Search Results (143)

Most existing cloud base height (CBH) retrieval algorithms are only applicable for daytime satellite observations due to their dependence on visible observations. This study presents a novel algorithm to retrieve day and night CBH using infrared observations of the geostationary Advanced Himawari Imager (AHI). The algorithm is featured by integrating deep learning techniques with a physical model. The algorithm first utilizes a convolutional neural network-based model to extract cloud top height (CTH) and cloud water path (CWP) from the AHI infrared observations. Then, a physical model is introduced to relate cloud geometric thickness (CGT) to CWP by constructing a look-up table of effective cloud water content (ECWC). Thus, the CBH can be obtained by subtracting CGT from CTH. The results demonstrate good agreement between our AHI CBH retrievals and the spaceborne active remote sensing measurements, with a mean bias of −0.14 ± 1.26 km for CloudSat-CALIPSO observations at daytime and −0.35 ± 1.84 km for EarthCARE measurements at nighttime. Additional validation against ground-based millimeter wave cloud radar (MMCR) measurements further confirms the effectiveness and reliability of the proposed algorithm across varying atmospheric conditions and temporal scales. Full article

Accurate estimation of ambient PM_2.5 concentrations is crucial for assessing air quality and health risks, particularly in regions with limited ground-based monitoring. Satellite-retrieved data products, such as top-of-atmosphere reflectance (TOAR) and aerosol optical depth (AOD), are widely used for PM_2.5 estimation. However, complex atmospheric conditions cause retrieval gaps in TOAR and AOD products, limiting their reliability. This study introduced a spatiotemporal convolutional approach to fill sampling gaps in TOAR and AOD data from the Himawari-8 geostationary satellite over the Yangtze River Delta (YRD) in 2016. Four machine-learning models (random forest, extreme gradient boosting, gradient boosting, and support vector regression) were used to estimate hourly PM_2.5 concentrations by integrating gap-filled and original TOAR and AOD data with meteorological variables. The random forest model trained on gap-filled TOAR data yielded the highest predictive accuracy (R² = 0.75, RMSE = 18.30 μg m⁻³). Significant seasonal variations in PM_2.5 estimates were found, with TOAR-based models outperforming AOD-based models. Furthermore, we observed that a substantial portion of the YRD population in non-attainment areas is at risk of cardiovascular disease due to chronic PM_2.5 exposure. This study suggests that TOAR-based models offer more reliable PM_2.5 estimates, enhancing air-quality assessments and public health-risk evaluations. Full article

Satellite-based precipitation products (SPPs) have become a crucial source of quantitative global precipitation data. Geostationary Orbit (GEO) satellites provide high spatiotemporal resolution but tend to have lower accuracy, while Low Earth Orbit (LEO) satellites provide more precise precipitation estimates but suffer from lower temporal resolution due to their limited observation frequency. This study proposes an efficient algorithm for integrating and enhancing precipitation estimates from multiple satellite observations. The target domain includes the Full Disk (FD) and the extended East Asia (EA) regions, both of which are observable by GEO satellites, such as Himawari-8, serving as the GEO platform in this study. The algorithm involves four steps: pre-data preparation, LEO morphing, adjustment, and final merging. It produces Early and Late composite products with 10-min temporal and up to 2 km spatial resolution and significantly reduces latency compared to IMERG. Specifically, the Early and Late products can be generated with approximate latencies of 90 min and 270 min, respectively—much faster than Integrated Multi-satellite Retrievals for GPM (IMERG)’s Early (4-h) and Late (14-h) products. A key feature of the proposed method is the use of accuracy-based weighting derived from radar-based validation, enabling dynamic merging that reflects the reliability of each satellite observation. Statistical validation using Global Telecommunication System (GTS) precipitation data confirmed the positive impact of the proposed bias correction and merging method. In particular, the Late product achieved accuracy comparable to or higher than that of IMERG Early and IMERG Late, despite its significantly shorter latency. However, its accuracy was still lower than that of IMERG Final, which benefits from additional gauge-based correction but is released with a delay of several months. Full article

Atmospheric aerosols play an important role in the ecological environment, climate change, and human health. Aerosol optical depth (AOD) is the main measurement of aerosols. The next-generation geostationary satellite Himawari-8, loaded with the Advanced Himawari Imager (AHI), provides observation-based estimates of the hourly AOD. However, a highly accurate evaluation of AOD using AHI is still limited. In this paper, we establish a Stacked Denoising AutoEncoder (SDAE) model to retrieve highly accurate AOD using AHI. We explore the SDAE to retrieve AOD by taking the ground-observed AOD as the output and taking the AHI image, the month, hour, latitude, and longitude as the input data. This approach was tested in the Hubei province of China. Traditional machine learning methods such as Extreme Learning Machines (ELMs), BackPropagation Neural Networks (BPNNs), and Support Vector Machines (SVMs) are also used to evaluate model performance. The results show that the proposed method has the highest accuracy. We also compare the proposed method with ground-observed AOD measurements at the Wuhan University site, showing good consistency between the satellite-retrieved AOD and the ground-observed value. The study of the spatiotemporal change pattern of the hourly AOD in the Hubei province shows that the algorithm has good stability in the face of changes in the angle and intensity of sunlight. Full article

High-resolution (2 km), high-frequency (hourly) SST data of the Advanced Himawari Imager (AHI) flown onboard the Japanese Himawari-8 geostationary satellite were used to derive the monthly climatology of temperature fronts in the South China Sea. The SST data from 2015 to 2022 were processed with the Belkin–O’Reilly algorithm to generate maps of SST gradient magnitude GM. The GM maps were log-transformed to enhance contrasts in digital maps and reveal additional features (fronts). The combination of high-resolution, cloud-free, four-day-composite SST imagery from AHI, the advanced front-preserving gradient algorithm BOA, and digital contrast enhancement with the log-transformation of SST gradients allowed us to identify numerous mesoscale/submesoscale fronts (including a few fronts that have never been reported) and document their month-to-month variability and spatial patterns. The spatiotemporal variability of SST fronts was analyzed in detail in five regions: (1) In the Taiwan Strait, six fronts were identified: the China Coastal Front, Taiwan Bank Front, Changyun Ridge Front, East Penghu Channel Front, and Eastern/Western Penghu Islands fronts; (2) the Guangdong Shelf is dominated by the China Coastal Front in winter, with the eastern and western Guangdong fronts separated by the Pearl River outflow in summer; (3) Hainan Island is surrounded by upwelling fronts of various nature (wind-driven coastal and topographic) and tidal mixing fronts; in the western Beibu Gulf, the Red River Outflow Front extends southward as the Vietnam Coastal Front, while the northern Beibu Gulf features a tidal mixing front off the Guangxi coast; (4) Off SE Vietnam, the 11°N coastal upwelling gives rise to a summertime front, while the Mekong Outflow and associated front extend seasonally toward Cape Camau, close to the Gulf of Thailand Entrance Front; (5) In the Luzon Strait, the Kuroshio Front manifests as a chain of three fronts across the Babuyan Islands, while west of Luzon Island a broad offshore frontal zone persists in winter. The summertime eastward jet (SEJ) off SE Vietnam is documented from five-day mean SST data. The SEJ emerges in June–September off the 11°N coastal upwelling center and extends up to 114°E. The zonally oriented SEJ is observed to be located between two large gyres, each about 300 km in diameter. Full article

Wind field retrieval, a crucial component of weather forecasting, has been significantly enhanced by recent advances in deep learning. However, existing approaches that are primarily focused on wind speed retrieval are limited by their inability to achieve real-time, full-coverage retrievals at large scales. To address this problem, we propose a novel multi-scale fusion retrieval (MFR) method, leveraging geostationary observation satellites. At the mesoscale, MFR incorporates a cloud-to-wind transformer model, which employs local self-attention mechanisms to extract detailed wind field features. At large scales, MFR incorporates a multi-encoder coordinate U-net model, which incorporates multiple encoders and utilises coordinate information to fuse meso- to large-scale features, enabling accurate and regionally complete wind field retrievals, while reducing the computational resources required. The MFR method was validated using Level 1 data from the Himawari-8 satellite, covering a geographic range of 0–60°N and 100–160°E, at a resolution of 0.25°. Wind field retrieval was accomplished within seconds using a single graphics processing unit. The mean absolute error of wind speed obtained by the MFR was 0.97 m/s, surpassing the accuracy of the CFOSAT and HY-2B Level 2B wind field products. The mean absolute error for wind direction achieved by the MFR was 23.31°, outperforming CFOSAT Level 2B products and aligning closely with HY-2B Level 2B products. The MFR represents a pioneering approach for generating initial fields for large-scale grid forecasting models. Full article

High-resolution (2 km) high-frequency (hourly) SST data from 2015 to 2021 provided by the Advanced Himawari Imager (AHI) onboard the Japanese Himawari-8 geostationary satellite were used to study spatial and temporal variability of the China Coastal Front (CCF) in the South China Sea. The SST data were processed with the Belkin and O’Reilly (2009) algorithm to generate monthly maps of the CCF’s intensity (defined as SST gradient magnitude GM) and frontal frequency (FF). The horizontal structure of the CCF was investigated from cross-frontal distributions of SST along 11 fixed lines that allowed us to determine inshore and offshore boundaries of the CCF and calculate the CCF’s strength (defined as total cross-frontal step of SST). Combined with the results of Part 1 of this study, where the CCF was documented in the East China Sea, the new results reported in this paper allowed the CCF to be traced from the Yangtze Bank to Hainan Island. The CCF is continuous in winter, when its intensity peaks at 0.15 °C/km (based on monthly data). In summer, when the Guangdong Coastal Current reverses and flows eastward, the CCF’s intensity is reduced to 0.05 °C/km or less, especially off western Guangdong, where the CCF vanishes almost completely. Owing to its breadth (50–100 km, up to 200 km in the Taiwan Strait), the CCF is a very strong front, especially in winter, when the total SST step across the CCF peaks at 9 °C in the Taiwan Strait. The CCF’s strength decreases westward to 6 °C off eastern Guangdong, 5 °C off western Guangdong, and 2 °C off Hainan Island, all in mid-winter. Full article

As a contribution to the Integrated Marine Observing System (IMOS), the Bureau of Meteorology introduces new reprocessed Himawari-8 satellite-derived Sea Surface Temperature (SST) products. The Radiative Transfer Model and a Bayesian cloud clearing method is used to retrieve SSTs every 10 min from the geostationary satellite Himawari-8. An empirical Sensor Specific Error Statistics (SSES) model, introduced herein, is applied to calculate bias and standard deviation for the retrieved SSTs. The SST retrieval and compositing method, along with validation results, are discussed. The monthly statistics for comparisons of Himawari-8 Level 2 Product (L2P) skin SST against in situ SST quality monitoring (iQuam) in situ SST datasets, adjusted for thermal stratification, showed a mean bias of −0.2/−0.1 K and a standard deviation of 0.4–0.7 K for daytime/night-time after bias correction, where satellite zenith angles were less than 60° and the quality level was greater than 2. For ease of use, these native resolution SST data have been composited using a method introduced herein that retains retrieved measurements, to hourly, 4-hourly and daily SST products, and projected onto the rectangular IMOS 0.02 degree grid. On average, 4-hourly products cover ≈$10\%$ more of the IMOS domain, while one-night composites cover ≈$25\%$ more of the IMOS domain than a typical 1 h composite. All available Himawari-8 data have been reprocessed for the September 2015–December 2022 period. The 10 min temporal resolution of the newly developed Himawari-8 SST data enables a daily composite with enhanced spatial coverage, effectively filling in SST gaps caused by transient clouds occlusion. Anticipated benefits of the new Himawari-8 products include enhanced data quality for applications like IMOS OceanCurrent and investigations into marine thermal stress, marine heatwaves, and ocean upwelling in near-coastal regions. Full article

The Advanced Geostationary Radiation Imager (AGRI) sensor on board the geostationary satellite Fengyun-4B (FY-4B) is capable of capturing particles in different phases in the atmospheric environment and acquiring aerosol observation data with high spatial and temporal resolution. To understand the quality of the Land Aerosol (LDA) product of AGRI and its application prospects, we conducted a comprehensive evaluation of the AGRI LDA AOD. Using the 550 nm AGRI LDA AOD (550 nm) of nearly 1 year (1 October 2022 to 30 September 2023) to compare with the Aerosol Robotic Network (AERONET), MODIS MAIAC, and Himawari-9/AHI AODs. Results show the erratic algorithmic performance of AGRI LDA AOD, the correlation coefficient (R), mean error (Bias), root mean square error (RMSE), and the percentage of data with errors falling within the expected error envelope of $\pm (0.05+0.15\times {\mathrm{AOD}}_{AERONET})$ (within EE15) of the LDA AOD dataset are 0.55, 0.328, 0.533, and 34%, respectively. The LDA AOD appears to be overestimated easily in the southern and western regions of China and performs poorly in the offshore areas, with an R of 0.43, a Bias of 0.334, a larger RMSE of 0.597, and a global climate observing system fraction (GCOSF) percentage of 15% compared to the inland areas (R = 0.60, Bias = 0.163, RMSE = 0.509, GCOSF = 17%). Future improvements should focus on surface reflectance calculation, water vapor attenuation, and more suitable aerosol model selection to improve the algorithm’s accuracy. Full article

Through the application of cloud top temperature data and the extraction of supercooled cloud information in cloud-type data from the next-generation Himawari-8 geostationary satellite with high spatial–temporal resolution, a quantitative evaluation of the forecasting performance of the weather modification model named the Cloud and Precipitation Explicit Forecasting System (CPEFS) was conducted. The evaluation, based on selected forecast cases from 8 days in September and October 2018 initialized at 00 and 12 UTC every day, focused especially on the forecasting performance in supercooled clouds (vertical integrated supercooled liquid water, VISL > 0), including the comprehensive spatial distribution of cloud top temperature (CTT) and 3 h precipitation over 0.1 mm (R3 > 0.1). The results indicated that the forecasting performance for VISL > 0 was relatively good, with the Threat Score (TS) ranging from 0.46 to 0.67. The forecasts initialized at 12 UTC slightly outperformed the forecasts initialized at 00 UTC. Additionally, the corresponding spatial Anomaly Correlation Coefficient (ACC) of CTT between forecasts and observations was 0.23, and the TS for R3 > 0.1 reached as high as 0.87. For a mix of cold and warm cloud systems, there was a correlation between the forecasting performance of VISL > 0 and CTT. The trends in the TS for VISL > 0 and the ACC of CTT aligned with the forecast lead-time. Full article

Accurate cloud detection is a crucial initial stage in optical satellite remote sensing. In this study, a daytime cloud mask model is proposed for the Advanced Geostationary Radiation Imager (AGRI) onboard the Fengyun 4A (FY-4A) satellite based on a deep learning approach. The model, named “Convolutional and Attention-based Cloud Mask Net (CACM-Net)”, was trained using the 2021 dataset with CALIPSO data as the truth value. Two CACM-Net models were trained based on a satellite zenith angle (SZA) < 70° and >70°, respectively. The study evaluated the National Satellite Meteorological Center (NSMC) cloud mask product and compared it with the method established in this paper. The results indicate that CACM-Net outperforms the NSMC cloud mask product overall. Specifically, in the SZA < 70° subset, CACM-Net enhances accuracy, precision, and F1 score by 4.8%, 7.3%, and 3.6%, respectively, while reducing the false alarm rate (FAR) by approximately 7.3%. In the SZA > 70° section, improvements of 12.2%, 19.5%, and 8% in accuracy, precision, and F1 score, respectively, were observed, with a 19.5% reduction in FAR compared to NSMC. An independent validation dataset for January–June 2023 further validates the performance of CACM-Net. The results show improvements of 3.5%, 2.2%, and 2.8% in accuracy, precision, and F1 scores for SZA < 70° and 7.8%, 11.3%, and 4.8% for SZA > 70°, respectively, along with reductions in FAR. Cross-comparison with other satellite cloud mask products reveals high levels of agreement, with 88.6% and 86.3% matching results with the MODIS and Himawari-9 products, respectively. These results confirm the reliability of the CACM-Net cloud mask model, which can produce stable and high-quality FY-4A AGRI cloud mask results. Full article

The advancement of more precise remote sensing inversion technology for dust aerosols has long been a hot topic in the field of the atmospheric environment. In 2023, China experienced 18 dust-related weather events, predominantly in spring. These high-intensity and frequent dust events have attracted considerable attention. However, gridded observation data of dust intensity levels are not collected in current dust monitoring and forecasting operations. Based on the Himawari 9 geostationary satellite data, this study establishes a new method to identify spring dust events. This method integrates the brightness temperature difference method and the multiple infrared dust index, taking into account the response discrepancies of the multiple infrared dust index under various underlying surfaces. Furthermore, by obtaining dynamic background brightness temperature values eight times a day, threshold statistics are applied to analyze the correlation between the infrared difference dust index and ground-observed dust level, so as to establish a satellite-based near-surface dust intensity level identification algorithm. This algorithm aims to improve dust detection accuracy, and to provide more effective gridded observation support for dust forecasting and monitoring operations. The test results indicate that the algorithm can effectively identify the presence or absence of dust, with a misjudgment rate of less than 3%. With regard to dust intensity, the identification of blowing sand and floating dust aligns relatively well with ground-based observations, but notable uncertainties exist in determining a dust intensity of sand-storm level or above. Among these uncertainties, the differences between ground-based observations and satellite identification caused by non-grounded dust in the upper air, and the selection of dust identification thresholds, are two important error sources in the dust identification results of this study. Full article

Aerosols exert a significant influence on the brightness temperature observed in the thermal infrared (IR) channels, yet the specific contributions of various aerosol types remain underexplored. This study integrated the Copernicus Atmosphere Monitoring Service (CAMS) atmospheric composition reanalysis data into the Radiative Transfer for TOVS (RTTOV) model to quantify the aerosol effects on brightness temperature (BT) simulations for the Advanced Himawari Imager (AHI) aboard the Himawari-8 geostationary satellite. Two distinct experiments were conducted: the aerosol-aware experiment (AER), which accounted for aerosol radiative effects, and the control experiment (CTL), in which aerosol radiative effects were omitted. The CTL experiment results reveal uniform negative bias (observation minus background (O-B)) across all six IR channels of the AHI, with a maximum deviation of approximately −1 K. Conversely, the AER experiment showed a pronounced reduction in innovation, which was especially notable in the 10.4 μm channel, where the bias decreased by 0.7 K. The study evaluated the radiative effects of eleven aerosol species, all of which demonstrated cooling effects in the AHI’s six IR channels, with dust aerosols contributing the most significantly (approximately 86%). In scenarios dominated by dust, incorporating the radiative effect of dust aerosols could correct the brightness temperature bias by up to 2 K, underscoring the substantial enhancement in the BT simulation for the 10.4 μm channel during dust events. Jacobians were calculated to further examine the RTTOV simulations’ sensitivity to aerosol presence. A clear temporal and spatial correlation between the dust concentration and BT simulation bias corroborated the critical role of the infrared channel data assimilation on geostationary satellites in capturing small-scale, rapidly developing pollution processes. Full article

Machine learning methods have been recognized as rapid methods for satellite-based aerosol retrievals but have not been widely applied in geostationary satellites. In this study, we developed a wide and deep learning model to retrieve the aerosol optical depth (AOD) using Himawari-8. Compared to traditional deep learning methods, we embedded a “wide” modeling component and tested the proposed model across China using independent training (2016–2018) and test (2019) datasets. The results showed that the “wide” model improves the accuracy and enhances model interpretability. The estimates exhibited better accuracy (R² = 0.81, root-mean-square errors (RMSEs) = 0.19, and within the estimated error (EE) = 63%) than those of the deep-only models (R² = 0.78, RMSE = 0.21, within the EE = 58%). In comparison with extreme gradient boosting (XGBoost) and Himawari-8 V2.1 AOD products, there were also significant improvements. In addition to higher accuracy, the interpretability of the proposed model was superior to that of the deep-only model. Compared with other seasons, higher contributions of spring to the AOD concentrations were interpreted. Based on the application of the wide and deep learning model, the near-real-time variation of the AOD over China could be captured with an ultrafine temporal resolution. Full article

Fine particulate matter with an aerodynamic diameter less than 2.5 µm ($\mathrm{P}{\mathrm{M}}_{2.5}$) profoundly affects environmental systems, human health and economic structures. Multi-source data and advanced machine or deep-learning methods have provided a new chance for estimating the $\mathrm{P}{\mathrm{M}}_{2.5}$ concentrations at a high spatiotemporal resolution. In this paper, the Random Forest (RF) algorithm was applied to estimate hourly $\mathrm{P}{\mathrm{M}}_{2.5}$ of the North China area (Beijing–Tianjin–Hebei, BTH) based on the next-generation geostationary meteorological satellite Himawari-8/AHI (Advanced Himawari Imager) aerosol optical depth (AOD) products. To improve the estimation of $\mathrm{P}{\mathrm{M}}_{2.5}$ concentration across large areas, we construct a method for co-weighting the environmental similarity and the geographical distances by using an attention mechanism so that it can efficiently characterize the influence of spatial–temporal information hidden in adjacent ground monitoring sites. In experiment results, the hourly $\mathrm{P}{\mathrm{M}}_{2.5}$ estimates are well correlated with ground measurements in BTH, with a coefficient of determination (${\mathrm{R}}^2$) of 0.887, a root-mean-square error (RMSE) of 18.31 $\mathsf{\mu }$g/${\mathrm{m}}^3$, and a mean absolute error (MAE) of 11.17 µg/${\mathrm{m}}^3$, indicating good model performance. In addition, this paper makes a comprehensive analysis of the effectiveness of multi-source data in the estimation process, in this way, to simplify the model structure and improve the estimation efficiency of the model while ensuring its accuracy. Full article