Exploring Global Land Coarse-Mode Aerosol Changes from 2001-2021 Using a New Spatiotemporal Coaction Deep-Learning Model

Coarse-mode aerosol optical depths (cAODs) are critical for understanding the impact of coarse particle sizes, especially dust aerosols, on climate. Currently, the limited data length and high uncertainty of satellite products diminish the applicability of cAOD for climate research. Here, we propose a spatiotemporal coaction deep-learning model (SCAM) for the retrieval of global land cAOD (500 nm) from 2001-2021. In contrast to conventional deep-learning models, the SCAM considers the impacts of spatiotemporal feature interactions and can simultaneously describe linear and nonlinear relationships for retrievals. Based on these unique characteristics, the SCAM considerably improved global daily cAOD accuracies and coverages (R = 0.82, root-mean-square error [RMSE] = 0.04). Compared to official products from the multiangle imaging spectroradiometer (MISR), the moderate resolution imaging spectroradiometer (MODIS), and the polarization and directionality of Earth's reflectances (POLDER) instrument, as well as the physical-deep learning (Phy-DL) derived cAOD, the SCAM cAOD improved the monthly R from 0.44 to 0.88 and more accurately captured over the desert regions. Based on the SCAM cAOD, daily dust cases decreased over the Sahara, Thar Desert, Gobi Desert, and Middle East during 2001-2021 (>3 × 10-3/year). The SCAM-retrieved cAOD can contribute considerably to resolving the climate change uncertainty related to coarse-mode aerosols. Our proposed method is highly valuable for reducing uncertainties regarding coarse aerosols and climate interactions.

Retrieval uncertainty and consistency of Suomi-NPP VIIRS deep blue and dark target aerosol products under diverse aerosol loading scenarios over South Asia

Retrieval accuracy and stability of two operational aerosol retrieval algorithms, Deep Blue (DB) and Dark Target (DT), applied on Visible Infrared Imaging Radiometer Suite (VIIRS) on-board Suomi National Polar-orbiting Partnership (S-NPP) satellite were evaluated over South Asia. The region is reported to be highly challenging to accurate estimation of satellite-based aerosol optical properties due to variations in surface reflectance, complex aerosol system and regional meteorology. Performance of both algorithms were initially evaluated by comparing their ability to retrieve aerosol signal over the complex geographical region under specific air pollution emission scenario. Thereafter, retrieval accuracy was investigated against 10 AERONET sites across South Asia, selected based on their geography and predominance aerosol types, from year 2012-2021. Geo-spatial analysis indicates DB to efficiently retrieve fine aerosol features over bright arid surfaces, and for smoke/dust dominating events whereas DT was better to identify small fire events under dark vegetated surface. Both algorithms however, indicate unsatisfactory retrieval accuracy against AERONET having 56-59% of valid retrievals with high RMSE (0.30-0.33) and bias. Overall, DB slightly underpredicted AOD with -0.02 mean bias (MB) whereas DT overpredicted AOD (MB: 0.13), with seasonality in their retrieval efficiency against AERONET. Time-series analysis indicates stability in retrieving AOD and match-up number for both algorithms. Retrieval bias of DB and DT AOD against AERONET AOD under diverse aerosol loading, aerosol size, scattering/absorbing aerosol, and surface vegetation coverage scenarios revealed DT to be more influenced by these conditions. Error analysis indicates at low AOD (≤0.2), accuracy of both DB and DT were subject to underlying vegetation coverage. At AOD>0.2, DB performed well in retrieving coarse aerosols whereas DT was superior when fine aerosols dominated. Overall, accuracy of both VIIRS algorithms require further refinement to continue MODIS AOD legacy over South Asia.

Retrieval of total and fine mode aerosol optical depth by an improved MODIS Dark Target algorithm

Total and fine mode aerosol optical depth (AODT and AODF), as well as the fine mode fraction (FMF = AODF/AODT), are critical variables for climate change and atmospheric environment studies. The retrievals with high accuracy from satellite observations, particularly FMF and AODF over land, remain challenging. This study aims to improve the Moderate-resolution Imaging Spectro-radiometer (MODIS) land dark target (DT) algorithm for retrieving AODT, AODF, and FMF on a global scale. Based on the fact that the underestimated surface reflectance (SR) could overestimate the AODT and underestimate the aerosol size parameter in the DT algorithm, two robust schemes were developed to improve SR determination: the first (NEW1 DT) used the top of the atmosphere reflectance instead of SR at 2.12 µm; the second (NEW2 DT) used eleven-year MODIS data to establish a monthly spectral SR relationship model (2.12-0.47 and 2.12-0.65 µm) database at pixel-by-pixel scale. Then a novel lookup table approach based on the physical process was proposed to retrieve the AODF and FMF. The new MODIS AODT, FMF, and AODF were compared to AERosol RObotic NETwork (AERONET) retrievals. Results showed that the root mean square error (RMSE) was 0.096-0.103, 0.098-0.099, and 0.167-0.180 for the new AODTs, AODFs, and FMFs, respectively, which were better than that of the Collection 6.1 (C6.1) DT (0.117, 0.235, and 0.426) in the validation by global AERONET sites. From the validation results, NEW2 DT provided better AODT and coarse mode AOD retrievals, while NEW1 DT had better AODF and FMF performances. The spatial patterns of AODF, FMF, and AODC of the new DT algorithms were comparable to those of the Polarization and Directionality of the Earth's Reflectances aerosol product. Hence, the new algorithms have the potential to provide global AODT, FMF, and AODF products over land to the scientific community with high accuracy using long-term MODIS data.

Long-term climatology and spatial trends of absorption, scattering, and total aerosol optical depths over East Africa during 2001-2019

The unprecedented increase in anthropogenic activities, coupled with the prevailing climatic conditions, has increased the aerosol load over East Africa (EA). Given this, the present study examined the trends in total, absorption, scattering, and total aerosol extinction optical depth (TAOD, AAOD, SAOD, and TAEOD) over EA, alongside trends in single scattering albedo (SSA). For this purpose, the AOD of different optical properties retrieved from multiple sensors and the Modern-Era Retrospective Analysis for Research and Applications (MERRA-2) model between January 2001 to December 2019 were utilized to estimate trends and assess their statistical significance. The spatial patterns of seasonal mean AOD from the Moderate-resolution Imaging Spectroradiometer (MODIS) sensor and MERRA-2 model were generally characterized with high (>0.35) and low (<0.2) AOD centers over EA observed during the local dry and wet seasons, respectively. Also, the spatial trend analysis revealed a general increase in TAOD, being positive and significant over the arid and semi-arid zones of the northeastern part of EA, which is majorly dominated by locally derived dust. The local dry (wet) months generally experienced positive (negative) trends in TAOD, associated with seasonal cycles of rainfall. High and significant positive trends in AAOD were dominated over the study domain, attributed to an increased amount of biomass burning, variations in soil moisture, and changes in the rainfall pattern. The trends in TAEOD showed a distinct pattern, except over some months that depicted significant increasing trends attributed to changes in climatic conditions and anthropogenic activities. At last, the study domain exhibited decreasing trends in SSA, signifying strong absorption of direct solar radiation resulting in a warming effect. The study revealed patterns of trends in aerosol optical properties and forms the basis for further research in aerosols over EA.

Accuracy, stability, and continuity of AVHRR, SeaWiFS, MODIS, and VIIRS deep blue long-term land aerosol retrieval in Asia

The deep blue (DB) aerosol algorithm applied to four satellite instruments, AVHRR, SeaWiFS, MODIS, and VIIRS, produced a long-term aerosol data set since 1989. This study first evaluated and compared the accuracy, stability, and continuity of four DB aerosol optical depth (AOD) products in Asia using AErosol RObotic NETwork measurements. Then, the regional AOD spatial distributions, coverages, and series trends are analyzed. The results show that VIIRS DB has the highest accuracy and stability, with an expected error (EE, ±(0.05 + 20%)) of 76.59% and stability of approximately 0.027 per decade. The performance of MODIS DB is slightly worse than that of VIIRS. However, their AOD pattern, coverage, and trend are comparable. The performance of AVHRR (EE = 58.10%) and the stability of SeaWiFS (0.093 per decade) are less good. Therefore, SeaWiFS DB data should be used with caution for trend analysis. The AOD accuracy and coverage together determine the AOD pattern and the continuity of multi-sensor data. In addition to consistent algorithm accuracy, it is necessary to consider the influences in sensor sampling and inappropriate-pixel screening schemes in the joint multi-sensor analysis. Encouragingly, although multiple DB products have different AOD averages of regional series, their changing trends are consistent. Error analysis shows that the AOD bias characteristic is different in different surface conditions. This indicates that the surface reflectance estimated by the DB algorithm using different techniques is divergent, which may be the direction for the improvement of the algorithm.

New global aerosol fine-mode fraction data over land derived from MODIS satellite retrievals

The space-borne measured fine-mode aerosol optical depth (fAOD) is a gross index of column-integrated anthropogenic particulate pollutants, especially over the populated land. The fAOD is the product of the AOD and the fine-mode fraction (FMF). While there exist numerous global AOD products derived from many different satellite sensors, there have been much fewer, if any, global FMF products with a quality good enough to understand their spatiotemporal variations. This is key to understanding the global distribution and spatiotemporal variations of air pollutants, as well as their impacts on global environmental and climate changes. Modifying our newly developed retrieval algorithm to the latest global-scale Moderate Resolution Imaging Spectroradiometer (MODIS) aerosol product (Collection 6.1), a global 10-year FMF product is generated and analyzed here. We first validate the product through comparisons with the FMF derived from Aerosol Robotic Network (AERONET) measurements. Among our 169,313 samples, the satellite-derived FMFs agreed with the AERONET spectral deconvolution algorithm (SDA)-retrieved FMFs with a root-mean-square error (RMSE) of 0.22. Analyzed using this new product are the global patterns and interannual and seasonal variations of the FMF over land. In general, the FMF is large (>0.80) over Mexico, Myanmar, Laos, southern China, and Africa and less than 0.5 in the Sahelian and Sudanian zones of northern Africa. Seasonally, higher FMF values occur in summer and autumn. The linear trend in the satellite-derived and AERONET FMFs for different countries was explored. The upward trend in the FMFs was particularly strong over Australia since 2008. This study provides a new global view of changes in FMFs using a new satellite product that could help improve our understanding of air pollution around the world.

Can We Detect the Brownness or Greenness of the Congo Rainforest Using Satellite-Derived Surface Albedo? A Study on the Role of Aerosol Uncertainties

The ability of spatial remote sensing in the visible domain to properly detect the slow transitions in the Earth’s vegetation is often a subject of debate. The reason behind this is that the satellite products often used to calculate vegetation indices such as surface albedo or reflectance, are not always correctly decontaminated from atmospheric effects. In view of the observed decline in vegetation over the Congo during the last decade, this study investigates how effectively satellite-derived variables can contribute to the answering of this question. In this study, we use two satellite-derived surface albedo products, three satellite-derived aerosol optical depth (AOD) products, two model-derived AOD products, and synthetic observations from radiative transfer simulations. The study discusses the important discrepancies (of up to 70%) found between these satellite surface albedo products in the visible domain over this region. We conclude therefore that the analysis of trends in vegetation properties based on satellite observations in the visible domain such as NDVI (normalized difference vegetation index), calculated from reflectance or albedo variables, is still quite questionable over tropical forest regions such as the Congo. Moreover, this study demonstrates that there is a significant increase (of up to 14%) in total aerosols within the last decade over the Congo. We note that if these changes in aerosol loads are not correctly taken into account in the retrieval of surface albedo, a greenness change of the surface properties (decrease of visible albedo) of around 8% could be artificially detected. Finally, the study also shows that neglecting strong aerosol emissions due to volcano eruptions could lead to an artificial increase of greenness over the Congo of more than 25% in the year of the eruptions and up to 16% during the 2–3 years that follow.

Discerning the pre-monsoon urban atmosphere aerosol characteristic and its potential source type remotely sensed by AERONET over the Bengal Gangetic plain

In the present study, we evaluated the pre-monsoon urban atmosphere (UA) aerosol characteristics remotely sensed by Aerosol Robotic Network (AERONET) over the Bengal Gangetic plain (BGP) at Kolkata (KOL) and their implication in potential source types and spatiotemporal features. About 70% of the AERONET-sensed aerosol optical depth at 0.50 μ m, AOD_0.5 (Angstrom exponent, α at 0.44-0.87 μ m) during the pre-monsoon period (February to June) was greater than 0.50 (≤ 1); the pre-monsoon mean of AOD_0.5 (α) was 0.73 (0.83) which was found being slightly higher (lower) than nearby AERONET stations (Dhaka/Bhola) located over the eastern Ganges basin. The volume geometric mean radius for the fine mode (FM) (coarse mode, CM) UA aerosol from AERONET retrievals was estimated to be 0.14-0.17 (2.24-2.75) μ m. The spectral distribution of the monthly mean of UA aerosol single-scattering albedo (SSA) exhibited an increasing trend with an increase in wavelength throughout all wavelengths during April, unlike the rest of the pre-monsoon months. Investigation of aerosol types indicated the pre-dominance of dust during April and a mixture of urban/open burning with mixed desert dust during the rest of the pre-monsoon months. Potential aerosol source fields were identified over the Indo-Gangetic Plain (IGP), east coast, northwestern India, and oceanic regions; these were estimated at elevated layers of atmosphere during April and May but that at surface layers during February and June. Comparison of aerosol characteristics over the BGP (at Kolkata, KOL) with that at six other coincident AERONET sites over India revealed mean AOD at KOL being 11 to 91% higher than the rest of the AERONET stations, with the relative increase at KOL being the highest during March; this was attributed to persistent high values of both FM and CM AOD unlike the rest of the stations. The monthly mean of SSA was the lowest at KOL among AERONET stations, during February and March. Comparison of the AOD from the AERONET aerosol retrievals over the BGP UA with the coincident Moderate Resolution Imaging Spectroradiometer (MODIS) latest retrievals (C005 and C006) indicated a moderate correlation between the two retrievals; discrepancy in MODIS-retrieved relative distribution of FM and CM AOD was inferred compared to AERONET in the UA.

Intercomparison of Aerosol Optical Thickness Derived from MODIS and in Situ Ground Datasets over Jaipur, a Semi-arid Zone in India

The first detailed seasonal validation has been carried out for the Moderate Resolution Imaging Spectroradiometer (MODIS) Terra and Aqua satellites Level 2.0 Collection Version 5.1 AOT (τMODIS) with Aerosol Robotic Network (AERONET) Level 2.0 AOT (τAERONET) for the years 2009-2012 over semi-arid region Jaipur, northwestern India. The correlation between τMODIS versus τAERONET at 550 nm is determined with different spatial and temporal size windows. The τMODIS overestimates τAERONET within a range of +0.06 ± 0.24 during the pre-monsoon (April-June) season, while it underestimates the τAERONET with -0.04 ± 0.12 and -0.05 ± 0.18 during dry (December-March) and post-monsoon (October-November) seasons, respectively. Correlation without (with) error envelope has been found for pre-monsoon at 0.71 (0.89), post-monsoon at 0.76 (0.94), and dry season at 0.78 (0.95). τMODIS is compared to τAERONET at three more ground AERONET stations in India, i.e., Kanpur, Gual Pahari, and Pune. Furthermore, the performance of MODIS Deep Blue and Aqua AOT550 nm (τDB550 nm and τAqua550 nm) with τAERONET is also evaluated for all considered sites over India along with a U.S. desert site at White Sand, Tularosa Basin, NM. The statistical results reveal that τAqua550 nm performs better over Kanpur and Pune, whereas τDB550 nm performs better over Jaipur, Gual Pahari, and White Sand High Energy Laser Systems Test Facility (HELSTF) (U.S. site).