传输模型,反演出研究区气溶胶光学厚度。采用偏最小二乘法,基于气溶胶空间分布的估算结果,模拟出研究区PM_(2.5)的空间分布。基于GIS和改进的高斯大气扩散模型,进行了研究区PM_(2.5)的扩散模拟。根据我国PM_(2.5)检测网的空气质量新标准,对沈阳市的雾霾进行等级了划分。     

相对湿度对吸收性气溶胶指数的影响

为提高风云三号气象卫星(FY-3)紫外臭氧总量探测仪(TOU)观测数据得到的吸收性气溶胶指数AAI(Absorbing Aerosol Index)或AI(Aerosol Index)的可靠性,需要了解AAI指数与相对湿度之间的内在关系。本研究利用大气辐射传输模型DAK(Doubling-Adding KNMI)分别模拟了在城市气溶胶和乡村气溶胶模式下AAI同相对湿度之间的关系,并将结果同已观测到的实际结果进行对比。结果发现,在相对湿度呈高值时AAI指数出现很大的变化,但相对湿度对两种气溶胶模型的影响具有相反的效应,分析显示当大气中含有吸收性气溶胶如含碳类气溶胶(在中国北部的污染过程中很常见)时,AAI结果对RH(Relative Humidity)有很强的依赖。在应用AAI指数产品检测污染过程中需要注意气溶胶的具体类型和相对湿度的影响,必要时在高相对湿度过程中进行数据校正或剔除。     

北京地区Landsat 8 OLI高空间分辨率气溶胶光学厚度反演

卫星气溶胶光学厚度(AOD)反演中,传统暗目标方法在反射率较低的水体、浓密植被覆盖区域取得了较好效果,在反射率较高且结构复杂的高反射地表上空目前多采用深蓝算法,但存在空间分辨率较低,对细节分布描述性较差等问题。为解决这一问题,本文首先以5年(2008年—2012年)长时间序列MODIS地表反射率产品为基础,采用最小值合成法建立500 m分辨率逐月地表反射率产品数据集,然后利用地物波谱库中典型地物波谱数据,分析建立MODIS与Landsat 8 OLI传感器蓝光波段反射率转换模型,最后北京地区AERONET地基观测数据确定了气溶胶光学物理参数,并反演获取了北京地区上空500 m分辨率的AOD分布。为验证反演算法的精度,分别将反演结果同AERONET及MODIS/Terra气溶胶产品(MOD04)进行交叉对比,同时利用相关系数R,均方根误差RMSE,平均绝对误差MAE以及MODIS AOD产品预期误差EE共4个指标进行衡量。结果表明:算法反演获取的AOD与AERONET观测值具有较高的一致性,各指标分别为R=0.963,RMSE=0.156,MAE=0.097,EE=85.3%,稍优于MOD04产品(R=0.962,RMSE=0.158,MAE=0.101,EE=75.8%),并且有效的对比点数也高于MOD04。通过与地基观测相比,卫星遥感获取的高分辨率城市地区AOD精度可作为定量评估城市空气质量的有效依据。     

NPPVIIRS数据反演气溶胶光学厚度

利用NPP卫星的VIIRS传感器数据,基于暗像元法反演陆地气溶胶光学厚度AOD。首先,根据红外波段的归一化植被指数NDVI来对暗像元进行识别;然后,利用6S软件进行辐射传输计算构建查找表;最后,根据VIIRS数据从查找表插值得到AOD,并对其进行海拔校正。选取华北地区作为反演实验区,获得了2013年9月1日的气溶胶分布。利用AERONET北京站太阳光度计地基观测结果对反演结果对比验证,发现二者具有显著的相关性,相关系数达到0.7920。将2013年9月1日的MODIS AOD产品与本研究反演的AOD进行比对,发现二者分布趋势一致,相关系数为0.7059,相关性显著。反演结果表明,本文算法反演陆地AOD效果较好,为大气颗粒物环境监测提供了良好方法手段和数据源。     

Machine Learning and Bias Correction of MODIS Aerosol Optical Depth

Machine-learning approaches (neural networks and support vector machines) are used to explore the reasons for a persistent bias between aerosol optical depth (AOD) retrieved from the MODerate resolution Imaging Spectroradiometer (MODIS) and the accurate ground-based Aerosol Robotic Network. While this bias falls within the expected uncertainty of the MODIS algorithms, there is room for algorithm improvement. The results of the machine-learning approaches suggest a link between the MODIS AOD biases and surface type. MODIS-derived AOD may be showing dependence on the surface type either because of the link between surface type and surface reflectance or because of the covariance between aerosol properties and surface type.      

Satellite-based terrestrial production efficiency modeling

Production efficiency models (PEMs) are based on the theory of light use efficiency (LUE) which states that a relatively constant relationship exists between photosynthetic carbon uptake and radiation receipt at the canopy level. Challenges remain however in the application of the PEM methodology to global net primary productivity (NPP) monitoring. The objectives of this review are as follows: 1) to describe the general functioning of six PEMs (CASA; GLO-PEM; TURC; C-Fix; MOD17; and BEAMS) identified in the literature; 2) to review each model to determine potential improvements to the general PEM methodology; 3) to review the related literature on satellite-based gross primary productivity (GPP) and NPP modeling for additional possibilities for improvement; and 4) based on this review, propose items for coordinated research.     

灰霾污染状况下气溶胶组分及辐射效应的遥感估算

本文利用地基遥感观测数据和气溶胶模型,分析了北京地区2013年1月份灰霾期间气溶胶中不同成分的光学与物理化学特性,并使用大气辐射传输模式估算了气溶胶中不同成分对直接辐射强迫的贡献。结果显示,在严重灰霾时期,大气中的硫酸盐等水溶性气溶胶光学厚度（440nm）显著增加,对气溶胶光学厚度的贡献达到70%以上,同时黑碳气溶胶对总光学厚度（440nm）的贡献也呈现增长趋势,是晴朗天气下的8倍以上。严重灰霾（28日）时气溶胶的直接辐射强迫显著增加,但气溶胶中不同成分对辐射效应的贡献不同,黑碳成分在气溶胶对大气层的辐射强迫中占主要贡献（57%）,而水溶性成分则在气溶胶对地表的辐射强迫中占主要贡献（60%）。     

中国东部气溶胶光学厚度季节变化的数值模拟

气候模式对气溶胶光学厚度AOD的合理模拟,是模拟研究气溶胶气候效应的前提。利用在线耦合的区域气候—大气化学—气溶胶耦合模式系统RIEMS-Chem,模拟研究了2010年中国东部地区AOD的季节变化情况。模拟结果与卫星搭载的中分辨率成像光谱仪(MODIS)的反演资料和地基气溶胶观测网(AERONET)的站点观测资料分别进行了一年四季的详细对比,检验结果显示尽管模拟值有所低估,模式仍然能够合理地反映AOD的季节变化情况和空间分布特征,与AERONET站点观测值相比,整体相关系数为0.6。MODIS反演和相应模拟结果均显示,中国东部地区AOD整体水平夏季最大,春季次之,秋、冬季最小,华北平原、四川盆地和华中地区是AOD的主要大值区。只考虑日间AOD时,其季节分布特征略有不同,在华北平原地区,日间AOD夏季最大(1.1—1.5),在长江中下游流域地区,日间AOD则在春季最大(1.1—1.7);在中国东部,日间AOD的大值在夏、冬两季分别主要分布在长江以北、以南地区,而在春、秋两季则主要位于长江中下游流域。     

南京地区大气气溶胶综合观测与对比分析

利用CE-318太阳光度计、MPL激光雷达与卫星观测数据,分别采用光谱消光法、Fernald方法以及MODIS暗像元法(DDV)反演南京地区气溶胶光学厚度,并进行了对比分析。通过研究分析3月3日、6日卫星反演气溶胶光学厚度的空间分布图,发现长江流域附近以及市区(除老山、中山陵等山区地带之外)的AOD较高。3月3日太阳光度计、激光雷达与卫星数据在站点位置(南京信息工程大学,118.7°E,32.2°N)的AOD值分别为0.455、0.289、0.4;3月6日的AOD值分别为0.373、0.267、0.25。通过对比分析3月至9月之间的多天数据,可得3种数据计算所得AOD相差不大,说明卫星与激光雷达反演数据相对可靠。其中,3月3日与3月6日的太阳光度计数据显示,观测地区出现常见的两种AOD变化类型:一种是早晚高,中午低;一种是早低晚高。此外,激光雷达所得数据结果随着时间的变化幅度较大,且可以在有云的天气条件下探测气溶胶;本文利用激光雷达数据计算出的9 km以下AOD值多数在0.3左右,3月3日与3月6日两天之中,2 km以下较脏,出现了一些气溶胶层,6km以上相对比较干净,个别时段6 km以上高空存在云层。与地基观测相比,卫星虽然时间分辨率虽然低,但是对于大面积的趋势分析却有着绝对的优势。在今后的气溶胶观测发展中,结合三者的优势,有助于以较高精度,大面积反演大气气溶胶空间分布情况,获得较准确的气溶胶参数。     

珠三角地区长时间序列气溶胶时空变化特征分析

本文通过对2000年—2013年长时间序列的MODIS气溶胶产品进行统计,分析了珠三角地区气溶胶光学厚度(AOD)和细粒子光学厚度(FAOD)的空间分布特征以及年度和季节变化特点,有助于深入研究珠三角地区颗粒物污染水平变化及颗粒物的排放与输送等。研究结果显示珠三角地区中部为AOD高值区,东西两翼地区为AOD低值区。AOD和FAOD的最高值通常分别出现在春季和秋季,最低值则通常都出现在冬季。2006年之后,珠三角地区大气气溶胶总消光虽在部分年份仍有反弹上升的现象出现,但已有明显降低。然而,该地区细粒子消光在2000年—2012年期间则呈逐年增加的趋势,且其空间差异性也越加显著,细颗粒物污染仍需进一步控制。     

Dust Radiative Effects Over Global Oceans

Using one year of moderate resolution imaging spectroradiometer (MODIS) and clouds and the Earth's radiant energy system (CERES) data, we provide a satellite-based assessment of top-of-atmosphere (TOA) cloud-free shortwave and longwave dust radiative effects over global oceans from the Terra satellite. Over global cloud-free oceans, the dust net radiative effect is -0.7 plusmn0.2 W middotm^-2, and the TOA dust shortwave radiative effect (SWRE) dominates the longwave radiative effect (LWRE). Globally, the annual mean dust contribution to the total MODIS level 2 aerosol optical thickness (AOT, at 550 nm) is about 30% with a dust SWRE of -0.7 plusmn0.2 W middotm^-2 and LWRE of 0.03 plusmn0.02 W middotm^-2. Averaged over all seasons, the cloud-free diurnal mean dust radiative efficiency is -33 plusmn5 W middotm^-2 middottau^-1, and there is a remarkable linear relationship between the CERES SWRE and the MODIS AOT. This is the first satellite-based assessment of dust net radiative effect over the global oceans and will serve as a useful constraint for numerical modeling analysis.     

FY-3A/MERSI海上沙尘天气气溶胶光学厚度反演

利用FY-3A/MERSI资料,结合MODIS C005算法的海上气溶胶模型,研究了中国渤海、黄海以及东海海域沙尘天气气溶胶光学厚度反演方法。通过MERSI反演的气溶胶光学厚度与MODIS C005气溶胶光学厚度产品(MOD04)对比,发现MERSI气溶胶光学厚度反演结果存在较大偏差。分析认为,这种偏差是由MODIS C005算法的海上气溶胶模型对MERSI气溶胶光学厚度反演不完全适用造成的。鉴于此,本文研究引进了一种沙尘气溶胶模型,并将其与MODIS C005算法的粗粒子气溶胶模型按照一定比例混合,形成了改进的气溶胶模型。利用改进气溶胶模型再次反演海上沙尘天气气溶胶光学厚度,反演结果与MOD04一致性较好,说明改进气溶胶模型能有效地提高MERSI定量反演沙尘气溶胶的能力。     

利用深蓝算法从HJ-1数据反演陆地气溶胶

大气气溶胶是环境空气污染监测的重要指标,在利用环境一号卫星CCD相机进行气溶胶监测时,暗目标法和结构函数法都有相应的不足。本文从Hsu等人(2004)提出的深蓝算法出发,以MODIS的地表反射率产品为基础建立反射率库,并利用地面观测数据分析了各种典型地物在CCD相机与MODIS蓝波段反射率之间关系,提出了将MODIS地表反射率修正到CCD相机的方法,进而实现地气解耦,反演气溶胶光学厚度。选择北京地区为实验区,进行了算法实验,并用AERONET/PHOTONS北京站的数据进行了验证,结果表明,(1)光学厚度较大时(>0.5),深蓝算法精度能够较好的满足环境一号卫星CCD相机对气溶胶日常监测的要求;(2)气溶胶模式会对结果产生较大的影响,尤其是城市型气溶胶。     

Spatiotemporal mixed effects modeling for the estimation of PM2.5 from MODIS AOD over the Indian subcontinent

ABSTRACT

The physical processes associated with the constituents of the troposphere, such as aerosols have an immediate impact on human health. This study employs a novel method to calibrate Aerosol Optical Depth (AOD) obtained from the MODerate resolution Imaging Spectrometer (MODIS – Terra satellite) for estimating surface PM_2.5 concentration. The Combined Deep Blue Deep Target daily product from the MODIS AOD data acquired across the Indian Subcontinent was used as input, and the daily averaged PM_2.5pollution level data obtained from 33 monitoring stations spread across the country was used for calibration. Mixed Effect Models (MEM) is a linear model to deal with non-independent data from multiple levels or hierarchy using fixed and random effects of dependent parameters. MEM was applied to the dataset obtained for the period from January to August 2017. The MEM considers a fixed and random component, where the random components model the daily variations of the AOD – PM_2.5 relationships, site-specific adjustment parameters, temporal (meteorological) variables such as temperature, and spatial variables such as the percentage of agricultural area, forest cover, barren land and road density with the resolution of 10 km × 10 km. Estimation accuracy was improved from an R² value of 0.66 from our earlier study (when PM_2.5 was modeled against only AOD and site-specific parameters) toR² value of 0.75 upon the inclusion of spatiotemporal (meteorological) variables with increased % within Expected Error from 18% to 35%, reduced Mean Bias Error from 3.22 to 0.11 and reduced RMSE from 29.11 to 20.09. We also found that spline interpolation performed better than IDW and Kriging inefficiently estimating the PM_2.5 concentrations wherever there were missing AOD data. The estimated minimum PM_2.5 is 93 ± 25μg/m³ which itself is in the upper limit of the hazardous level while the maximum is estimated as 170 ± 70μg/m³. The study has thus made it possible to determine the daily spatial variations of PM_2.5 concentrations across the Indian subcontinent utilizing satellite-based AOD data.     

Seasonality and Latitudinal Variation of Water Vapour and Aerosol Optical Thickness and Their relationship Over Tropical Indian Ocean

Aerosol and water vapour are very important element in the Earth’s climate system which has direct role in the Earth’s radiation budget. In this paper the seasonality, latitudinal distribution and the relationship of aerosol optical thickness (AOD) and water vapour (WV) using MODIS Level 3 monthly data from 2001 to 2008 are analysed. The analysis shows that AOD (0.55 μm) values reach maximum during southwest monsoon and remain minimum during northeast monsoon period. The Equatorial Indian Ocean shows minimum AOD (0.115 to 0.153) throughout the year compared to Arabian Sea (0.208 to 0.613) and Bay of Bengal (0.214 to 0.351). Arabian Sea shows high variation and maximum value of AOD compared to Bay of Bengal and Equatorial Indian Ocean. During southwest monsoon WV over Bay of Bengal was found higher in concentration compared to Arabian Sea and Equitorial Indian Ocean throughout the study period. Comparison between Arabian Sea (2.98 cm to 5.07 cm) and Bay of Bengal (3.49 cm to 5.94 cm) shows that WV concentration is less in Arabian Sea throughout the year. The analysis of correlation between WV and AOD was found to be inconsistent. However, AOD and WV shows a strong positive correlation for whole year (Mean R² =0.90) in the Equitorial Indian Ocean region except in the months of January, February and March. In general, the correlation between WV and AOD is found to be strongly positive for oceanic aerosol (sea salt) in low water vapour condition.     

GF-4增强型地表反射率库支持法的气溶胶光学厚度反演

针对GF-4等国产卫星气溶胶光学厚度反演算法存在的地表反射率估计困难、云像元污染等问题,本文发展了一种增强型地表反射率库支持的气溶胶光学厚度反演方法,改进了云筛选与地表反射率确定方案,在考虑GF-4逐像元成像角度的情况下,使用6SV模型与MOD09-CMA数据对季度尺度上的GF-4 PMS传感器数据进行大气校正,提出了百分比最小值均值法建立地表反射率库,并据此建立了NDVI与红蓝反射率关系模型,根据地表反射率的分布特点,当NDVI小于0.2的时候使用地表反射率库估计地表反射率,而当NDVI大于0.2时,则使用NDVI来估计地表反射率。使用MOD04气溶胶模式时空分布确定气溶胶参数。在京津冀地区开展气溶胶光学厚度反演实验,使用Aeronet站点数据与MOD04产品对反演结果进行了对比验证,与Aeronet相关系数R为0.964,均方根误差RMSE为0.13,满足±(0.05+0.2τ)的点多于78.9%,相关系数与均方根误差优于MODIS暗目标法产品,满足期望误差线的数量优于MODIS暗目标与深蓝算法产品。