Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Validation of Two MODIS Aerosols Algorithms with SKYNET and Prospects for Future Climate Satellites Such as the GCOM-C/SGLI

Validation of Two MODIS Aerosols Algorithms with SKYNET and Prospects for Future Climate... Hindawi Publishing Corporation Advances in Meteorology Volume 2013, Article ID 508064, 16 pages http://dx.doi.org/10.1155/2013/508064 Research Article Validation of Two MODIS Aerosols Algorithms with SKYNET and Prospects for Future Climate Satellites Such as the GCOM-C/SGLI 1 2 3 2 Jules R. Dim, Tamio Takamura, Akiko Higurashi, Pradeep Kathri, 3 4 Nobuyuki Kikuchi, and Takahashi Y. Nakajima Earth Observation Research Center/JAXA, 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan Center for Environmental Remote Sensing (CEReS), Chiba University, 1-33 Yayoi-Cho, Inage-Ku, Chiba 263-8522, Japan Center for Environmental Measurement and Analysis, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan Department of Network and Computer Engineering, Tokai University, 2-28-4 Tomigaya, Shibuya-ku, Tokyo 151-0063, Japan Correspondence should be addressed to Jules R. Dim; rosutando@yahoo.com Received 14 April 2013; Revised 10 June 2013; Accepted 11 June 2013 Academic Editor: Harry D. Kambezidis Copyright © 2013 Jules R. Dim et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Potential improvements of aerosols algorithms for future climate-oriented satellites such as the coming Global Change Observation Mission Climate/Second generation Global Imager (GCOM-C/SGLI) are discussed based on a validation study of three years’ (2008–2010) daily aerosols properties, that is, the aerosol optical thickness (AOT) and the Angstrom ¨ exponent (AE) retrieved from two MODIS algorithms. The ground-truth data used for this validation study are aerosols measurements from 3 SKYNET ground sites. The results obtained show a good agreement between the ground-truth data AOT and that of one of the satellites’ algorithms, then a systematic overestimation (around 0.2) by the other satellites’ algorithm. The examination of the AE shows a clear underestimation (by around 0.2–0.3) by both satellites’ algorithms. eTh uncertainties explaining these ground-satellites’ algorithms discrepancies are examined: the cloud contamination aeff cts differently the aerosols properties (AOT and AE) of both satellites’ algorithms due to the retrieval scale differences between these algorithms. eTh deviation of the real part of the refractive index values assumed by the satellites’ algorithms from that of the ground tends to decrease the accuracy of the AOT of both satellites’ algorithms. eTh asymmetry factor (AF) of the ground tends to increase the AE ground-satellites discrepancies as well. 1. Introduction and local variation. The most common aerosols optical and physical properties used for this characterization are the The determination of the optical properties of aerosols and AOT, theAE, theparticlesizedistribution, thesinglescatter- their size distribution around the globe has been a significant ing albedo (SSA), the aerosol phase function, the asymmetry contemporary research effort of late [ 1]. Some of the major factor (AF), the refractive index (RI), and so forth. To obtain factorsthathaveenabled this progress arethe better spectral these aerosols characteristics from satellites’ observations, a and spatial capacities of satellites and ground-based radiome- detailed model of the aerosols properties is required [2]. The ters, the improvement of the aerosol signal filtering methods, increasing number and spatial expansion of field measure- the better knowledge of the aerosols particles’ shapes, and so ment campaigns and ground sites’ coverage have helped to forth. The qualitative and quantitative importance of the data gather statistically meaningful data for the development of collected through the continuous monitoring of aerosols and aerosols models. Ground collected data not only serve as their daily global coverage, by various satellites, has permitted preliminary input for aerosols models but also are reliable a better characterization of the role of the aerosols in the evaluation and validation tools for aerosols products derived climate dynamics and the understanding of their temporal from satellites (GMS-5/SVISSR, NOAA-AVHRR, Terra- and 2 Advances in Meteorology Aqua-MODIS and MISR, OrbView-2/SeaWiFS, ENVISAT- settings andmotives of thestudy;(2) the description of the MERIS, etc.). eTh objective of these validation studies is ground data collection approach and the satellites’ algorithms to improve the quality of the aerosols properties retrieved main features;(3) the presentation of the main characteristics from satellites’ observations as well as the confidence level in of theaerosolspropertiesofthe studyarea; (4)the validation theseproducts. Sincethe launch of Terra- andAqua-MODIS, scheme of the satellites’ retrievals, and, discussion on the ground data from AERONEThavebeenusedtovalidate uncertainties plaguing these retrievals; (5) the lessons for and improve the MODIS AOT over ocean [3] and land future satellites’ aerosols products; (6) the summary of the [4, 5]. NOAA-AVHRR AOT data have been also evaluated results obtained. using sun photometers of AERONET and others [6, 7]. The retrieval quality over the ocean of POLDER/ADEOS aerosols measurements has been also conducted through comparisons 2. Ground Data Collection and with AERONET sun photometer data [8]. Satellites’ Algorithms The increasing accuracy of aerosols products, as a result of all these evaluation/validation and correction studies, The analyses conducted in this study aim at validating couldbeusedtoreducethe uncertaintiesassociatedwiththe tropospheric aerosols’ retrievals from two algorithms, using aerosol radiative forcing of the global climate [9]. Due to the similar observations from both Terra- and Aqua-MODIS complex and diverse composition, size, structure, and origin satellites.Thesesatelliteshavespectrallycompatiblechannels of aerosols and though a lot of progress has been made in the with the future GCOM-C/SGLI satellite. This validation study satellites’ observations, there are still various challenges for a is conducted through a comparison of the satellites’ retrievals globally acceptable accuracy of the aerosol optical and micro- with ground-truth measurements from three SKYNET sites: ∘ ∘ physical properties of different types of aerosols. The high Fukue (32.752 N latitude, 128.682 Elongitude,and 50m ∘ ∘ spectral and spatial resolutions of climate-oriented satellites altitude), Hedo (26.867 N latitude, 128.249 Elongitude,and ∘ ∘ such as the GCOM-C/SGLI, scheduled to be launched around 65 m altitude), and Miyako (24.737 Nlatitude, 125.327 E 2015 by the Japan Aerospace Exploration Agency (JAXA), longitude, and 50 m altitude). es Th e sites are islands located aim at accomplishing better distinctions between aerosols in the Pacicfi East Asian region, south of Japan. eTh choice particles and other atmospheric components. However, the of these three locations for the satellites’ retrieved aerosols programmed long lifespan of such satellites (3 series) and properties’ evaluation study derives from the fact that they the connectivity between similar/different satellites’ series are at the confluence of one of the most complex mixtures will pose other accuracy issues proper to long-term climate of aerosols types. It is known that the Asian atmospheric observations [10–12] that may only be alleviated with better environment has been worsened rapidly by Asian dust, accuracy retrieval algorithms. anthropogenic gases, and aerosols in recent years [13]; it In the present study, the potential performance of future has been also the subject of various aerosols projects. Some satellites’ products is discussed through a validation study of the examples are the Asian-Pacific Regional Aerosol of aerosol retrievals by present satellites’ (Terra- and Aqua- Characterization Experiment (ACE-Asia) [14–16]which was MODIS satellites) algorithms using spectrally compatible a multiplatform project where aerosols, radiative u fl xes channels with those of the coming GCOM-C/SGLI. This val- measurements were conducted over the East China Sea; idation is conducted against a three-year (2008–2010) daily and the Atmospheric Brown Cloud East Asian Regional ground-truth dataset from three SKYNET observation sites Experiment 2005 [17, 18], with the study of aerosols radiative (Fukue,Hedo, andMiyako).Thesegroundsites arelocated characteristics and aerosol direct radiative forcing. in the Pacific East Asian region, an area at the confluence of eTh SKYNET,whose data areusedtovalidateagainst actively changing and mixed aerosols (natural and anthro- the satellite retrievals in this study, is a well-developed and pogenic particles resulting from rapid industrialization). eTh maintained network of climate radiation stations with ground retrievals examined are from two algorithms using similar sites spread all over Japan and the East Asian region [17, 19]. observations, Terra- and Aqua-MODIS satellites’ calibrated It serves for the monitoring of aerosol and cloud proper- radiances (MOD/MYD021, version 5.1). eTh rfi st algorithm ties as well as other weather/climate variables. The basic is named here as MODIS-GLI, as it was previously applied aerosols properties at the SKYNET sites are obtained from to the Global Imager (GLI) onboard the Advanced Earth radiation measurements performed with sky radiometers. Observation Satellite-II (ADEOS-II). eTh second algorithm These radiometers record direct sky radiance intensities at is the MODIS-NASA algorithm, used for MODIS stan- 11 predenfi edwavelengths (315,340,380,400,500,675,870, dard products’ retrievals by the National Aeronautics and 940, 1020, 1600, and 2200 nm) and scattering angles. eTh Service Administration (NASA). The aerosols properties measurement frequency is 10 or 15 minutes. The solar direct- investigated are the aerosol optical thickness (AOT) and beam transmission and sky radiance obtained from sky the Angstrom ¨ exponent (AE). The differences between the radiometer measurements are used for aerosols properties satellites’ retrievals and ground observations are analyzed, such as theAOT,the aerosolcolumnsizedistributions,the issues plaguing the accuracy of each of these satellites’ complex refractive indices, the single scattering albedo, and algorithms are discussed, and lessons for future satellites’ the asymmetry factor, at multiple wavelengths. A program algorithms improvements, such as the algorithm of the future package named Skyrad.pack [20]isusedfor theretrieval of GCOM-SGLI, are drawn. To conduct this work, the following these parameters. The SKYNET helps to extend the global steps are adopted:(1) the introduction, highlighting the basic aerosol monitoring from the ground to locations not always Advances in Meteorology 3 covered by the better known AErosol RObotic NETwork observed and theoretical reflectance, are iteratively searched. (AERONET). Previous studies have shown a good match The 𝛿 parameter, representing the contribution of the large between the SKYNET and the AERONET aerosols mea- particles in the aerosol total load, is used to calculate the AE. surements in areas where both networks coexist, confirming For the MODIS-NASA algorithm, the retrieval scheme the quality and consistency of the SKYNET measurements. over the ocean, described by Kaufmann and Tanre[ ´ 30], One of these is a study conducted over the Loess Plateau uses seven wavelengths from 0.47 to 2.13𝜇 mtoretrievethe of Northwest China. eTh study compares the aerosol opti- AOT and the volume distribution (in the range of 0.08– cal depths at two common wavelengths (675 and 870 nm) 5𝜇 m radius) from measured radiances. LUTs from radiative observed by the CIMEL sun photometer of the AERONET transfer calculations of the MODIS radiance spectral bands and those observed by the POM-01 sky radiometer of the are used to retrieve aerosols parameters at each particle size SKYNET,based on 703clear-sky measurements during the distribution mode: the small particle mode or accumulation year 2009 [21]. The study shows a high correlation coefficient mode particles (5 cases) and the large or coarse mode (0.99) and a low relative difference of about 3% between particles (6 cases). The AOT and the volume distribution are themeasurementsofthe twoinstruments.Bietal. [22]also the primarily derived parameters from the algorithm, while showed a good match between the AOT retrieved by both the AE is (as in the MODIS-GLI algorithm) a secondarily networks at four major channels (440, 675, 870, and 1020). derived parameter. The AOT is obtained from the measured Both the SKYNET and AERONET aerosols networks radiances, by linear interpolation between vfi e optical thick- have been very oen ft used to validate and refine aerosols ness values of each aerosol particle mode combination. eTh models for the satellites’ retrievals [3, 4, 9, 23–26]. The two modes that give the best selected value (ratio between the multiplication of ground stations and the improvement of modes) are the selected aerosol models for the retrieval. thequality of thedataobtainedwiththe introduction of new parameterization schemes in the radiative transfer derivation 3. Aerosols Characteristics codes and a more rigorous cloud mask application have increased the confidence level in the use of these ground- Prior to the evaluation of the satellite retrievals against based data for satellite products’ validation purpose [25, 27]. ground data, a quick description of some of the major In this study, daily aerosol retrievals from both Terra- characteristics of the aerosols distribution observed in the and Aqua-MODIS at 0.5𝜇 m wavelength are analyzed. eTh region covering the ground sites selected for this study may matching time and space area between these satellite data be useful to understand the aerosols of the region. Figure 1 andthose from thegroundstationsare±30 min and 25 km shows the prevailing pattern of the AOT and AE distributions around each ground site (time and space average are applied), in the region, as retrieved by the MODIS-GLI algorithm respectively. eTh parameters studied are the AOT at 0.5 𝜇 m during winter (month of January 2010), and the geographical andthe AE from visiblechannels. eTh MODIS-NASA data location of the SKYNET ground sites selected. In this gur fi e, (MOD/MYD04, version 5.1) produce the AOT at 0.55𝜇 m. the aerosol distribution shows high AOTs near the coastal Based on the wavelength relationship between the AE and area of the southwest (adjacent to China) and in the northwest the AOT, the latter are interpolated to 0.5𝜇 m, in order to area representing the sea of Japan. But higher AE (finer match the retrievals from the MODIS-GLI algorithm and the particles) are visible in the former than the latter. Closer ground data AOT. ranges of both AOTand AE canbeseeninanorth-south The basic retrieval scheme of the MODIS-GLI algorithm band running along the SKYNET sites, with however thicker over the ocean, described in Nakajima and Higurashi [23], layers and n fi er particles in the north than the south. For Higurashi and Nakajima [28], and Higurashi et al. [29], the three-year data set matching the satellites’ overpass time, uses the visible (0.660𝜇 m) and near-infra-red (0.870𝜇 m) the northern site (Fukue) has the thickest aerosols (mean channels to retrieve theAOT.Thesatellite-receivedradiances AOT of 0.30±0.23) as well as the finest particles (mean are synthesized in four look-up tables (LUTs) derived from AE of0.97±0.35). The other 2 sites (Hedo and Miyako) radiative transfer calculations under the Mie theory scheme. located in the south have mean AOT values of 0.26± 0.20 and The algorithm assumes a complex refractive index (RI =1.5– 0.27±0.15,respectivelyand AE of0.95±0.35 and0.89±0.36, 0.005𝑖 ) and a bimodal size distribution expressed as respectively. As we move from the north to the south, that is, 2 further from the major coastal area and towards the central (ln𝑟− ln𝑟 ) 𝑑𝑉 (𝑟 ) parts of the Pacific Ocean ( Figure 1), the aerosol particle size =∑𝐶 exp{− }, (1) 𝑑 ln𝑟 2𝜎 𝑛=1 𝑛 increases (AE values get smaller). where𝑛 is thenumberoflognormal aerosolmodes,𝑟 is the median or geometric mean radius at each mode,𝜎 is the 4. Evaluation of Satellites’ Derived variance or width of the mode,𝐶 represents the particle peak Aerosols Data volume concentration of the mode; this is an undetermined parameter to be retrieved from satellite-received radiances. 4.1. Analyses and Results. The previous section examined the The AOT and the peak ratio 𝛿 of thesizedistributionare aerosols characteristics of the study eld fi , based on ground retrieved by comparing measured and theoretical (precalcu- observations. Now, let us use these characteristics to evaluate lated LUTs) reflectance values. eTh optimal values of AOT the AOT and AE retrievals from the satellites’ algorithms. and𝛿 , that minimize the root mean square deviation between Figures 2 and 3 present 1 : 1 scatter plots between daily 4 Advances in Meteorology January 2010 Terra-MODIS AOT from GLI algorithm Table 1: Error analyses showing the root mean square error 40N (RMSE) between the ground data and the satellite derived data for the MODIS-GLI algorithm (GLI RMSE) and the MODIS-NASA 38N algorithm (NASA RMSE), and the simple differences between the ground data and the MODIS-GLI (GLI Dif), then the MODIS- 1.8 36N NASA (NASA Dif) satellites’ algorithms. 1.6 34N AE AOT AE AOT 1.4 Sites RMSE GLI RMSE GLI RMSE NASA RMSE NASA 32N Fukue 0.411 0.214 0.321 0.138 1.2 Fukue Hedo 0.535 0.234 0.388 0.147 30N 1 Miyako 0.566 0.233 0.384 0.089 28N Sites Dif GLI Dif GLI Dif NASA Dif NASA 0.8 Fukue 0.181 −0.194 0.065 0.009 26N Hedo 0.6 Hedo 0.248 −0.203 0.150 −0.010 Miyako 0.145 −0.202 0.086 −0.008 24N 0.4 Miyako 22N 0.2 to that of MODIS-GLI. In both MODIS-GLI and MODIS- 20N NASA cases, the dispersion around the 1 : 1 correlation line 120E 122E 124E 126E 128E 130E 132E 134E 136E 138E 140E tends to increase at large AOT values (above 1). (a) For the AE, the graphs of Figure 3 show that the satellites’ retrievals from both algorithms are clearly underestimated. January 2010 Terra-MODIS AE from GLI algorithm 40N This underestimation is slightly larger with the MODIS-GLI data than with the MODIS-NASA, and its magnitude seems 38N 2 to vary with the proportion of small/large particles’ size of the aerosol load present in the ground data. eTh distribution of 1.8 36N ground AE data appears as two overlapping datasets (around 1.6 a cutoff value of 0.8) and, the increase in the number of 34N ground data with AE > 0.8 against those with AE < 0.8 1.4 leads to an increasing underestimation of the AE by the 32N satellite retrievals. In the large particle group, the satellite 1.2 Fukue 30N particles are smaller in size than the ground data, while in the 1 smallparticlegroup,the satellitedataare larger in size.And 28N as, in general, small particles tend to be in higher number 0.8 than large particles, among the ground data, the AE would 26N be underestimated. The determination coefficient in the AE Hedo 0.6 comparison is lower than that in the AOT comparison for 24N 0.4 both MODIS-GLI and MODIS-NASA cases. Among the 3 Miyako ground sites, the best match between the ground and both 22N 0.2 satellites’ algorithms AOT and AE is at Fukue (northernmost site). 20N 120E 122E 124E 126E 128E 130E 132E 134E 136E 138E 140E Table 1 shows the discrepancies between the ground measurements and the satellites’ retrievals, expressed as the (b) simple difference and the RMSE (root mean square error) of Figure 1: Geographical location of the three SKYNET radiation sites the AOT and AE. At the 3 sites, there are higher AOT RMSEs (Fukue, Hedo, Miyako) used in this study to evaluate the satellite and biases (against ground observations) with the MODIS- retrievals, over an image of satellite derived aerosols properties from GLI algorithm (ranges 0.214–0.234 and−0.194–−0.203, resp.) the MODIS-GLI algorithm, for the month of January. compared to the MODIS-NASA algorithm (ranges 0.089– 0.147 and−0.008–0.009, resp.). eTh AOT bias of the MODIS- GLI algorithm is consistently negative, that is, the average aerosols measurements from the ground (SKYNET sites) AOT from this algorithm is overestimated (by an average of timely matching the satellite retrievals (from MODIS-GLI around 0.2), while the bias of the MODIS-NASA shows only and MODIS-NASA algorithms) for the AOT and AE, respec- a slightly under/overestimated value of around±0.010. These tively, during the 3-year study period. Analyses of these plots AOT validation biases confirm the conclusions of Remer et show that the MODIS-GLI AOT values are systematically al. [9] indicating that the MODIS-NASA ocean algorithm overestimated compared to ground data, while the MODIS- has virtually no offset and little bias except for a slight NASA AOT values show a better match with these ground underprediction at high AOT, and the linear regression line data, as shown by the higher correlation coecffi ient compared follows the 1 : 1 line closely. Pacific Ocean Pacific Ocean Japan Japan Advances in Meteorology 5 AOT Fukue versus MODIS-GLI AOT Fukue versus MODIS-NASA 1.5 1.5 1 1 0.5 0.5 y = 0.7214x + 0.0825 y = 0.7676x + 0.2657 R = 0.45 R = 0.63 0 0 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Fukue Fukue (a) (b) AOT Hedo versus MODIS-GLI AOT Hedo versus MODIS-NASA 1.5 1.5 0.5 0.5 y = 0.837x + 0.2517 y = 0.6662x + 0.0951 R = 0.44 R = 0.57 0 0 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Hedo Hedo (c) (d) AOT Miyako versus MODIS-NASA AOT Miyako versus MODIS-GLI 1.5 1.5 0.5 0.5 y = 0.8722x + 0.0465 y = 0.8361x + 0.285 R = 0.89 R = 0.44 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Miyako Miyako (e) (f) Figure 2: Scatter diagrams between ground (Fukue, Hedo, Miyako) and satellites’ derived aerosols AOT. Ground data are on the 𝑥 -axis while satellite derived data (from both algorithms, MODIS-GLI and MODIS-NASA) are on the𝑦 -axis. AOT is the aerosol optical thickness. The uncertainty in the AE satellite estimation is expected to the total optical thickness magniefi s relative errors in the to be larger than that of the AOT. An expression of this retrievals of individual path radiances by up to 4 to 8 times uncertainty is [31] the original percentages [9]. The AE uncertainty values are much closer than the AOT errors; between the two satellites’ 1 Δ𝜏 Δ𝜏 algorithms, the RMSE and bias vary from 0.414–0.566 and 1 2 Δ𝛼= ( + ). (2) 0.145–0.248, respectively, for the MODIS-GLI, then 0.321– ln(𝜆 /𝜆 ) 𝜏 𝜏 2 1 1 2 0.388 and 0.065–0.150, respectively, for the MODIS-NASA ˚ (Table 1). Both algorithms show consistently a positive bias 𝛼 is the Angstrom ¨ exponent, and𝜏 and𝜏 are the aerosol 1 2 in the AE, that is, smaller AE (larger size of particles) than optical thickness at wavelengths𝜆 and𝜆 ;Δ𝛼 ,Δ𝜏 ,Δ𝜏 are 1 2 1 2 ˚ ground data. the errors in the Angstrom exponent and the aerosol optical Figure 4 presents the comparison between the frequency thickness at the wavelengths𝜆 and𝜆 ,respectively. 1 2 distribution of the ground measurements and the satellite As a rule, satellite retrievals of aerosol optical thickness derived data, for the AE and AOT. eTh MODIS-NASA are more robust than the corresponding retrievals of aerosol algorithm shows the closest AOT distribution range to that size; consequently, the fraction of the fine mode contribution MODIS-GLI MODIS-GLI MODIS-GLI MODIS-NASA MODIS-NASA MODIS-NASA 6 Advances in Meteorology AE Fukue versus MODIS-GLI AE Fukue versus MODIS-NASA 1.5 1.5 0.5 0.5 y = 0.3895x + 0.4473 y = 0.6295x + 0.3217 R = 0.37 R = 0.27 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Fukue Fukue (a) (b) AE Hedo versus MODIS-GLI AE Hedo versus MODIS-NASA 1.5 1.5 1 1 0.5 0.5 y = 0.1941x + 0.5047 y = 0.4503x + 0.353 R = 0.14 R = 0.25 0 0 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Hedo Hedo (c) (d) AE Miyako versus MODIS-GLI AE Miyako versus MODIS-NASA 2 2 1.5 1.5 1 1 0.5 0.5 y = 0.1362x + 0.5875 y = 0.3481x + 0.4774 R = 0.13 R = 0.2 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Miyako Miyako (e) (f) Figure 3: Scatter diagrams between ground (Fukue, Hedo, Miyako) and satellites’ derived aerosols AE. Ground data are on the 𝑥 -axis while satellite derived data (from both algorithms, MODIS-GLI and MODIS-NASA) are on the𝑦 -axis. The AE is the Angstrom ¨ exponent. of the ground data with however a consistently low sensitivity data distributions appear to be bimodal. This distribution to low AOT values (<0.1). eTh principal mode occurs at the contrasts with that of the two satellites’ algorithms, as they samerangesasthatofthegrounddata,thatis,around0.2–0.3. show at all sites a unimodal distribution with the modes con- The difference in the data distribution between this algorithm sistently at 0.8 (for the MODIS-GLI) and 1.0 (for the MODIS- andthe ground AOTdistributionatmostofthe ranges is NASA), and fewer data are available beyond these modes. below 5%. Contrary to the MODIS-NASA algorithm, the This contributes to the underestimation of the AE in the high MODIS-GLI algorithm shows its highest AOT peak range AE ranges by the 2 satellites’ algorithms, overwhelming the at >0.7, contrary also to the ground data peak, which is overestimation in the low AE ranges. around 0.3. This MODIS-GLI higher range peak depletes the eTh comparison of the AOT, then AE seasonal distri- number of data in the lower AOT ranges and contributes bution between the ground and satellite data, is shown at to the AOT overestimation by this algorithm. Concerning Figure 5, as the monthly average of each parameter for the 3- the AE, it shows a unimodal distribution (mode at 1.2) at year study period. The MODIS-GLI algorithm overestimates theFukue ground site whilethe Hedo andMiyakoground the AOT nearly equally each month. A slight inflection of MODIS-GLI MODIS-GLI MODIS-GLI MODIS-NASA MODIS-NASA MODIS-NASA Advances in Meteorology 7 AE (Fukue) AOT (Fukue) 35 35 5 5 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 >1.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 Ranges Ranges AOT (Hedo) AE (Hedo) 0.2 0.4 0.6 0.8 1 1.2 1.4 >1.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 Ranges Ranges AOT (Miyako) AE (Miyako) 35 35 25 25 20 20 15 15 10 10 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.2 0.4 0.6 0.8 1 1.2 1.4 >1.4 Ranges Ranges AOT ground AE ground AOT GLI AE GLI AE NASA AOT NASA (a) (b) Figure 4: Comparison between the frequency distribution of ground m easurements and satellite derived (GLI for the MODIS-GLI algorithm, NASA for the MODIS-NASA algorithm) aerosols properties (AE (a), AOT (b)). this overestimation is seen in summer, around July. There is and Miyako), show closer values between the ground and a better match each month between the MODIS-NASA and both the MODIS-GLI and MODIS-NASA data. eTh AE is the ground data. mostly underestimated by the satellites’ algorithms. But this For the AE, winter months (January, February, March) underestimation is generally within the ground data monthly and summer months (July, August, September), respectively, variation range (standard deviation represented by vertical for the northern site (Fukue) and the southern sites (Hedo blue bars on the map). Number of data (%) Number of data (%) Number of data (%) Number of data (%) Number of data (%) Number of data (%) 8 Advances in Meteorology AE (Fukue) AOT (Fukue) 1 1.5 0.8 0.6 0.4 0.5 0.2 0 0 (a) (b) AE (Hedo) AOT (Hedo) 1.5 0.8 0.6 0.4 0.5 0.2 0 0 (c) (d) AOT (Miyako) AE (Miyako) 1 1.5 0.8 0.6 0.4 0.5 0.2 Ground Ground GLI GLI NASA NASA (e) (f) Figure 5: Average monthly variation of the ground (Ground) AOT at the 3 SKYNET sites (Fukue, Hedo, Miyako) and the corresponding satelliteretrievals: MODIS-GLI(GLI) andMODIS-NASA (NASA).Theerrorsbars(blue)represent thestandarddeviation of themonthly ground data. AOT AOT AOT Jan. Jan. Jan. Feb. Feb. Feb. Mar. Mar. Mar. Apr. Apr. Apr. May. May. May. Jun. Jun. Jun. Jul. Jul. Jul. Aug. Aug. Aug. Sep. Sep. Sep. Oct. Oct. Oct. Nov. Nov. Nov. Dec. Dec. Dec. AE AE AE Jan. Jan. Jan. Feb. Feb. Feb. Mar. Mar. Mar. Apr. Apr. Apr. May. May. May. Jun. Jun. Jun. Jul. Jul. Jul. Aug. Aug. Aug. Sep. Sep. Sep. Oct. Oct. Oct. Nov. Nov. Nov. Dec. Dec. Dec. Advances in Meteorology 9 4.2. Impact of External Factors and Aerosol Model Assump- with MODIS-NASA) tends to increase, substantially (Dif tions on Satellites’ Retrievals. As mentioned before, a prior GLI) or slightly (Dif NASA) with the CF increase. eTh AOT step in the retrieval of aerosols properties is the choice of an bias observed here may surpass 0.3 (e.g., at the CF peak of aerosol model. This model is based on various assumptions. 0.7).Theslighttrend betweenthe DifNASAand theCF, eTh uncertainties in these assumptions will likely affect the foundinthisstudy conrfi msthe remarksofRemer et al. accuracy of the aerosols properties retrieved from satellite [9] that the MODIS monthly mean AOT over the ocean observations. Some of these uncertainties are the particle is not significantly cloud contaminated (based on the good size, refractive index, the single scattering albedo, the particle agreement between the MODIS-NASA ocean algorithm and sphericity [30], the satellite sensors’ calibration, the cloud 4 AERONET sites’ AOT). signal (thin clouds or adjacent cloud to the aerosols) inter- For the 3-year dataset, the average cloud amount is higher ference, the asymmetry factor, and so forth. In this section in the south (58.5% and 60.2% at Hedo and Miyako, resp.) we will review the contribution of these uncertainties on the than the north (44.2% at Fukue). And, coincidently, the cloud accuracy of the satellite aerosols retrievals. Initial evaluations contamination and the negative bias (Table 1)are loweratthe of Terra- and Aqua-MODIS ocean aerosols products, made latter area compared to the former areas. by Kaufman and Tanre[ ´ 30], suggest that the expected The AE and AOT larger biases in the MODIS-GLI than uncertainty for MODIS small AOTs ranges from 0.05 to 0.1 the MODIS-NASA, in relation to the cloud contamination, and increases to 20–30% for high AOTs. In Remer et al. [9], can be explained by the pixel size selection scale differences the uncertainty is given as±0.03±0.05 AOT. between the two algorithms. The MODIS-GLI algorithm retrieves aerosols properties on a pixel-by-pixel basis while the MODIS-NASA algorithm retrievals are made on 10 4.2.1. Cloud Contamination Eeff ct. Atmosphere areas where aerosols properties are retrieved are supposed to be cloud- × 10-pixel boxes (only pixels with the highest noncloudy free. Examination of Terra- and Aqua-MODIS cloud masks probability are selected and averaged). This selection process suggest that, even aer ft rigorous cloud screening, the resulting unfortunately reduces the number of data to be analyzed clear-sky dataset could still be contaminated (sometimes by but produces less cloud-contaminated products compared to the pixel-by-pixel retrieval approach adopted by the MODIS- 15% or more) by small clouds [32, 33]. Also, in cloud-free areas, adjacent cloud pixels to clear-sky areas can increase GLI algorithm. Kaufman and Tanre[ ´ 30]suggest that when (brightening effect) or decrease (shadowing effect) the inten- theaerosol productisnot givenonapixelbypixel basis 2 2 sity of the aerosol radiative signal. The brightening eeff ct is (0.5 × 0.5 km ) but rather over a grid of 10 × 10 km or expected to play a more prominent role in these aerosols 50× 50 km , clear sky areas used for the determination of areas. To discuss the possible eeff cts of cloud contamination aerosols can lead to a cloud contamination of aerosols and on the satellite retrievals, we will examine the variation of therefore contribute to measurement errors of these aerosols an independent variable, the average cloud fraction (CF) properties. Decreasing the spatial resolution for the aerosol in each spatial subarea (the CF is from the MOD/MYD06 product could help to reduce cloud contamination but also version 5.1 MODIS cloud product) where the MODIS-NASA enhance the signal-to-noise-ratio (SNR) for the AE retrieval algorithm retrieval was performed, against the change in AE [30, 34]. andAOT (bothatthe ground andsatellite levels)thenthe bias of the satellite retrievals, that is, the difference (Ground- 4.2.2. Particle Size and Shape Distribution Assumption. In MODIS-GLI or Dif GLI and Ground-MODIS-NASA or Dif order to convert the aerosol mass measurements to size distri- NASA). Figure 6 shows the results of these relationships. On butions, a model of the aerosol size distribution as a function this figure, the CF data are sorted in 0.1 ranges (i.e., 10% of aerosol mass is needed [35]. If “actual” aerosol properties CF). The maximum range is >0.7 (i.e., >70% CF). Let us are too different from those assumed by the model, then examine rfi st the possibility of cloud contamination on the the satellite retrieved products will be less accurate [36]. For AE.TheAEgroundvariation showsthatitisnearlysteady the 3-year study period and at the 3 ground sites examined the distribution is consistently bimodal: small particle mode (no specific trend) along the CF variation ranges, while both the MODIS-NASA and MODIS-GLI AE show a decreasing (accumulation) and large particle mode (coarse) radii. It was trend (larger sizes) with the CF increase. Consequently, the noticed that the small mode radius tends to move to higher AE difference ground-satellite (Dif GLI and Dif NASA) tends values with the increase of the particle volume. This creates to increase with the CF increase. A stronger AE decreasing a larger dispersion (as will be shown later with the standard trend at high CF is observable with the MODIS-GLI than deviation) of the mode. The average radii of the ne fi mode ± the MODIS-NASA at all sites. eTh se trends show that both standard deviation (STD) are0.15±1.75𝜇 m,0.16±1.63𝜇 m, satellites’ algorithm retrieved AE appear to be contaminated and0.16±1.72𝜇 mandthecoarsemodes4.18± 2.35𝜇 m,4.92± by clouds. 2.27𝜇 m, and4.72±2.21𝜇 mfor Fukue, Hedo,and Miyako, The AOT as the AE shows no specific trend in the respectively.Theaccumulationmoderadiusisnearlysimilar ground data variation with the CF. In contrast, the MODIS- at the3sites(0.15–0.16),and thecoarsemodeislargerinthe GLI AOT gradually increases with the CF increase while south than the north. The STDs within each mode range are the increase of the latter with the MODIS-NASA AOT also very comparable. eTh accumulation and coarse modes is slighter. Consequently, the difference in absolute value radiiusedinthe MODIS-GLIalgorithm are, respectively, between the ground and satellite AOT (Dif GLI for the 0.17±1.3,3.44±2.37. es Th e values are, respectively, higher difference with MODIS-GLI or Dif NASA for the difference and lower than those of the ground sites, while their STDs 10 Advances in Meteorology Fukue (CF versus AE) Fukue (CF versus AOT) 0.5 1.4 0.8 0.5 1.2 0.7 0.25 0.25 0.6 0.5 0.8 0 0.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.6 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.3 0.4 −0.25 −0.25 0.2 0.2 0.1 −0.5 −0.5 0 CF ranges CF ranges Dif GLI AE GLI AOT GLI Dif GLI Dif NASA AE NASA Dif NASA AOT NASA AE ground AOT ground Hedo (CF versus AE) Hedo (CF versus AOT) 0.5 1.4 0.5 0.8 1.2 0.7 0.25 0.6 0.25 0.5 0.8 0 0.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.6 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.3 0.4 −0.25 −0.25 0.2 0.2 0.1 −0.5 0 −0.5 0 CF ranges CF ranges AE GLI AOT GLI Dif GLI Dif GLI AE NASA AOT NASA Dif NASA Dif NASA Ground AE ground Miyako (CF versus AOT) Miyako (CF versus AE) 0.5 0.5 0.8 1.4 0.7 1.2 0.25 0.6 0.25 0.5 0.8 0 0.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.6 0.3 0.4 −0.25 −0.25 0.2 0.2 0.1 −0.5 −0.5 0 CF ranges CF ranges AE GLI AOT GLI Dif GLI Dif GLI AE NASA AOT NASA Dif NASA Dif NASA AE ground AOT ground (a) (b) Figure 6: Cloud fraction (CF) variation against aerosols properties retrievals from ground (AE and AOT ground), MODIS-GLI (AE and AOT GLI), and MODIS-NASA (AE and AOT NASA). en Th CF variation against satellite bias, that is, difference ground-satellite (ground AE or AOT-MODIS-GLI AE or AOT), (ground AE or AOT-MODIS-NASA AE or AOT), during the 3 years of observations. The AE or AOT scales are on the right while the scale of the Dif is on the left. Dif AE (ground − satellite) Dif AE (ground − satellite) Dif AE (ground − satellite) AE AE AE Dif AOT (ground − satellite) Dif AOT (ground − satellite) Dif AOT (ground − satellite) AOT AOT AOT Advances in Meteorology 11 Fukue (V /V ) S L Hedo (V /V ) S L 1.2 1.2 0.8 0.8 0.4 0.4 −0.4 −0.4 Accumulation mode volume ratio Accumulation mode volume ratio AE Dif GLI Dif GLI AE ground AE ground AE Dif NASA Dif NASA AE GLI AE GLI AE NASA AE NASA (a) (b) Miyako (V /V ) S L 1.2 0.8 0.4 −0.4 Accumulation mode volume ratio AE ground Dif GLI AE GLI Dif NASA AE NASA (c) Figure 7: AE and AE difference (ground-satellite) against the accumulation mode volume ratio ( 𝑉 /𝑉 )variation. 𝑆 𝐿 are lower and close to those of the ground observations. study. However no relationship could be established between The MODIS-NASA algorithm uses a combination of models the increase of ground-satellite discrepancies with either the whosemedianradius ± STD varies from 0.035 ± 0.40 to accumulation or the coarse mode radii. 0.10±0.60 for the accumulation mode and0.40 ± 0.60 to Figure 7 shows the histograms of the AE and AE differ- 1.00 ± 0.80 for the coarse mode, in the wavelengths ranges ence (ground-satellite) against the ratio accumulation (small) 0.47–0.86𝜇 m. These particle size mode values and STDs are mode volume against large mode volume (𝑉 /𝑉 ). The AE 𝑆 𝐿 lower than those of the corresponding ground data. Single difference tends to increase with the 𝑉 /𝑉 increase. As 𝑆 𝐿 process originating particles have smaller STDs, while mul- noticed before both satellites’ algorithms use smaller STDs tiple processes originating particles have wider log-normal at the accumulation mode than the ground observed STDs, distributions [37]. The smaller accumulation mode STDs of though the mode radius itself is not very dieff rent from that both satellites’ algorithms compared to the ground data may of the ground. eTh STD of the accumulation mode may imply that the aerosols models retrievals do not take into increase the uncertainty in the satellites retrieved AE by both account most of the processes at the origin of these particles. algorithms. As a consequence, the particle size of the satellite retrieved The particle shape (nonsphericity) can also be a potential aerosols may dieff r from that of the ground. In some areas contributor to the inaccuracy of satellite aerosols retrievals. like semiarid areas, the effective radius of the accumulation According to Levy et al. [39] the MODIS-NASA ocean mode particle may increase with increasing AOT [38]. This algorithm does not perform well in dust-laden atmosphere, tendency was also noticed in the ground data of the present and this problem is attributed to poor assumptions of the AE and Dif <0.1 0.15 0.2 AE and Dif 0.25 0.3 <0.1 0.35 0.15 >0.35 0.2 AE and Dif 0.25 0.3 <0.1 0.35 0.15 >0.35 0.2 0.25 0.3 0.35 >0.35 12 Advances in Meteorology nonsphericity of the particles in the dust-aerosol phase tendstodecreasebeyond1.5 (closertothe values used by the functions. The peak in dust (where particles are less spherical) MODIS-GLI algorithm). All these variations show that the amount occurring in the aerosol load of the study area during RIr is an important contributor to the enhancement of the spring does not particularly increase the difference between satellite AOT uncertainty. No trend was seen between the RIi the ground AOT or AE with the satellite derived data (as and the AOT RMSE nor between the RIr or RIi and the AE shown by Figure 5). Also the 1 : 1 good alignment of the AOT RMSE variation. of the MODIS-NASA algorithm with ground data does not 4.2.4. Impact of Scattering Parameters. The main property show a specific seasonal dependency. All of these may lead representing the scattering by aerosols is generally expressed to the conclusion that there is a less significant impact of the by the single scattering albedo (SSA). The scattering prefer- sphericity in the satellites’ algorithms examined compared ential direction may be either forward (+1 for pure forward to other algorithms. Furthermore, tests conducted with a scattering) or backward (–1 for pure backscattering). The nonspherical model, by Nakajima and Higurashi [23], on proportion of light scattered in the forward direction is the MODIS-GLI algorithm suggest that the nonspherical defined as the asymmetry factor (AF). For satellites’ aerosol scattering eect ff is limited in this algorithm. retrievals, the SSA and AF are precomputed using a Mie- based theory (spherical particles) scattering model. Irregu- 4.2.3. Impact of the Refractive Index. In a given location, larly shaped tropospheric particles may have larger SSA and atmospheric aerosols are characterized by their concentra- smaller AF than their equal volume spherical counterparts tion, their size distribution, their shape, their vertical profile, [40]. The average SSA ± STD and AF± STD observed by and by their chemical composition [2]. The latter is indicative the SKYNET ground sites are, respectively,0.95±0.06 and of the magnitude of the real and imaginary parts of the 0.69±0.04 for Fukue,0.98±0.05 and0.70±0.04 for Hedo, refractive index (RI) of aerosols. eTh real part of the RI and0.96±0.07 and0.71±0.05 for Miyako. Figure 9 shows (RIr) is spectrally related to the scattering of the radiation the variation of the AF against the AE and the AE Dif. eTh while the imaginary part (RIi) is a consequence of the ground AE consistently decreases with the increase of the AF absorption of radiation, with a null value meaning the particle at the three ground sites; that is, the increase in the forward does not absorb radiation. In the satellite aerosols retrieval scattering of light results from the increase of the particle size. scheme, the aerosol model assumes specific realistic values No such clear trend is observed between the ground AF and for the real and imaginary parts of the RI. In the wavelength both, the satellites’ AE of the GLI and NASA algorithms. As range 0.47–0.86𝜇 m, the MODIS-GLI algorithm uses aerosol a result, the ground-satellite AE discrepancies decrease as the models with a single RI (1.5–i0.005), based on field data AF increases that is, as the scattering of light in the forward analyses. eTh real part RI of this algorithm is close to general direction increases. eTh uncertainty in the determination of tropospheric aerosols indices [40]. Meanwhile, the MODIS- the AF appears then to impact both satellites’ algorithms AE NASA algorithm uses a combination of models where the retrievals. eTh ground AF slope change with the AF is not RIs are set to vary according to the area and the volume size as steep in the northern Fukue as in the southern Hedo and distribution mode ratio. The values used by this algorithm are Miyako. This is consistent with the increase in coarse particles derived from the sun/sky ground photometers of AERONET from the north to the south (increased forward radiation [9], 1.45–i0.0035 or 1.40–i0.0020 forthe nfi emodeand 1.35– scattering) and consequently the AE RMSE. No consistent i0.001 for the coarse mode, of all aerosols types except dust- correlationwasfoundbetweentheSSAandeither,theground like types where the RIs are 1.53–i0.003 and 1.53–i0.000 AE, the AOT or the error in the determination of the satellite for 0.47𝜇 mand 0.86𝜇 m wavelengths, respectively [41]. The retrieved AOT and AE. This could be related to the fact that average RIs at 0.5𝜇 m wavelength obtained from ground data the influence of error in the retrieved SSA is important only for the whole study period are 1.44–i0.0043, 1.43–i0.0016, for thick layers [36]. And the data used here are mostly thin 1.44–i0.0029 for Fukue, Hedo, and Miyako sites, respectively. aerosols layers. The MODIS-NASA real and imaginary parts are closer or within the ranges of the ground RI values than the MODIS- 4.2.5. Impact of Solar and Satellite Zenith Angles. An ideal GLI are to the latter. The following distribution of RI at the aerosol algorithm would retrieve AOTs of equal quality, ground siteswas observed (inbracketsare thepercentage independent of solar and observing geometry [42]. No of data at Fukue, Hedo,and Miyako,respectively, forthe consistent trend showing possible or signicfi ant satellite bias RI range considered): the real part RI: <1.40 (33%, 41%, was observed between both the AOT and AE (AOT and 36%), 1.4–1.5 (37%, 29%, 32%),>1.5 (30%, 30%, 32%); and AE RMSE) and the solar zenith angle or the satellite zenith the imaginary part RI:<0.002 (54%, 84%, 75%), 0.002–0.005 angle. Using collection 5 MODIS dark target data over land, (22%, 7%, 10%),>0.005 (24%, 9%, 15%). Only one-third of Levy et al. [42] suggest that there is only a slight over- and the data are beyond RIr 1.5, and less than 25% have RIi above underestimate of the AOT by∼0.01, respectively, on the sun- 0.005 (RI values considered for the MODIS-GLI). glint side and the sun-shadow side of the MODIS swath. eTh comparison with the AOT shows that ( Figure 8)the RIr tends to increase with the AOT increase, for the ground 5. Implications for Future and both satellites’ algorithms data. eTh AOT RMSE is high Satellites’ Algorithms at high RIr and vice versa for low RIr values such as 1.4 (closer to the values used in the MODIS-NASA algorithm). For the The RMSE and Dif for the AOT retrieved from the MODIS-GLI, the AOT RMSE is high in the 1.4–1.45 range and MODIS-GLI algorithm are higher compared to those of the Advances in Meteorology 13 Hedo (RIr 0.5 𝜇 m versus AOT) Fukue (RIr 0.5 𝜇 m versus AOT) 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 RIr 0.5 𝜇 m RIr 0.5 𝜇 m AOT ground AOT NASA AOT ground AOT NASA AOT GLI AOT RMSE AOT GLI AOT RMSE GLI GLI (a) (b) Miyako (RIr 0.5 𝜇 m versus AOT) 0.8 0.6 0.4 0.2 RIr 0.5 𝜇 m AOT ground AOT NASA AOT RMSE AOT GLI GLI (c) Figure 8: Real part of the refractive index (RIr) variation at the ground against ground AOT, and the satellite retrieved AOT and AOT RMSE. MODIS-NASA algorithm. The accuracies of both algorithms of output data. Changes in aerosols mixtures of the region for the AE are much closer to each other as well as lower would need a better representation. For example, this study than the AOT accuracies. Some of the factors explaining these showed that the aerosol particle radii deviations from the accuracy differences range from the cloud contamination, the mode should be wider in order to integrate the dynamic particle size and shape distribution, the refractive index, and processes at the origin of the aerosols of this region. Also, a the scattering parameters to the sun-satellite geometry. The wider range (than that already used by the MODIS-NASA lessons drawn from the possible impact of these parameters algorithm) for the refractive indices would be necessary on satelliteretrievalswillhelptoimprove theaccuracyof to reduce the uncertainty in the AE. eTh high spatial and future satellites’ aerosols products such as those from the spectralresolutionoftheGCOM-C/SGLIsatellite(e.g.,250 m coming GCOM-C/SGLI satellite. at all visible and thermal infrared channels) should help to better discriminate clouds from cloud-free areas than actual eTh cloud contamination was found to be stronger in the AOT of MODIS-GLI than that of MODIS-NASA. This satellites do and therefore reduce possible cloud contamina- was explained by the differences in the spatial resolution of tion of the aerosols products. The improvement in the AOT accuracy will certainly increase that of the AE. eTh increase the retrievals: pixel-by-pixel retrieval in the MODIS-GLI and pixels’ selection retrieval from boxes of 10× 10 pixels in the in the number of ground sites measurements of aerosols such as in the SKYNET, AERONET, and other networks, provid- MODIS-NASA. To avoid this contamination in the future ing better quality data and quantitative characterization of and in case the pixel-by-pixel approach is chosen, a cloud fraction analysis should accompany the aerosols retrievals aerosols, would be of great asset for the improvement of the assumptions used in aerosols models: predetermined particle andbeusedtocorrect theaerosolsproperties’ inaccuracies. The MODIS-NASA approach looks attractive as retrievals are size distribution (modes’ radii and volumes), refractive index, made only on the high probability clear pixels of a wider asymmetry factor, single scattering albedo, ambient humidity andparticlegrowth, andsoforth. pool of pixels, but the inconvenience is the lower amount AOT and RMSE 1.37 1.4 1.43 AOT and RMSE 1.46 1.49 1.37 1.52 1.4 >1.52 1.43 AOT and RMSE 1.46 1.49 1.37 1.52 1.4 >1.52 1.43 1.46 1.49 1.52 >1.52 14 Advances in Meteorology Fukue (AF versus AE) Hedo (AF versus AE) 1.2 1.2 0.8 0.8 0.4 0.4 −0.4 −0.4 AF AF AE Dif GLI AE Dif GLI AE ground AE ground AE Dif NASA AE Dif NASA AE GLI AE GLI AE NASA AE NASA (a) (b) Miyako (AF versus AE) 1.2 0.8 0.4 −0.4 AF AE Dif GLI AE ground AE Dif NASA AE GLI AE NASA (c) Figure 9: Aerosol asymmetry factor (AF) histogram distribution against AE and the satellites’ algorithms AE, and the difference ground- satellite. 6. Conclusion ground and MODIS-NASA AOT and AE underestimation by both algorithms. eTh large particle AE overestimation The quality of aerosols retrievals from future satellites was overwhelmed by the small particle underestimation. eTh was discussed in this study, through a validation study of seasonal variation of the aerosols parameters (AOT and AE) aerosols products from actual satellites (Terra- and Aqua- based on monthly averages showed similar trends/variations MODIS) having compatible channels with those of the though the direct values differed, sometimes with MODIS- coming GCOM-C/SGLI. Aerosols products derived from two NASA or more oen ft with MODIS-GLI. Also nearly constant algorithms (MODIS-GLI and MODIS-NASA) using these monthly AOT differences between the ground aerosols data satellites were evaluated against ground-truth data from 3 and the satellite retrievals were observed. The AE showed SKYNET sites located in the Pacific East Asian region. eTh closer values between the ground and satellite data during results obtained showed a systematic overestimation of the winter and spring for the Fukue site and summer and autumn AOTbythe MODIS-GLIand abetteragreement betweenthe for the southern sites (Hedo and Miyako). eTh eeff ct of cloud AE and Dif 0.63 0.65 0.67 0.69 AE and Dif 0.71 0.73 0.63 0.75 0.65 >0.75 0.67 0.69 AE and Dif 0.71 0.73 0.63 0.75 0.65 >0.75 0.67 0.69 0.71 0.73 0.75 >0.75 Advances in Meteorology 15 contamination on the satellites retrievals, examined through retrieval algorithm over Central Europe,” Atmospheric Measure- ment Techniques,vol.3,no. 5, pp.1255–1270,2010. the use of an independent variable, the cloud fraction, showed that cloud contamination was more pronounced on [8] P. Goloub, D. Tanre´, J. L. Deuze, M. Herman, A. Marchand, and F. Breon, “Validation of the first algorithm applied for deriving theAOT MODIS-GLIthanthatofMODIS-NASA. eTh AE the aerosol properties over the ocean using the POLDER/ analysis showed that this parameter decreased (particles get ADEOS measurements,” IEEE Transactions on Geoscience and larger) with the CF for both algorithms. The difference in Remote Sensing,vol.37, no.3,pp. 1586–1596, 1999. the AOT contamination between the two satellites’ algo- [9] L.A.Remer,Y.J.Kaufman,D.Tanre´ et al., “eTh MODIS aerosol rithms was found to be mostly due to the retrieval pixel algorithm, products, and validation,” Journal of the Atmospheric resolution difference: pixel-by-pixel for the MODIS-GLI, Sciences,vol.62, no.4,pp. 947–973, 2005. therefore higher probability of contamination compared to [10] A. K. Heidinger, V. R. Anne, and C. Dean, “Using MODIS the pixels’ selection from larger boxes for the MODIS-NASA. to estimate cloud contamination of the AVHRR data record,” Also, this difference could be explained by the fact that the JournalofAtmospheric andOceanic Technology,vol.19, no.5, MODIS-NASA uses wider spectral information to estimate pp.586–601,2002. the aerosol model and, may perform more accurate retrieval [11] M. Jin and R. E. Treadon, “Correcting the orbit drift eeff ct on of spectral AOT [36]. The AE distribution showed that when AVHRR land surface skin temperature measurements,” Inter- thegrounddatahaveabimodaldistribution, theuncertainty national Journal of Remote Sensing,vol.24, no.22, pp.4543– in the AE retrievals by the two satellites’ algorithms is 4558, 2003. higher.Theshape of theparticlesizedistributionexamined [12] J. R. Dim, H. Murakami, T. Y. Nakajima, B. Nordell, A. K. showed larger STDs than those considered by the satellites’ Heidinger, and T. Takamura, “eTh recent state of the climate: algorithms, leading to limited mixtures in the aerosols sizes driving components of cloud-type variability,” JournalofGeo- retrieved by the satellites’ algorithms. Another contributor physical Research D, vol. 116, no. 11, Article ID D11117, 2011. to the ground-satellite differences was the uncertainty on [13] T. Takemura, T. Nakajima, A. Higurashi, S. Ohta, and N. the satellite retrievals, due to the asymmetry factor (AF). Sugimoto, “Aerosol distributions and radiative forcing over the A negative correlation was found with both the ground AE Asian Pacific region simulated by Spectral Radiation-Trans- andthe AE dieff rence(ground-satellite)against theAF. eTh port Model for Aerosol Species (SPRINTARS),” Journal of increase of the light scattering in the forward direction is a Geophysical Research D,vol.108,no. 23,pp. 27–10, 2003. consequence of the increasing size of the aerosols particles. [14] B. C. Bush and F. P. J. Valero, “Surface aerosol radiative forcing at The real part of the refractive index was found to be correlated Gosan during the ACE-Asia campaign,” Journal of Geophysical with the AOT RMSE of both satellites’ algorithms and not Research, vol. 108, no. D23, p. 8660, 2003. with the AE RMSE. To end this study, some measures aimed [15] B.J.Huebert,T.Bates,P.B.Russell et al., “AnoverviewofACE at improving the accuracy of future satellites’ algorithms were Asia: strategies for quantifying the relationships between Asian discussed. These involve remedies to cloud contamination aerosols and their climatic impacts,” Journal of Geophysical and corrections of the particle size distribution and scattering Research, vol. 108, no. D23, p. 8633, 2003. models. [16] I. Uno, “Regional chemical weather forecasting system CFORS: model descriptions and analysis of surface observations at Japanese island stations during the ACE-ASIA experiment,” References Journal of Geophysical Research: Atmospheres, vol. 108, no. D23, [1] K. N. Liou, An Introduction to Atmospheric Radiation, Academic [17] T. Nakajima, S. Yoon, V. Ramanathan et al., “Overview of the Press, 2nd edition, 2002. Atmospheric Brown Cloud East Asian Regional Experiment [2] G.A.d’Almeida,P.Koepke, andE.P.Shettle, Atmospheric Aero- 2005 and a study of the aerosol direct radiative forcing in east sols, Global Climatology and Radiative Characteristics,Deepak, Asia,” Journal of Geophysical Research D,vol.112,no. 24,Article Hampton, Va, USA, 1991. ID D24S91, 2007. [3]L.A.Remer,D.Tanre´, Y. J. Kaufman et al., “Validation of [18] T. Takamura, N. Sugimoto, A. Shimizu et al., “Aerosol radiative MODIS aerosol retrieval over ocean,” Geophysical Research characteristics at Gosan, Korea, during the Atmospheric Brown Letters, vol. 29, no. 12, pp. MOD3-1–MOD3-4, 2002. Cloud East Asian Regional Experiment 2005,” Journal of Geo- [4] D.A.Chu,Y.J.Kaufman,C.Ichoku, L. A. Remer, D. Tanre, ´ physical Research D: Atmospheres,vol.112,no. 22,Article ID and B. N. Holben, “Validation of MODIS aerosol optical depth D22S36, 2007. retrieval over land,” Geophysical Research Letters,vol.29, no.12, [19] I. Sano,S.Mukai,M.Yamano, T. Takamura,T.Nakajima, and pp. MOD2-1–MOD2-4, 2002. B. Holben, “Calibration and validation of retrieved aerosol [5] C.Ichoku, D. AllenChu,S.Mattooetal.,“Aspatio-temporal properties based on aeronet and skynet,” Advances in Space approach for global validation and analysis of MODIS aerosol Research, vol. 32, no. 11, pp. 2159–2164, 2003. products,” Geophysical Research Letters,vol.29, no.12, pp.1–4, [20] T. Nakajima, G. Tonna, R. Rao, P. Boi, Y. Kaufman, and B. Holben, “Use of sky brightness measurements from ground for [6] A.Ignatov,L.Stowe,R.Singh,S.Sakerin, D. Kabanov, and remote sensing of particulate polydispersions,” Applied Optics, I. Dergileva, “Validation of NOAA/AVHRR aerosol retrievals vol. 35, no. 15, pp. 2672–2686, 1996. using sun-photometer measurements from R/V akademik ver- nadsky in 1991,” Advances in Space Research,vol.16, no.10, pp. [21] Y. Liu, J. Huang, G. Shi et al., “Aerosol optical properties 95–98, 1995. and radiative effect determined from sky-radiometer over [7] M.Riffler, C. Popp,A.Hauser, F. Fontana, andS.Wun- Loess Plateau of Northwest China,” Atmospheric Chemistry and derle, “Validation of a modified AVHRR aerosol optical depth Physics,vol.11, no.22, pp.11455–11463,2011. 16 Advances in Meteorology [22] J. Bi, Y. Liu, J. Huang et al., “Characteristics of dust aerosols the Royal Meteorological Society,vol.129,no. 590, pp.1395–1413, derived from sky-radiometer over Loess Plateau of Northwest 2003. China,” in Proceedings of the16thCEReS InternationalSympo- [38] L. A. Remer, D. Tanre´, andY.J.Kaufman,“Algorithmfor remote sium on Climate Change Studies through Activities of SKYNET sensing of tropospheric aerosols from MODIS,” collection 5,87 and Virtual Laboratory for Climate Diagnostics,Nago, Okinawa, pp, 1998. Japan, October 2010. [39] R. C. Levy,L.A.Remer,D.Tanre´ et al., “Evaluation of the [23] T. Nakajima and A. Higurashi, “A use of two-channel radi- MODIS retrievals of dust aerosol over the ocean during PRIDE,” ances for an aerosol characterization from space,” Geophysical J. Geophys. Res,vol.108,no. D19, 2003. Research Letters, vol. 25, no. 20, pp. 3815–3818, 1998. [40] J. B. Pollack and J. N. Cuzzi, “Scattering by nonspherical [24] A. Higurashi and T. Nakajima, “Detection of aerosol types over particles of size comparable to a wavelength: a new semi- theEastChinaSea near Japanfromfour-channelsatellite data,” empirical theory and its application to tropospheric aerosols,” Geophysical Research Letters,vol.29, no.17, pp.17–1, 2002. Journal of the Atmospheric Sciences,vol.37, pp.868–881,1980. [25] C. Ichoku,Y.J.Kaufman,L.A.Remer,and R. Levy,“Global [41] R. C. Levy,L.A.Remer,D.Tanre, ´ S. Mattoo, and Y. J. Kaufman, aerosol remote sensing from MODIS,” Advances in Space Re- “Algorithm for remote sensing of tropospheric aerosols over search,vol.34, no.4,pp. 820–827, 2004. dark targets from MODIS,” Collections005 and051:Revision2, Feb 2009, Product ID: MOD04/MYD04,96pp, 2009. [26] W. A. Abdou, D. J. Diner, J. V. Martonchik et al., “Comparison of coincident Multiangle Imaging Spectroradiometer and Mod- [42] R. C. Levy,L.A.Remer,R.G.Kleidman et al., “Global erate Resolution Imaging Spectroradiometer aerosol optical evaluation of the Collection 5 MODIS dark-target aerosol depths over land and ocean scenes containing Aerosol Robotic products over land,” Atmospheric Chemistry and Physics,vol.10, Network sites,” Journal of Geophysical Research D,vol.110,no. no.21, pp.10399–10420,2010. 10,pp. 1–12,2005. [27] O. V. Kalashnikova and R. A. Kahn, “Mineral dust plume evolution over the Atlantic from MISR and MODIS aerosol retrievals,” JournalofGeophysical Research D, vol. 113, no. 24, Article ID D24204, 2008. [28] A. Higurashi and T. Nakajima, “Development of a two-channel aerosol retrieval algorithm on a global scale using NOAA AVHRR,” Journal of the Atmospheric Sciences,vol.56, no.7,pp. 924–941, 1999. [29] A. Higurashi, T. Nakajima, B. N. Holben, A. Smirnov, R. Frouin, and B. Chatenet, “A study of global aerosol optical climatology with two-channel AVHRR remote sensing,” Journal of Climate, vol. 13,no. 12,pp. 2011–2027, 2000. [30] Y. J. Kaufman and D. Tanre, ´ “Algorithm for remote sensing of tropospheric aerosol from MODIS. Products: MOD04, MOD08. ATBD reference number: ATBD-MOD-02,” 1998, http://modis.gsfc.nasa.gov/data/atbd/atbd mod02.pdf. [31] F. Wagner and A. M. Silva, “Some considerations about Angstrom ¨ exponent distributions,” Atmospheric Chemistry and Physics,vol.8,no. 3, pp.481–489,2008. [32] T. W. Ault,K.P.Czajkowski, T. Benkoetal.,“Validation of the MODIS snow product and cloud mask using student and NWS cooperative station observations in the Lower Great Lakes Region,” Remote Sensing of Environment,vol.105,no. 4, pp.341– 353, 2006. [33] S. A. Ackerman,R.E.Holz, R. Frey,E.W.Eloranta, B. C. Maddux, and M. McGill, “Cloud detection with MODIS. Part II: validation,” JournalofAtmospheric andOceanic Technology, vol. 25,no. 7, pp.1073–1086,2008. [34] R. Santer and D. Ramon, MERIS Aerosol Remote Sensing over Land ATBD, 2011. [35] J. N. Porter and A. D. Clarke, “Aerosol size distribution models basedoninsitumeasurements,” Journal of Geophysical Research D,vol.102,no. 5, pp.6035–6045,1997. [36] A. A. Kokhanovsky, J. L. Deuze, ´ D. J. Diner et al., “The inter- comparison of major satellite aerosol retrieval algorithms using simulated intensity and polarization characteristics of reflected light,” Atmospheric Measurement Techniques,vol.3,pp. 909– 932, 2010. [37] L. Alados-Arboledas, H. Lyamani, and F. J. Olmo, “Aerosol size properties at Armilla, Granada (Spain),” Quarterly Journal of International Journal of Journal of Ecology Mining The Scientific Journal of Geochemistry Scientifica World Journal Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Earthquakes Paleontology Journal Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Petroleum Engineering Submit your manuscripts at http://www.hindawi.com International Journal of Geophysics Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Advances in Advances in Journal of Advances in International Journal of Meteorology ClimatologyGeology Oceanography Oceanography Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Applied & Journal of Environmental International Journal of Journal of International Journal of Computational Soil Science Mineralogy Geological Research Atmospheric Sciences Environmental Sciences Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com http://www.hindawi.com Volume 2014 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Advances in Meteorology Hindawi Publishing Corporation

Validation of Two MODIS Aerosols Algorithms with SKYNET and Prospects for Future Climate Satellites Such as the GCOM-C/SGLI

Loading next page...
 
/lp/hindawi-publishing-corporation/validation-of-two-modis-aerosols-algorithms-with-skynet-and-prospects-e604QendB4
Publisher
Hindawi Publishing Corporation
Copyright
Copyright © 2013 Jules R. Dim et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN
1687-9309
eISSN
1687-9317
DOI
10.1155/2013/508064
Publisher site
See Article on Publisher Site

Abstract

Hindawi Publishing Corporation Advances in Meteorology Volume 2013, Article ID 508064, 16 pages http://dx.doi.org/10.1155/2013/508064 Research Article Validation of Two MODIS Aerosols Algorithms with SKYNET and Prospects for Future Climate Satellites Such as the GCOM-C/SGLI 1 2 3 2 Jules R. Dim, Tamio Takamura, Akiko Higurashi, Pradeep Kathri, 3 4 Nobuyuki Kikuchi, and Takahashi Y. Nakajima Earth Observation Research Center/JAXA, 2-1-1 Sengen, Tsukuba, Ibaraki 305-8505, Japan Center for Environmental Remote Sensing (CEReS), Chiba University, 1-33 Yayoi-Cho, Inage-Ku, Chiba 263-8522, Japan Center for Environmental Measurement and Analysis, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan Department of Network and Computer Engineering, Tokai University, 2-28-4 Tomigaya, Shibuya-ku, Tokyo 151-0063, Japan Correspondence should be addressed to Jules R. Dim; rosutando@yahoo.com Received 14 April 2013; Revised 10 June 2013; Accepted 11 June 2013 Academic Editor: Harry D. Kambezidis Copyright © 2013 Jules R. Dim et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Potential improvements of aerosols algorithms for future climate-oriented satellites such as the coming Global Change Observation Mission Climate/Second generation Global Imager (GCOM-C/SGLI) are discussed based on a validation study of three years’ (2008–2010) daily aerosols properties, that is, the aerosol optical thickness (AOT) and the Angstrom ¨ exponent (AE) retrieved from two MODIS algorithms. The ground-truth data used for this validation study are aerosols measurements from 3 SKYNET ground sites. The results obtained show a good agreement between the ground-truth data AOT and that of one of the satellites’ algorithms, then a systematic overestimation (around 0.2) by the other satellites’ algorithm. The examination of the AE shows a clear underestimation (by around 0.2–0.3) by both satellites’ algorithms. eTh uncertainties explaining these ground-satellites’ algorithms discrepancies are examined: the cloud contamination aeff cts differently the aerosols properties (AOT and AE) of both satellites’ algorithms due to the retrieval scale differences between these algorithms. eTh deviation of the real part of the refractive index values assumed by the satellites’ algorithms from that of the ground tends to decrease the accuracy of the AOT of both satellites’ algorithms. eTh asymmetry factor (AF) of the ground tends to increase the AE ground-satellites discrepancies as well. 1. Introduction and local variation. The most common aerosols optical and physical properties used for this characterization are the The determination of the optical properties of aerosols and AOT, theAE, theparticlesizedistribution, thesinglescatter- their size distribution around the globe has been a significant ing albedo (SSA), the aerosol phase function, the asymmetry contemporary research effort of late [ 1]. Some of the major factor (AF), the refractive index (RI), and so forth. To obtain factorsthathaveenabled this progress arethe better spectral these aerosols characteristics from satellites’ observations, a and spatial capacities of satellites and ground-based radiome- detailed model of the aerosols properties is required [2]. The ters, the improvement of the aerosol signal filtering methods, increasing number and spatial expansion of field measure- the better knowledge of the aerosols particles’ shapes, and so ment campaigns and ground sites’ coverage have helped to forth. The qualitative and quantitative importance of the data gather statistically meaningful data for the development of collected through the continuous monitoring of aerosols and aerosols models. Ground collected data not only serve as their daily global coverage, by various satellites, has permitted preliminary input for aerosols models but also are reliable a better characterization of the role of the aerosols in the evaluation and validation tools for aerosols products derived climate dynamics and the understanding of their temporal from satellites (GMS-5/SVISSR, NOAA-AVHRR, Terra- and 2 Advances in Meteorology Aqua-MODIS and MISR, OrbView-2/SeaWiFS, ENVISAT- settings andmotives of thestudy;(2) the description of the MERIS, etc.). eTh objective of these validation studies is ground data collection approach and the satellites’ algorithms to improve the quality of the aerosols properties retrieved main features;(3) the presentation of the main characteristics from satellites’ observations as well as the confidence level in of theaerosolspropertiesofthe studyarea; (4)the validation theseproducts. Sincethe launch of Terra- andAqua-MODIS, scheme of the satellites’ retrievals, and, discussion on the ground data from AERONEThavebeenusedtovalidate uncertainties plaguing these retrievals; (5) the lessons for and improve the MODIS AOT over ocean [3] and land future satellites’ aerosols products; (6) the summary of the [4, 5]. NOAA-AVHRR AOT data have been also evaluated results obtained. using sun photometers of AERONET and others [6, 7]. The retrieval quality over the ocean of POLDER/ADEOS aerosols measurements has been also conducted through comparisons 2. Ground Data Collection and with AERONET sun photometer data [8]. Satellites’ Algorithms The increasing accuracy of aerosols products, as a result of all these evaluation/validation and correction studies, The analyses conducted in this study aim at validating couldbeusedtoreducethe uncertaintiesassociatedwiththe tropospheric aerosols’ retrievals from two algorithms, using aerosol radiative forcing of the global climate [9]. Due to the similar observations from both Terra- and Aqua-MODIS complex and diverse composition, size, structure, and origin satellites.Thesesatelliteshavespectrallycompatiblechannels of aerosols and though a lot of progress has been made in the with the future GCOM-C/SGLI satellite. This validation study satellites’ observations, there are still various challenges for a is conducted through a comparison of the satellites’ retrievals globally acceptable accuracy of the aerosol optical and micro- with ground-truth measurements from three SKYNET sites: ∘ ∘ physical properties of different types of aerosols. The high Fukue (32.752 N latitude, 128.682 Elongitude,and 50m ∘ ∘ spectral and spatial resolutions of climate-oriented satellites altitude), Hedo (26.867 N latitude, 128.249 Elongitude,and ∘ ∘ such as the GCOM-C/SGLI, scheduled to be launched around 65 m altitude), and Miyako (24.737 Nlatitude, 125.327 E 2015 by the Japan Aerospace Exploration Agency (JAXA), longitude, and 50 m altitude). es Th e sites are islands located aim at accomplishing better distinctions between aerosols in the Pacicfi East Asian region, south of Japan. eTh choice particles and other atmospheric components. However, the of these three locations for the satellites’ retrieved aerosols programmed long lifespan of such satellites (3 series) and properties’ evaluation study derives from the fact that they the connectivity between similar/different satellites’ series are at the confluence of one of the most complex mixtures will pose other accuracy issues proper to long-term climate of aerosols types. It is known that the Asian atmospheric observations [10–12] that may only be alleviated with better environment has been worsened rapidly by Asian dust, accuracy retrieval algorithms. anthropogenic gases, and aerosols in recent years [13]; it In the present study, the potential performance of future has been also the subject of various aerosols projects. Some satellites’ products is discussed through a validation study of the examples are the Asian-Pacific Regional Aerosol of aerosol retrievals by present satellites’ (Terra- and Aqua- Characterization Experiment (ACE-Asia) [14–16]which was MODIS satellites) algorithms using spectrally compatible a multiplatform project where aerosols, radiative u fl xes channels with those of the coming GCOM-C/SGLI. This val- measurements were conducted over the East China Sea; idation is conducted against a three-year (2008–2010) daily and the Atmospheric Brown Cloud East Asian Regional ground-truth dataset from three SKYNET observation sites Experiment 2005 [17, 18], with the study of aerosols radiative (Fukue,Hedo, andMiyako).Thesegroundsites arelocated characteristics and aerosol direct radiative forcing. in the Pacific East Asian region, an area at the confluence of eTh SKYNET,whose data areusedtovalidateagainst actively changing and mixed aerosols (natural and anthro- the satellite retrievals in this study, is a well-developed and pogenic particles resulting from rapid industrialization). eTh maintained network of climate radiation stations with ground retrievals examined are from two algorithms using similar sites spread all over Japan and the East Asian region [17, 19]. observations, Terra- and Aqua-MODIS satellites’ calibrated It serves for the monitoring of aerosol and cloud proper- radiances (MOD/MYD021, version 5.1). eTh rfi st algorithm ties as well as other weather/climate variables. The basic is named here as MODIS-GLI, as it was previously applied aerosols properties at the SKYNET sites are obtained from to the Global Imager (GLI) onboard the Advanced Earth radiation measurements performed with sky radiometers. Observation Satellite-II (ADEOS-II). eTh second algorithm These radiometers record direct sky radiance intensities at is the MODIS-NASA algorithm, used for MODIS stan- 11 predenfi edwavelengths (315,340,380,400,500,675,870, dard products’ retrievals by the National Aeronautics and 940, 1020, 1600, and 2200 nm) and scattering angles. eTh Service Administration (NASA). The aerosols properties measurement frequency is 10 or 15 minutes. The solar direct- investigated are the aerosol optical thickness (AOT) and beam transmission and sky radiance obtained from sky the Angstrom ¨ exponent (AE). The differences between the radiometer measurements are used for aerosols properties satellites’ retrievals and ground observations are analyzed, such as theAOT,the aerosolcolumnsizedistributions,the issues plaguing the accuracy of each of these satellites’ complex refractive indices, the single scattering albedo, and algorithms are discussed, and lessons for future satellites’ the asymmetry factor, at multiple wavelengths. A program algorithms improvements, such as the algorithm of the future package named Skyrad.pack [20]isusedfor theretrieval of GCOM-SGLI, are drawn. To conduct this work, the following these parameters. The SKYNET helps to extend the global steps are adopted:(1) the introduction, highlighting the basic aerosol monitoring from the ground to locations not always Advances in Meteorology 3 covered by the better known AErosol RObotic NETwork observed and theoretical reflectance, are iteratively searched. (AERONET). Previous studies have shown a good match The 𝛿 parameter, representing the contribution of the large between the SKYNET and the AERONET aerosols mea- particles in the aerosol total load, is used to calculate the AE. surements in areas where both networks coexist, confirming For the MODIS-NASA algorithm, the retrieval scheme the quality and consistency of the SKYNET measurements. over the ocean, described by Kaufmann and Tanre[ ´ 30], One of these is a study conducted over the Loess Plateau uses seven wavelengths from 0.47 to 2.13𝜇 mtoretrievethe of Northwest China. eTh study compares the aerosol opti- AOT and the volume distribution (in the range of 0.08– cal depths at two common wavelengths (675 and 870 nm) 5𝜇 m radius) from measured radiances. LUTs from radiative observed by the CIMEL sun photometer of the AERONET transfer calculations of the MODIS radiance spectral bands and those observed by the POM-01 sky radiometer of the are used to retrieve aerosols parameters at each particle size SKYNET,based on 703clear-sky measurements during the distribution mode: the small particle mode or accumulation year 2009 [21]. The study shows a high correlation coefficient mode particles (5 cases) and the large or coarse mode (0.99) and a low relative difference of about 3% between particles (6 cases). The AOT and the volume distribution are themeasurementsofthe twoinstruments.Bietal. [22]also the primarily derived parameters from the algorithm, while showed a good match between the AOT retrieved by both the AE is (as in the MODIS-GLI algorithm) a secondarily networks at four major channels (440, 675, 870, and 1020). derived parameter. The AOT is obtained from the measured Both the SKYNET and AERONET aerosols networks radiances, by linear interpolation between vfi e optical thick- have been very oen ft used to validate and refine aerosols ness values of each aerosol particle mode combination. eTh models for the satellites’ retrievals [3, 4, 9, 23–26]. The two modes that give the best selected value (ratio between the multiplication of ground stations and the improvement of modes) are the selected aerosol models for the retrieval. thequality of thedataobtainedwiththe introduction of new parameterization schemes in the radiative transfer derivation 3. Aerosols Characteristics codes and a more rigorous cloud mask application have increased the confidence level in the use of these ground- Prior to the evaluation of the satellite retrievals against based data for satellite products’ validation purpose [25, 27]. ground data, a quick description of some of the major In this study, daily aerosol retrievals from both Terra- characteristics of the aerosols distribution observed in the and Aqua-MODIS at 0.5𝜇 m wavelength are analyzed. eTh region covering the ground sites selected for this study may matching time and space area between these satellite data be useful to understand the aerosols of the region. Figure 1 andthose from thegroundstationsare±30 min and 25 km shows the prevailing pattern of the AOT and AE distributions around each ground site (time and space average are applied), in the region, as retrieved by the MODIS-GLI algorithm respectively. eTh parameters studied are the AOT at 0.5 𝜇 m during winter (month of January 2010), and the geographical andthe AE from visiblechannels. eTh MODIS-NASA data location of the SKYNET ground sites selected. In this gur fi e, (MOD/MYD04, version 5.1) produce the AOT at 0.55𝜇 m. the aerosol distribution shows high AOTs near the coastal Based on the wavelength relationship between the AE and area of the southwest (adjacent to China) and in the northwest the AOT, the latter are interpolated to 0.5𝜇 m, in order to area representing the sea of Japan. But higher AE (finer match the retrievals from the MODIS-GLI algorithm and the particles) are visible in the former than the latter. Closer ground data AOT. ranges of both AOTand AE canbeseeninanorth-south The basic retrieval scheme of the MODIS-GLI algorithm band running along the SKYNET sites, with however thicker over the ocean, described in Nakajima and Higurashi [23], layers and n fi er particles in the north than the south. For Higurashi and Nakajima [28], and Higurashi et al. [29], the three-year data set matching the satellites’ overpass time, uses the visible (0.660𝜇 m) and near-infra-red (0.870𝜇 m) the northern site (Fukue) has the thickest aerosols (mean channels to retrieve theAOT.Thesatellite-receivedradiances AOT of 0.30±0.23) as well as the finest particles (mean are synthesized in four look-up tables (LUTs) derived from AE of0.97±0.35). The other 2 sites (Hedo and Miyako) radiative transfer calculations under the Mie theory scheme. located in the south have mean AOT values of 0.26± 0.20 and The algorithm assumes a complex refractive index (RI =1.5– 0.27±0.15,respectivelyand AE of0.95±0.35 and0.89±0.36, 0.005𝑖 ) and a bimodal size distribution expressed as respectively. As we move from the north to the south, that is, 2 further from the major coastal area and towards the central (ln𝑟− ln𝑟 ) 𝑑𝑉 (𝑟 ) parts of the Pacific Ocean ( Figure 1), the aerosol particle size =∑𝐶 exp{− }, (1) 𝑑 ln𝑟 2𝜎 𝑛=1 𝑛 increases (AE values get smaller). where𝑛 is thenumberoflognormal aerosolmodes,𝑟 is the median or geometric mean radius at each mode,𝜎 is the 4. Evaluation of Satellites’ Derived variance or width of the mode,𝐶 represents the particle peak Aerosols Data volume concentration of the mode; this is an undetermined parameter to be retrieved from satellite-received radiances. 4.1. Analyses and Results. The previous section examined the The AOT and the peak ratio 𝛿 of thesizedistributionare aerosols characteristics of the study eld fi , based on ground retrieved by comparing measured and theoretical (precalcu- observations. Now, let us use these characteristics to evaluate lated LUTs) reflectance values. eTh optimal values of AOT the AOT and AE retrievals from the satellites’ algorithms. and𝛿 , that minimize the root mean square deviation between Figures 2 and 3 present 1 : 1 scatter plots between daily 4 Advances in Meteorology January 2010 Terra-MODIS AOT from GLI algorithm Table 1: Error analyses showing the root mean square error 40N (RMSE) between the ground data and the satellite derived data for the MODIS-GLI algorithm (GLI RMSE) and the MODIS-NASA 38N algorithm (NASA RMSE), and the simple differences between the ground data and the MODIS-GLI (GLI Dif), then the MODIS- 1.8 36N NASA (NASA Dif) satellites’ algorithms. 1.6 34N AE AOT AE AOT 1.4 Sites RMSE GLI RMSE GLI RMSE NASA RMSE NASA 32N Fukue 0.411 0.214 0.321 0.138 1.2 Fukue Hedo 0.535 0.234 0.388 0.147 30N 1 Miyako 0.566 0.233 0.384 0.089 28N Sites Dif GLI Dif GLI Dif NASA Dif NASA 0.8 Fukue 0.181 −0.194 0.065 0.009 26N Hedo 0.6 Hedo 0.248 −0.203 0.150 −0.010 Miyako 0.145 −0.202 0.086 −0.008 24N 0.4 Miyako 22N 0.2 to that of MODIS-GLI. In both MODIS-GLI and MODIS- 20N NASA cases, the dispersion around the 1 : 1 correlation line 120E 122E 124E 126E 128E 130E 132E 134E 136E 138E 140E tends to increase at large AOT values (above 1). (a) For the AE, the graphs of Figure 3 show that the satellites’ retrievals from both algorithms are clearly underestimated. January 2010 Terra-MODIS AE from GLI algorithm 40N This underestimation is slightly larger with the MODIS-GLI data than with the MODIS-NASA, and its magnitude seems 38N 2 to vary with the proportion of small/large particles’ size of the aerosol load present in the ground data. eTh distribution of 1.8 36N ground AE data appears as two overlapping datasets (around 1.6 a cutoff value of 0.8) and, the increase in the number of 34N ground data with AE > 0.8 against those with AE < 0.8 1.4 leads to an increasing underestimation of the AE by the 32N satellite retrievals. In the large particle group, the satellite 1.2 Fukue 30N particles are smaller in size than the ground data, while in the 1 smallparticlegroup,the satellitedataare larger in size.And 28N as, in general, small particles tend to be in higher number 0.8 than large particles, among the ground data, the AE would 26N be underestimated. The determination coefficient in the AE Hedo 0.6 comparison is lower than that in the AOT comparison for 24N 0.4 both MODIS-GLI and MODIS-NASA cases. Among the 3 Miyako ground sites, the best match between the ground and both 22N 0.2 satellites’ algorithms AOT and AE is at Fukue (northernmost site). 20N 120E 122E 124E 126E 128E 130E 132E 134E 136E 138E 140E Table 1 shows the discrepancies between the ground measurements and the satellites’ retrievals, expressed as the (b) simple difference and the RMSE (root mean square error) of Figure 1: Geographical location of the three SKYNET radiation sites the AOT and AE. At the 3 sites, there are higher AOT RMSEs (Fukue, Hedo, Miyako) used in this study to evaluate the satellite and biases (against ground observations) with the MODIS- retrievals, over an image of satellite derived aerosols properties from GLI algorithm (ranges 0.214–0.234 and−0.194–−0.203, resp.) the MODIS-GLI algorithm, for the month of January. compared to the MODIS-NASA algorithm (ranges 0.089– 0.147 and−0.008–0.009, resp.). eTh AOT bias of the MODIS- GLI algorithm is consistently negative, that is, the average aerosols measurements from the ground (SKYNET sites) AOT from this algorithm is overestimated (by an average of timely matching the satellite retrievals (from MODIS-GLI around 0.2), while the bias of the MODIS-NASA shows only and MODIS-NASA algorithms) for the AOT and AE, respec- a slightly under/overestimated value of around±0.010. These tively, during the 3-year study period. Analyses of these plots AOT validation biases confirm the conclusions of Remer et show that the MODIS-GLI AOT values are systematically al. [9] indicating that the MODIS-NASA ocean algorithm overestimated compared to ground data, while the MODIS- has virtually no offset and little bias except for a slight NASA AOT values show a better match with these ground underprediction at high AOT, and the linear regression line data, as shown by the higher correlation coecffi ient compared follows the 1 : 1 line closely. Pacific Ocean Pacific Ocean Japan Japan Advances in Meteorology 5 AOT Fukue versus MODIS-GLI AOT Fukue versus MODIS-NASA 1.5 1.5 1 1 0.5 0.5 y = 0.7214x + 0.0825 y = 0.7676x + 0.2657 R = 0.45 R = 0.63 0 0 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Fukue Fukue (a) (b) AOT Hedo versus MODIS-GLI AOT Hedo versus MODIS-NASA 1.5 1.5 0.5 0.5 y = 0.837x + 0.2517 y = 0.6662x + 0.0951 R = 0.44 R = 0.57 0 0 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Hedo Hedo (c) (d) AOT Miyako versus MODIS-NASA AOT Miyako versus MODIS-GLI 1.5 1.5 0.5 0.5 y = 0.8722x + 0.0465 y = 0.8361x + 0.285 R = 0.89 R = 0.44 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Miyako Miyako (e) (f) Figure 2: Scatter diagrams between ground (Fukue, Hedo, Miyako) and satellites’ derived aerosols AOT. Ground data are on the 𝑥 -axis while satellite derived data (from both algorithms, MODIS-GLI and MODIS-NASA) are on the𝑦 -axis. AOT is the aerosol optical thickness. The uncertainty in the AE satellite estimation is expected to the total optical thickness magniefi s relative errors in the to be larger than that of the AOT. An expression of this retrievals of individual path radiances by up to 4 to 8 times uncertainty is [31] the original percentages [9]. The AE uncertainty values are much closer than the AOT errors; between the two satellites’ 1 Δ𝜏 Δ𝜏 algorithms, the RMSE and bias vary from 0.414–0.566 and 1 2 Δ𝛼= ( + ). (2) 0.145–0.248, respectively, for the MODIS-GLI, then 0.321– ln(𝜆 /𝜆 ) 𝜏 𝜏 2 1 1 2 0.388 and 0.065–0.150, respectively, for the MODIS-NASA ˚ (Table 1). Both algorithms show consistently a positive bias 𝛼 is the Angstrom ¨ exponent, and𝜏 and𝜏 are the aerosol 1 2 in the AE, that is, smaller AE (larger size of particles) than optical thickness at wavelengths𝜆 and𝜆 ;Δ𝛼 ,Δ𝜏 ,Δ𝜏 are 1 2 1 2 ˚ ground data. the errors in the Angstrom exponent and the aerosol optical Figure 4 presents the comparison between the frequency thickness at the wavelengths𝜆 and𝜆 ,respectively. 1 2 distribution of the ground measurements and the satellite As a rule, satellite retrievals of aerosol optical thickness derived data, for the AE and AOT. eTh MODIS-NASA are more robust than the corresponding retrievals of aerosol algorithm shows the closest AOT distribution range to that size; consequently, the fraction of the fine mode contribution MODIS-GLI MODIS-GLI MODIS-GLI MODIS-NASA MODIS-NASA MODIS-NASA 6 Advances in Meteorology AE Fukue versus MODIS-GLI AE Fukue versus MODIS-NASA 1.5 1.5 0.5 0.5 y = 0.3895x + 0.4473 y = 0.6295x + 0.3217 R = 0.37 R = 0.27 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Fukue Fukue (a) (b) AE Hedo versus MODIS-GLI AE Hedo versus MODIS-NASA 1.5 1.5 1 1 0.5 0.5 y = 0.1941x + 0.5047 y = 0.4503x + 0.353 R = 0.14 R = 0.25 0 0 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Hedo Hedo (c) (d) AE Miyako versus MODIS-GLI AE Miyako versus MODIS-NASA 2 2 1.5 1.5 1 1 0.5 0.5 y = 0.1362x + 0.5875 y = 0.3481x + 0.4774 R = 0.13 R = 0.2 0 0.5 1 1.5 2 0 0.5 1 1.5 2 Miyako Miyako (e) (f) Figure 3: Scatter diagrams between ground (Fukue, Hedo, Miyako) and satellites’ derived aerosols AE. Ground data are on the 𝑥 -axis while satellite derived data (from both algorithms, MODIS-GLI and MODIS-NASA) are on the𝑦 -axis. The AE is the Angstrom ¨ exponent. of the ground data with however a consistently low sensitivity data distributions appear to be bimodal. This distribution to low AOT values (<0.1). eTh principal mode occurs at the contrasts with that of the two satellites’ algorithms, as they samerangesasthatofthegrounddata,thatis,around0.2–0.3. show at all sites a unimodal distribution with the modes con- The difference in the data distribution between this algorithm sistently at 0.8 (for the MODIS-GLI) and 1.0 (for the MODIS- andthe ground AOTdistributionatmostofthe ranges is NASA), and fewer data are available beyond these modes. below 5%. Contrary to the MODIS-NASA algorithm, the This contributes to the underestimation of the AE in the high MODIS-GLI algorithm shows its highest AOT peak range AE ranges by the 2 satellites’ algorithms, overwhelming the at >0.7, contrary also to the ground data peak, which is overestimation in the low AE ranges. around 0.3. This MODIS-GLI higher range peak depletes the eTh comparison of the AOT, then AE seasonal distri- number of data in the lower AOT ranges and contributes bution between the ground and satellite data, is shown at to the AOT overestimation by this algorithm. Concerning Figure 5, as the monthly average of each parameter for the 3- the AE, it shows a unimodal distribution (mode at 1.2) at year study period. The MODIS-GLI algorithm overestimates theFukue ground site whilethe Hedo andMiyakoground the AOT nearly equally each month. A slight inflection of MODIS-GLI MODIS-GLI MODIS-GLI MODIS-NASA MODIS-NASA MODIS-NASA Advances in Meteorology 7 AE (Fukue) AOT (Fukue) 35 35 5 5 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 >1.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 Ranges Ranges AOT (Hedo) AE (Hedo) 0.2 0.4 0.6 0.8 1 1.2 1.4 >1.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 Ranges Ranges AOT (Miyako) AE (Miyako) 35 35 25 25 20 20 15 15 10 10 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.2 0.4 0.6 0.8 1 1.2 1.4 >1.4 Ranges Ranges AOT ground AE ground AOT GLI AE GLI AE NASA AOT NASA (a) (b) Figure 4: Comparison between the frequency distribution of ground m easurements and satellite derived (GLI for the MODIS-GLI algorithm, NASA for the MODIS-NASA algorithm) aerosols properties (AE (a), AOT (b)). this overestimation is seen in summer, around July. There is and Miyako), show closer values between the ground and a better match each month between the MODIS-NASA and both the MODIS-GLI and MODIS-NASA data. eTh AE is the ground data. mostly underestimated by the satellites’ algorithms. But this For the AE, winter months (January, February, March) underestimation is generally within the ground data monthly and summer months (July, August, September), respectively, variation range (standard deviation represented by vertical for the northern site (Fukue) and the southern sites (Hedo blue bars on the map). Number of data (%) Number of data (%) Number of data (%) Number of data (%) Number of data (%) Number of data (%) 8 Advances in Meteorology AE (Fukue) AOT (Fukue) 1 1.5 0.8 0.6 0.4 0.5 0.2 0 0 (a) (b) AE (Hedo) AOT (Hedo) 1.5 0.8 0.6 0.4 0.5 0.2 0 0 (c) (d) AOT (Miyako) AE (Miyako) 1 1.5 0.8 0.6 0.4 0.5 0.2 Ground Ground GLI GLI NASA NASA (e) (f) Figure 5: Average monthly variation of the ground (Ground) AOT at the 3 SKYNET sites (Fukue, Hedo, Miyako) and the corresponding satelliteretrievals: MODIS-GLI(GLI) andMODIS-NASA (NASA).Theerrorsbars(blue)represent thestandarddeviation of themonthly ground data. AOT AOT AOT Jan. Jan. Jan. Feb. Feb. Feb. Mar. Mar. Mar. Apr. Apr. Apr. May. May. May. Jun. Jun. Jun. Jul. Jul. Jul. Aug. Aug. Aug. Sep. Sep. Sep. Oct. Oct. Oct. Nov. Nov. Nov. Dec. Dec. Dec. AE AE AE Jan. Jan. Jan. Feb. Feb. Feb. Mar. Mar. Mar. Apr. Apr. Apr. May. May. May. Jun. Jun. Jun. Jul. Jul. Jul. Aug. Aug. Aug. Sep. Sep. Sep. Oct. Oct. Oct. Nov. Nov. Nov. Dec. Dec. Dec. Advances in Meteorology 9 4.2. Impact of External Factors and Aerosol Model Assump- with MODIS-NASA) tends to increase, substantially (Dif tions on Satellites’ Retrievals. As mentioned before, a prior GLI) or slightly (Dif NASA) with the CF increase. eTh AOT step in the retrieval of aerosols properties is the choice of an bias observed here may surpass 0.3 (e.g., at the CF peak of aerosol model. This model is based on various assumptions. 0.7).Theslighttrend betweenthe DifNASAand theCF, eTh uncertainties in these assumptions will likely affect the foundinthisstudy conrfi msthe remarksofRemer et al. accuracy of the aerosols properties retrieved from satellite [9] that the MODIS monthly mean AOT over the ocean observations. Some of these uncertainties are the particle is not significantly cloud contaminated (based on the good size, refractive index, the single scattering albedo, the particle agreement between the MODIS-NASA ocean algorithm and sphericity [30], the satellite sensors’ calibration, the cloud 4 AERONET sites’ AOT). signal (thin clouds or adjacent cloud to the aerosols) inter- For the 3-year dataset, the average cloud amount is higher ference, the asymmetry factor, and so forth. In this section in the south (58.5% and 60.2% at Hedo and Miyako, resp.) we will review the contribution of these uncertainties on the than the north (44.2% at Fukue). And, coincidently, the cloud accuracy of the satellite aerosols retrievals. Initial evaluations contamination and the negative bias (Table 1)are loweratthe of Terra- and Aqua-MODIS ocean aerosols products, made latter area compared to the former areas. by Kaufman and Tanre[ ´ 30], suggest that the expected The AE and AOT larger biases in the MODIS-GLI than uncertainty for MODIS small AOTs ranges from 0.05 to 0.1 the MODIS-NASA, in relation to the cloud contamination, and increases to 20–30% for high AOTs. In Remer et al. [9], can be explained by the pixel size selection scale differences the uncertainty is given as±0.03±0.05 AOT. between the two algorithms. The MODIS-GLI algorithm retrieves aerosols properties on a pixel-by-pixel basis while the MODIS-NASA algorithm retrievals are made on 10 4.2.1. Cloud Contamination Eeff ct. Atmosphere areas where aerosols properties are retrieved are supposed to be cloud- × 10-pixel boxes (only pixels with the highest noncloudy free. Examination of Terra- and Aqua-MODIS cloud masks probability are selected and averaged). This selection process suggest that, even aer ft rigorous cloud screening, the resulting unfortunately reduces the number of data to be analyzed clear-sky dataset could still be contaminated (sometimes by but produces less cloud-contaminated products compared to the pixel-by-pixel retrieval approach adopted by the MODIS- 15% or more) by small clouds [32, 33]. Also, in cloud-free areas, adjacent cloud pixels to clear-sky areas can increase GLI algorithm. Kaufman and Tanre[ ´ 30]suggest that when (brightening effect) or decrease (shadowing effect) the inten- theaerosol productisnot givenonapixelbypixel basis 2 2 sity of the aerosol radiative signal. The brightening eeff ct is (0.5 × 0.5 km ) but rather over a grid of 10 × 10 km or expected to play a more prominent role in these aerosols 50× 50 km , clear sky areas used for the determination of areas. To discuss the possible eeff cts of cloud contamination aerosols can lead to a cloud contamination of aerosols and on the satellite retrievals, we will examine the variation of therefore contribute to measurement errors of these aerosols an independent variable, the average cloud fraction (CF) properties. Decreasing the spatial resolution for the aerosol in each spatial subarea (the CF is from the MOD/MYD06 product could help to reduce cloud contamination but also version 5.1 MODIS cloud product) where the MODIS-NASA enhance the signal-to-noise-ratio (SNR) for the AE retrieval algorithm retrieval was performed, against the change in AE [30, 34]. andAOT (bothatthe ground andsatellite levels)thenthe bias of the satellite retrievals, that is, the difference (Ground- 4.2.2. Particle Size and Shape Distribution Assumption. In MODIS-GLI or Dif GLI and Ground-MODIS-NASA or Dif order to convert the aerosol mass measurements to size distri- NASA). Figure 6 shows the results of these relationships. On butions, a model of the aerosol size distribution as a function this figure, the CF data are sorted in 0.1 ranges (i.e., 10% of aerosol mass is needed [35]. If “actual” aerosol properties CF). The maximum range is >0.7 (i.e., >70% CF). Let us are too different from those assumed by the model, then examine rfi st the possibility of cloud contamination on the the satellite retrieved products will be less accurate [36]. For AE.TheAEgroundvariation showsthatitisnearlysteady the 3-year study period and at the 3 ground sites examined the distribution is consistently bimodal: small particle mode (no specific trend) along the CF variation ranges, while both the MODIS-NASA and MODIS-GLI AE show a decreasing (accumulation) and large particle mode (coarse) radii. It was trend (larger sizes) with the CF increase. Consequently, the noticed that the small mode radius tends to move to higher AE difference ground-satellite (Dif GLI and Dif NASA) tends values with the increase of the particle volume. This creates to increase with the CF increase. A stronger AE decreasing a larger dispersion (as will be shown later with the standard trend at high CF is observable with the MODIS-GLI than deviation) of the mode. The average radii of the ne fi mode ± the MODIS-NASA at all sites. eTh se trends show that both standard deviation (STD) are0.15±1.75𝜇 m,0.16±1.63𝜇 m, satellites’ algorithm retrieved AE appear to be contaminated and0.16±1.72𝜇 mandthecoarsemodes4.18± 2.35𝜇 m,4.92± by clouds. 2.27𝜇 m, and4.72±2.21𝜇 mfor Fukue, Hedo,and Miyako, The AOT as the AE shows no specific trend in the respectively.Theaccumulationmoderadiusisnearlysimilar ground data variation with the CF. In contrast, the MODIS- at the3sites(0.15–0.16),and thecoarsemodeislargerinthe GLI AOT gradually increases with the CF increase while south than the north. The STDs within each mode range are the increase of the latter with the MODIS-NASA AOT also very comparable. eTh accumulation and coarse modes is slighter. Consequently, the difference in absolute value radiiusedinthe MODIS-GLIalgorithm are, respectively, between the ground and satellite AOT (Dif GLI for the 0.17±1.3,3.44±2.37. es Th e values are, respectively, higher difference with MODIS-GLI or Dif NASA for the difference and lower than those of the ground sites, while their STDs 10 Advances in Meteorology Fukue (CF versus AE) Fukue (CF versus AOT) 0.5 1.4 0.8 0.5 1.2 0.7 0.25 0.25 0.6 0.5 0.8 0 0.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.6 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.3 0.4 −0.25 −0.25 0.2 0.2 0.1 −0.5 −0.5 0 CF ranges CF ranges Dif GLI AE GLI AOT GLI Dif GLI Dif NASA AE NASA Dif NASA AOT NASA AE ground AOT ground Hedo (CF versus AE) Hedo (CF versus AOT) 0.5 1.4 0.5 0.8 1.2 0.7 0.25 0.6 0.25 0.5 0.8 0 0.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.6 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.3 0.4 −0.25 −0.25 0.2 0.2 0.1 −0.5 0 −0.5 0 CF ranges CF ranges AE GLI AOT GLI Dif GLI Dif GLI AE NASA AOT NASA Dif NASA Dif NASA Ground AE ground Miyako (CF versus AOT) Miyako (CF versus AE) 0.5 0.5 0.8 1.4 0.7 1.2 0.25 0.6 0.25 0.5 0.8 0 0.4 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.1 0.2 0.3 0.4 0.5 0.6 0.7 >0.7 0.6 0.3 0.4 −0.25 −0.25 0.2 0.2 0.1 −0.5 −0.5 0 CF ranges CF ranges AE GLI AOT GLI Dif GLI Dif GLI AE NASA AOT NASA Dif NASA Dif NASA AE ground AOT ground (a) (b) Figure 6: Cloud fraction (CF) variation against aerosols properties retrievals from ground (AE and AOT ground), MODIS-GLI (AE and AOT GLI), and MODIS-NASA (AE and AOT NASA). en Th CF variation against satellite bias, that is, difference ground-satellite (ground AE or AOT-MODIS-GLI AE or AOT), (ground AE or AOT-MODIS-NASA AE or AOT), during the 3 years of observations. The AE or AOT scales are on the right while the scale of the Dif is on the left. Dif AE (ground − satellite) Dif AE (ground − satellite) Dif AE (ground − satellite) AE AE AE Dif AOT (ground − satellite) Dif AOT (ground − satellite) Dif AOT (ground − satellite) AOT AOT AOT Advances in Meteorology 11 Fukue (V /V ) S L Hedo (V /V ) S L 1.2 1.2 0.8 0.8 0.4 0.4 −0.4 −0.4 Accumulation mode volume ratio Accumulation mode volume ratio AE Dif GLI Dif GLI AE ground AE ground AE Dif NASA Dif NASA AE GLI AE GLI AE NASA AE NASA (a) (b) Miyako (V /V ) S L 1.2 0.8 0.4 −0.4 Accumulation mode volume ratio AE ground Dif GLI AE GLI Dif NASA AE NASA (c) Figure 7: AE and AE difference (ground-satellite) against the accumulation mode volume ratio ( 𝑉 /𝑉 )variation. 𝑆 𝐿 are lower and close to those of the ground observations. study. However no relationship could be established between The MODIS-NASA algorithm uses a combination of models the increase of ground-satellite discrepancies with either the whosemedianradius ± STD varies from 0.035 ± 0.40 to accumulation or the coarse mode radii. 0.10±0.60 for the accumulation mode and0.40 ± 0.60 to Figure 7 shows the histograms of the AE and AE differ- 1.00 ± 0.80 for the coarse mode, in the wavelengths ranges ence (ground-satellite) against the ratio accumulation (small) 0.47–0.86𝜇 m. These particle size mode values and STDs are mode volume against large mode volume (𝑉 /𝑉 ). The AE 𝑆 𝐿 lower than those of the corresponding ground data. Single difference tends to increase with the 𝑉 /𝑉 increase. As 𝑆 𝐿 process originating particles have smaller STDs, while mul- noticed before both satellites’ algorithms use smaller STDs tiple processes originating particles have wider log-normal at the accumulation mode than the ground observed STDs, distributions [37]. The smaller accumulation mode STDs of though the mode radius itself is not very dieff rent from that both satellites’ algorithms compared to the ground data may of the ground. eTh STD of the accumulation mode may imply that the aerosols models retrievals do not take into increase the uncertainty in the satellites retrieved AE by both account most of the processes at the origin of these particles. algorithms. As a consequence, the particle size of the satellite retrieved The particle shape (nonsphericity) can also be a potential aerosols may dieff r from that of the ground. In some areas contributor to the inaccuracy of satellite aerosols retrievals. like semiarid areas, the effective radius of the accumulation According to Levy et al. [39] the MODIS-NASA ocean mode particle may increase with increasing AOT [38]. This algorithm does not perform well in dust-laden atmosphere, tendency was also noticed in the ground data of the present and this problem is attributed to poor assumptions of the AE and Dif <0.1 0.15 0.2 AE and Dif 0.25 0.3 <0.1 0.35 0.15 >0.35 0.2 AE and Dif 0.25 0.3 <0.1 0.35 0.15 >0.35 0.2 0.25 0.3 0.35 >0.35 12 Advances in Meteorology nonsphericity of the particles in the dust-aerosol phase tendstodecreasebeyond1.5 (closertothe values used by the functions. The peak in dust (where particles are less spherical) MODIS-GLI algorithm). All these variations show that the amount occurring in the aerosol load of the study area during RIr is an important contributor to the enhancement of the spring does not particularly increase the difference between satellite AOT uncertainty. No trend was seen between the RIi the ground AOT or AE with the satellite derived data (as and the AOT RMSE nor between the RIr or RIi and the AE shown by Figure 5). Also the 1 : 1 good alignment of the AOT RMSE variation. of the MODIS-NASA algorithm with ground data does not 4.2.4. Impact of Scattering Parameters. The main property show a specific seasonal dependency. All of these may lead representing the scattering by aerosols is generally expressed to the conclusion that there is a less significant impact of the by the single scattering albedo (SSA). The scattering prefer- sphericity in the satellites’ algorithms examined compared ential direction may be either forward (+1 for pure forward to other algorithms. Furthermore, tests conducted with a scattering) or backward (–1 for pure backscattering). The nonspherical model, by Nakajima and Higurashi [23], on proportion of light scattered in the forward direction is the MODIS-GLI algorithm suggest that the nonspherical defined as the asymmetry factor (AF). For satellites’ aerosol scattering eect ff is limited in this algorithm. retrievals, the SSA and AF are precomputed using a Mie- based theory (spherical particles) scattering model. Irregu- 4.2.3. Impact of the Refractive Index. In a given location, larly shaped tropospheric particles may have larger SSA and atmospheric aerosols are characterized by their concentra- smaller AF than their equal volume spherical counterparts tion, their size distribution, their shape, their vertical profile, [40]. The average SSA ± STD and AF± STD observed by and by their chemical composition [2]. The latter is indicative the SKYNET ground sites are, respectively,0.95±0.06 and of the magnitude of the real and imaginary parts of the 0.69±0.04 for Fukue,0.98±0.05 and0.70±0.04 for Hedo, refractive index (RI) of aerosols. eTh real part of the RI and0.96±0.07 and0.71±0.05 for Miyako. Figure 9 shows (RIr) is spectrally related to the scattering of the radiation the variation of the AF against the AE and the AE Dif. eTh while the imaginary part (RIi) is a consequence of the ground AE consistently decreases with the increase of the AF absorption of radiation, with a null value meaning the particle at the three ground sites; that is, the increase in the forward does not absorb radiation. In the satellite aerosols retrieval scattering of light results from the increase of the particle size. scheme, the aerosol model assumes specific realistic values No such clear trend is observed between the ground AF and for the real and imaginary parts of the RI. In the wavelength both, the satellites’ AE of the GLI and NASA algorithms. As range 0.47–0.86𝜇 m, the MODIS-GLI algorithm uses aerosol a result, the ground-satellite AE discrepancies decrease as the models with a single RI (1.5–i0.005), based on field data AF increases that is, as the scattering of light in the forward analyses. eTh real part RI of this algorithm is close to general direction increases. eTh uncertainty in the determination of tropospheric aerosols indices [40]. Meanwhile, the MODIS- the AF appears then to impact both satellites’ algorithms AE NASA algorithm uses a combination of models where the retrievals. eTh ground AF slope change with the AF is not RIs are set to vary according to the area and the volume size as steep in the northern Fukue as in the southern Hedo and distribution mode ratio. The values used by this algorithm are Miyako. This is consistent with the increase in coarse particles derived from the sun/sky ground photometers of AERONET from the north to the south (increased forward radiation [9], 1.45–i0.0035 or 1.40–i0.0020 forthe nfi emodeand 1.35– scattering) and consequently the AE RMSE. No consistent i0.001 for the coarse mode, of all aerosols types except dust- correlationwasfoundbetweentheSSAandeither,theground like types where the RIs are 1.53–i0.003 and 1.53–i0.000 AE, the AOT or the error in the determination of the satellite for 0.47𝜇 mand 0.86𝜇 m wavelengths, respectively [41]. The retrieved AOT and AE. This could be related to the fact that average RIs at 0.5𝜇 m wavelength obtained from ground data the influence of error in the retrieved SSA is important only for the whole study period are 1.44–i0.0043, 1.43–i0.0016, for thick layers [36]. And the data used here are mostly thin 1.44–i0.0029 for Fukue, Hedo, and Miyako sites, respectively. aerosols layers. The MODIS-NASA real and imaginary parts are closer or within the ranges of the ground RI values than the MODIS- 4.2.5. Impact of Solar and Satellite Zenith Angles. An ideal GLI are to the latter. The following distribution of RI at the aerosol algorithm would retrieve AOTs of equal quality, ground siteswas observed (inbracketsare thepercentage independent of solar and observing geometry [42]. No of data at Fukue, Hedo,and Miyako,respectively, forthe consistent trend showing possible or signicfi ant satellite bias RI range considered): the real part RI: <1.40 (33%, 41%, was observed between both the AOT and AE (AOT and 36%), 1.4–1.5 (37%, 29%, 32%),>1.5 (30%, 30%, 32%); and AE RMSE) and the solar zenith angle or the satellite zenith the imaginary part RI:<0.002 (54%, 84%, 75%), 0.002–0.005 angle. Using collection 5 MODIS dark target data over land, (22%, 7%, 10%),>0.005 (24%, 9%, 15%). Only one-third of Levy et al. [42] suggest that there is only a slight over- and the data are beyond RIr 1.5, and less than 25% have RIi above underestimate of the AOT by∼0.01, respectively, on the sun- 0.005 (RI values considered for the MODIS-GLI). glint side and the sun-shadow side of the MODIS swath. eTh comparison with the AOT shows that ( Figure 8)the RIr tends to increase with the AOT increase, for the ground 5. Implications for Future and both satellites’ algorithms data. eTh AOT RMSE is high Satellites’ Algorithms at high RIr and vice versa for low RIr values such as 1.4 (closer to the values used in the MODIS-NASA algorithm). For the The RMSE and Dif for the AOT retrieved from the MODIS-GLI, the AOT RMSE is high in the 1.4–1.45 range and MODIS-GLI algorithm are higher compared to those of the Advances in Meteorology 13 Hedo (RIr 0.5 𝜇 m versus AOT) Fukue (RIr 0.5 𝜇 m versus AOT) 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 RIr 0.5 𝜇 m RIr 0.5 𝜇 m AOT ground AOT NASA AOT ground AOT NASA AOT GLI AOT RMSE AOT GLI AOT RMSE GLI GLI (a) (b) Miyako (RIr 0.5 𝜇 m versus AOT) 0.8 0.6 0.4 0.2 RIr 0.5 𝜇 m AOT ground AOT NASA AOT RMSE AOT GLI GLI (c) Figure 8: Real part of the refractive index (RIr) variation at the ground against ground AOT, and the satellite retrieved AOT and AOT RMSE. MODIS-NASA algorithm. The accuracies of both algorithms of output data. Changes in aerosols mixtures of the region for the AE are much closer to each other as well as lower would need a better representation. For example, this study than the AOT accuracies. Some of the factors explaining these showed that the aerosol particle radii deviations from the accuracy differences range from the cloud contamination, the mode should be wider in order to integrate the dynamic particle size and shape distribution, the refractive index, and processes at the origin of the aerosols of this region. Also, a the scattering parameters to the sun-satellite geometry. The wider range (than that already used by the MODIS-NASA lessons drawn from the possible impact of these parameters algorithm) for the refractive indices would be necessary on satelliteretrievalswillhelptoimprove theaccuracyof to reduce the uncertainty in the AE. eTh high spatial and future satellites’ aerosols products such as those from the spectralresolutionoftheGCOM-C/SGLIsatellite(e.g.,250 m coming GCOM-C/SGLI satellite. at all visible and thermal infrared channels) should help to better discriminate clouds from cloud-free areas than actual eTh cloud contamination was found to be stronger in the AOT of MODIS-GLI than that of MODIS-NASA. This satellites do and therefore reduce possible cloud contamina- was explained by the differences in the spatial resolution of tion of the aerosols products. The improvement in the AOT accuracy will certainly increase that of the AE. eTh increase the retrievals: pixel-by-pixel retrieval in the MODIS-GLI and pixels’ selection retrieval from boxes of 10× 10 pixels in the in the number of ground sites measurements of aerosols such as in the SKYNET, AERONET, and other networks, provid- MODIS-NASA. To avoid this contamination in the future ing better quality data and quantitative characterization of and in case the pixel-by-pixel approach is chosen, a cloud fraction analysis should accompany the aerosols retrievals aerosols, would be of great asset for the improvement of the assumptions used in aerosols models: predetermined particle andbeusedtocorrect theaerosolsproperties’ inaccuracies. The MODIS-NASA approach looks attractive as retrievals are size distribution (modes’ radii and volumes), refractive index, made only on the high probability clear pixels of a wider asymmetry factor, single scattering albedo, ambient humidity andparticlegrowth, andsoforth. pool of pixels, but the inconvenience is the lower amount AOT and RMSE 1.37 1.4 1.43 AOT and RMSE 1.46 1.49 1.37 1.52 1.4 >1.52 1.43 AOT and RMSE 1.46 1.49 1.37 1.52 1.4 >1.52 1.43 1.46 1.49 1.52 >1.52 14 Advances in Meteorology Fukue (AF versus AE) Hedo (AF versus AE) 1.2 1.2 0.8 0.8 0.4 0.4 −0.4 −0.4 AF AF AE Dif GLI AE Dif GLI AE ground AE ground AE Dif NASA AE Dif NASA AE GLI AE GLI AE NASA AE NASA (a) (b) Miyako (AF versus AE) 1.2 0.8 0.4 −0.4 AF AE Dif GLI AE ground AE Dif NASA AE GLI AE NASA (c) Figure 9: Aerosol asymmetry factor (AF) histogram distribution against AE and the satellites’ algorithms AE, and the difference ground- satellite. 6. Conclusion ground and MODIS-NASA AOT and AE underestimation by both algorithms. eTh large particle AE overestimation The quality of aerosols retrievals from future satellites was overwhelmed by the small particle underestimation. eTh was discussed in this study, through a validation study of seasonal variation of the aerosols parameters (AOT and AE) aerosols products from actual satellites (Terra- and Aqua- based on monthly averages showed similar trends/variations MODIS) having compatible channels with those of the though the direct values differed, sometimes with MODIS- coming GCOM-C/SGLI. Aerosols products derived from two NASA or more oen ft with MODIS-GLI. Also nearly constant algorithms (MODIS-GLI and MODIS-NASA) using these monthly AOT differences between the ground aerosols data satellites were evaluated against ground-truth data from 3 and the satellite retrievals were observed. The AE showed SKYNET sites located in the Pacific East Asian region. eTh closer values between the ground and satellite data during results obtained showed a systematic overestimation of the winter and spring for the Fukue site and summer and autumn AOTbythe MODIS-GLIand abetteragreement betweenthe for the southern sites (Hedo and Miyako). eTh eeff ct of cloud AE and Dif 0.63 0.65 0.67 0.69 AE and Dif 0.71 0.73 0.63 0.75 0.65 >0.75 0.67 0.69 AE and Dif 0.71 0.73 0.63 0.75 0.65 >0.75 0.67 0.69 0.71 0.73 0.75 >0.75 Advances in Meteorology 15 contamination on the satellites retrievals, examined through retrieval algorithm over Central Europe,” Atmospheric Measure- ment Techniques,vol.3,no. 5, pp.1255–1270,2010. the use of an independent variable, the cloud fraction, showed that cloud contamination was more pronounced on [8] P. Goloub, D. Tanre´, J. L. Deuze, M. Herman, A. Marchand, and F. Breon, “Validation of the first algorithm applied for deriving theAOT MODIS-GLIthanthatofMODIS-NASA. eTh AE the aerosol properties over the ocean using the POLDER/ analysis showed that this parameter decreased (particles get ADEOS measurements,” IEEE Transactions on Geoscience and larger) with the CF for both algorithms. The difference in Remote Sensing,vol.37, no.3,pp. 1586–1596, 1999. the AOT contamination between the two satellites’ algo- [9] L.A.Remer,Y.J.Kaufman,D.Tanre´ et al., “eTh MODIS aerosol rithms was found to be mostly due to the retrieval pixel algorithm, products, and validation,” Journal of the Atmospheric resolution difference: pixel-by-pixel for the MODIS-GLI, Sciences,vol.62, no.4,pp. 947–973, 2005. therefore higher probability of contamination compared to [10] A. K. Heidinger, V. R. Anne, and C. Dean, “Using MODIS the pixels’ selection from larger boxes for the MODIS-NASA. to estimate cloud contamination of the AVHRR data record,” Also, this difference could be explained by the fact that the JournalofAtmospheric andOceanic Technology,vol.19, no.5, MODIS-NASA uses wider spectral information to estimate pp.586–601,2002. the aerosol model and, may perform more accurate retrieval [11] M. Jin and R. E. Treadon, “Correcting the orbit drift eeff ct on of spectral AOT [36]. The AE distribution showed that when AVHRR land surface skin temperature measurements,” Inter- thegrounddatahaveabimodaldistribution, theuncertainty national Journal of Remote Sensing,vol.24, no.22, pp.4543– in the AE retrievals by the two satellites’ algorithms is 4558, 2003. higher.Theshape of theparticlesizedistributionexamined [12] J. R. Dim, H. Murakami, T. Y. Nakajima, B. Nordell, A. K. showed larger STDs than those considered by the satellites’ Heidinger, and T. Takamura, “eTh recent state of the climate: algorithms, leading to limited mixtures in the aerosols sizes driving components of cloud-type variability,” JournalofGeo- retrieved by the satellites’ algorithms. Another contributor physical Research D, vol. 116, no. 11, Article ID D11117, 2011. to the ground-satellite differences was the uncertainty on [13] T. Takemura, T. Nakajima, A. Higurashi, S. Ohta, and N. the satellite retrievals, due to the asymmetry factor (AF). Sugimoto, “Aerosol distributions and radiative forcing over the A negative correlation was found with both the ground AE Asian Pacific region simulated by Spectral Radiation-Trans- andthe AE dieff rence(ground-satellite)against theAF. eTh port Model for Aerosol Species (SPRINTARS),” Journal of increase of the light scattering in the forward direction is a Geophysical Research D,vol.108,no. 23,pp. 27–10, 2003. consequence of the increasing size of the aerosols particles. [14] B. C. Bush and F. P. J. Valero, “Surface aerosol radiative forcing at The real part of the refractive index was found to be correlated Gosan during the ACE-Asia campaign,” Journal of Geophysical with the AOT RMSE of both satellites’ algorithms and not Research, vol. 108, no. D23, p. 8660, 2003. with the AE RMSE. To end this study, some measures aimed [15] B.J.Huebert,T.Bates,P.B.Russell et al., “AnoverviewofACE at improving the accuracy of future satellites’ algorithms were Asia: strategies for quantifying the relationships between Asian discussed. These involve remedies to cloud contamination aerosols and their climatic impacts,” Journal of Geophysical and corrections of the particle size distribution and scattering Research, vol. 108, no. D23, p. 8633, 2003. models. [16] I. Uno, “Regional chemical weather forecasting system CFORS: model descriptions and analysis of surface observations at Japanese island stations during the ACE-ASIA experiment,” References Journal of Geophysical Research: Atmospheres, vol. 108, no. D23, [1] K. N. Liou, An Introduction to Atmospheric Radiation, Academic [17] T. Nakajima, S. Yoon, V. Ramanathan et al., “Overview of the Press, 2nd edition, 2002. Atmospheric Brown Cloud East Asian Regional Experiment [2] G.A.d’Almeida,P.Koepke, andE.P.Shettle, Atmospheric Aero- 2005 and a study of the aerosol direct radiative forcing in east sols, Global Climatology and Radiative Characteristics,Deepak, Asia,” Journal of Geophysical Research D,vol.112,no. 24,Article Hampton, Va, USA, 1991. ID D24S91, 2007. [3]L.A.Remer,D.Tanre´, Y. J. Kaufman et al., “Validation of [18] T. Takamura, N. Sugimoto, A. Shimizu et al., “Aerosol radiative MODIS aerosol retrieval over ocean,” Geophysical Research characteristics at Gosan, Korea, during the Atmospheric Brown Letters, vol. 29, no. 12, pp. MOD3-1–MOD3-4, 2002. Cloud East Asian Regional Experiment 2005,” Journal of Geo- [4] D.A.Chu,Y.J.Kaufman,C.Ichoku, L. A. Remer, D. Tanre, ´ physical Research D: Atmospheres,vol.112,no. 22,Article ID and B. N. Holben, “Validation of MODIS aerosol optical depth D22S36, 2007. retrieval over land,” Geophysical Research Letters,vol.29, no.12, [19] I. Sano,S.Mukai,M.Yamano, T. Takamura,T.Nakajima, and pp. MOD2-1–MOD2-4, 2002. B. Holben, “Calibration and validation of retrieved aerosol [5] C.Ichoku, D. AllenChu,S.Mattooetal.,“Aspatio-temporal properties based on aeronet and skynet,” Advances in Space approach for global validation and analysis of MODIS aerosol Research, vol. 32, no. 11, pp. 2159–2164, 2003. products,” Geophysical Research Letters,vol.29, no.12, pp.1–4, [20] T. Nakajima, G. Tonna, R. Rao, P. Boi, Y. Kaufman, and B. Holben, “Use of sky brightness measurements from ground for [6] A.Ignatov,L.Stowe,R.Singh,S.Sakerin, D. Kabanov, and remote sensing of particulate polydispersions,” Applied Optics, I. Dergileva, “Validation of NOAA/AVHRR aerosol retrievals vol. 35, no. 15, pp. 2672–2686, 1996. using sun-photometer measurements from R/V akademik ver- nadsky in 1991,” Advances in Space Research,vol.16, no.10, pp. [21] Y. Liu, J. Huang, G. Shi et al., “Aerosol optical properties 95–98, 1995. and radiative effect determined from sky-radiometer over [7] M.Riffler, C. Popp,A.Hauser, F. Fontana, andS.Wun- Loess Plateau of Northwest China,” Atmospheric Chemistry and derle, “Validation of a modified AVHRR aerosol optical depth Physics,vol.11, no.22, pp.11455–11463,2011. 16 Advances in Meteorology [22] J. Bi, Y. Liu, J. Huang et al., “Characteristics of dust aerosols the Royal Meteorological Society,vol.129,no. 590, pp.1395–1413, derived from sky-radiometer over Loess Plateau of Northwest 2003. China,” in Proceedings of the16thCEReS InternationalSympo- [38] L. A. Remer, D. Tanre´, andY.J.Kaufman,“Algorithmfor remote sium on Climate Change Studies through Activities of SKYNET sensing of tropospheric aerosols from MODIS,” collection 5,87 and Virtual Laboratory for Climate Diagnostics,Nago, Okinawa, pp, 1998. Japan, October 2010. [39] R. C. Levy,L.A.Remer,D.Tanre´ et al., “Evaluation of the [23] T. Nakajima and A. Higurashi, “A use of two-channel radi- MODIS retrievals of dust aerosol over the ocean during PRIDE,” ances for an aerosol characterization from space,” Geophysical J. Geophys. Res,vol.108,no. D19, 2003. Research Letters, vol. 25, no. 20, pp. 3815–3818, 1998. [40] J. B. Pollack and J. N. Cuzzi, “Scattering by nonspherical [24] A. Higurashi and T. Nakajima, “Detection of aerosol types over particles of size comparable to a wavelength: a new semi- theEastChinaSea near Japanfromfour-channelsatellite data,” empirical theory and its application to tropospheric aerosols,” Geophysical Research Letters,vol.29, no.17, pp.17–1, 2002. Journal of the Atmospheric Sciences,vol.37, pp.868–881,1980. [25] C. Ichoku,Y.J.Kaufman,L.A.Remer,and R. Levy,“Global [41] R. C. Levy,L.A.Remer,D.Tanre, ´ S. Mattoo, and Y. J. Kaufman, aerosol remote sensing from MODIS,” Advances in Space Re- “Algorithm for remote sensing of tropospheric aerosols over search,vol.34, no.4,pp. 820–827, 2004. dark targets from MODIS,” Collections005 and051:Revision2, Feb 2009, Product ID: MOD04/MYD04,96pp, 2009. [26] W. A. Abdou, D. J. Diner, J. V. Martonchik et al., “Comparison of coincident Multiangle Imaging Spectroradiometer and Mod- [42] R. C. Levy,L.A.Remer,R.G.Kleidman et al., “Global erate Resolution Imaging Spectroradiometer aerosol optical evaluation of the Collection 5 MODIS dark-target aerosol depths over land and ocean scenes containing Aerosol Robotic products over land,” Atmospheric Chemistry and Physics,vol.10, Network sites,” Journal of Geophysical Research D,vol.110,no. no.21, pp.10399–10420,2010. 10,pp. 1–12,2005. [27] O. V. Kalashnikova and R. A. Kahn, “Mineral dust plume evolution over the Atlantic from MISR and MODIS aerosol retrievals,” JournalofGeophysical Research D, vol. 113, no. 24, Article ID D24204, 2008. [28] A. Higurashi and T. Nakajima, “Development of a two-channel aerosol retrieval algorithm on a global scale using NOAA AVHRR,” Journal of the Atmospheric Sciences,vol.56, no.7,pp. 924–941, 1999. [29] A. Higurashi, T. Nakajima, B. N. Holben, A. Smirnov, R. Frouin, and B. Chatenet, “A study of global aerosol optical climatology with two-channel AVHRR remote sensing,” Journal of Climate, vol. 13,no. 12,pp. 2011–2027, 2000. [30] Y. J. Kaufman and D. Tanre, ´ “Algorithm for remote sensing of tropospheric aerosol from MODIS. Products: MOD04, MOD08. ATBD reference number: ATBD-MOD-02,” 1998, http://modis.gsfc.nasa.gov/data/atbd/atbd mod02.pdf. [31] F. Wagner and A. M. Silva, “Some considerations about Angstrom ¨ exponent distributions,” Atmospheric Chemistry and Physics,vol.8,no. 3, pp.481–489,2008. [32] T. W. Ault,K.P.Czajkowski, T. Benkoetal.,“Validation of the MODIS snow product and cloud mask using student and NWS cooperative station observations in the Lower Great Lakes Region,” Remote Sensing of Environment,vol.105,no. 4, pp.341– 353, 2006. [33] S. A. Ackerman,R.E.Holz, R. Frey,E.W.Eloranta, B. C. Maddux, and M. McGill, “Cloud detection with MODIS. Part II: validation,” JournalofAtmospheric andOceanic Technology, vol. 25,no. 7, pp.1073–1086,2008. [34] R. Santer and D. Ramon, MERIS Aerosol Remote Sensing over Land ATBD, 2011. [35] J. N. Porter and A. D. Clarke, “Aerosol size distribution models basedoninsitumeasurements,” Journal of Geophysical Research D,vol.102,no. 5, pp.6035–6045,1997. [36] A. A. Kokhanovsky, J. L. Deuze, ´ D. J. Diner et al., “The inter- comparison of major satellite aerosol retrieval algorithms using simulated intensity and polarization characteristics of reflected light,” Atmospheric Measurement Techniques,vol.3,pp. 909– 932, 2010. [37] L. Alados-Arboledas, H. Lyamani, and F. J. Olmo, “Aerosol size properties at Armilla, Granada (Spain),” Quarterly Journal of International Journal of Journal of Ecology Mining The Scientific Journal of Geochemistry Scientifica World Journal Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Earthquakes Paleontology Journal Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Journal of Petroleum Engineering Submit your manuscripts at http://www.hindawi.com International Journal of Geophysics Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Advances in Advances in Journal of Advances in International Journal of Meteorology ClimatologyGeology Oceanography Oceanography Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 Applied & Journal of Environmental International Journal of Journal of International Journal of Computational Soil Science Mineralogy Geological Research Atmospheric Sciences Environmental Sciences Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Hindawi Publishing Corporation Volume 2014 Hindawi Publishing Corporation http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com http://www.hindawi.com Volume 2014

Journal

Advances in MeteorologyHindawi Publishing Corporation

Published: Aug 14, 2013

References