Advanced Baseline Imager Research Articles

Contrail altitude estimation using GOES-16 ABI data and deep learning

Abstract. The climate impact of persistent aircraft contrails is currently estimated to be comparable to that due to aviation-emitted CO2. A potential near-term and low-cost mitigation option is contrail avoidance, which involves rerouting aircraft around ice-supersaturated regions, preventing the formation of persistent contrails. Current forecasting methods for these regions of ice supersaturation have been found to be inaccurate when compared to in situ measurements. Further assessment and improvements of the quality of these predictions can be realized by comparison with observations of persistent contrails, such as those found in satellite imagery. In order to further enable comparison between these observations and contrail predictions, we develop a deep learning algorithm to estimate contrail altitudes based on GOES-16 Advanced Baseline Imager (ABI) infrared imagery. This algorithm is trained using a dataset of 3267 contrails found within Cloud–Aerosol Lidar with Orthogonal Polarization (CALIOP) data and achieves a root mean square error (RMSE) of 570 m. The altitude estimation algorithm outputs probability distributions for the contrail top altitude in order to represent predictive uncertainty. The 95 % confidence intervals constructed using these distributions, which are shown to contain approximately 95 % of the contrail data points, are found to be 2.2 km thick on average. These intervals are found to be 34.1 % smaller than the 95 % confidence intervals constructed using flight altitude information alone, which are 3.3 km thick on average. Furthermore, we show that the contrail altitude estimates are consistent in time and, in combination with contrail detections, can be used to observe the persistence and three-dimensional (3D) evolution of contrail-forming regions from satellite images alone.

Extracting Wetlands in Coastal Louisiana from the Operational VIIRS and GOES-R Flood Products

Visible Infrared Imaging Radiometer Suite (VIIRS) and Advanced Baseline Imager (GOES-R ABI) flood products have been widely used by the National Weather Service (NWS) for river flood monitoring, and by the Federal Emergency Management Agency (FEMA) for rescue and relief efforts. Some water bodies, like wetlands, are detected as water but not marked as permanent or normal water, which may result in their misclassification as floodwaters by VIIRS and GOES-R flood products. These water bodies generally do not cause significant property damage or fatalities, but they can complicate the identification of truly hazardous floods. This study utilizes the severe Louisiana flood event caused by Hurricane Ida to demonstrate how to differentiate wetlands from real-hazard flooding. Since Hurricane Ida made landfall in 2021, and there was no major flood event in 2022, VIIRS and ABI flood data from 2021 and 2022 were selected. The difference in annual total flooding days between 2021 and 2022 was calculated and combined with long-time flood frequency to distinguish non-hazard floodwaters due to wetlands identified from real-hazard floods caused by the hurricane. The results were compared with the wetlands from the change detection analysis. The confusion matrix analysis indicated an accuracy of 91.58%, precision of 89.97%, and F1-score of 76.63% for the VIIRS flood products. For the GOES-R ABI flood products, the confusion matrix analysis yielded an accuracy of 86.88%, precision of 97.49%, and F1-score of 75.21%. The accuracy and F1-score values for the GOES-R ABI flood products are slightly lower than those for the VIIRS flood products, possibly due to their lower spatial resolution, but still within a feasible range.

Retrieval of boundary layer precipitable water from GOES ABI using machine learning techniques

Abstract Low-level moisture is an important ingredient for forecasting severe storms, especially over the Great Plains where severe storms often develop along the dryline. Although ground-based observation systems such as lidar or radiosondes on weather balloons provide accurate information on low-level moisture, data are provided at limited locations, and the low temporal resolution of radiosondes makes it difficult to track a rapidly developing dryline. Geostationary satellites provide high spatial and temporal observation, but the channels of current geostationary satellites are mostly sensitive to water vapor at mid to upper levels. However, the split window difference (SWD) between the “clean” window channel (10.3 μm) and the “dirty” window channel (12.3 μm) is commonly used for estimating low-level water vapor. However, this estimation is complicated by surface temperature contributions, dependence on the lapse rate, and nonlinear relationships between SWD and moisture. This study applies machine learning techniques to infer boundary layer precipitable water (BLPW) from Geostationary Operational Environmental Satellite (GOES) Advanced Baseline Imager (ABI) data. Since there are few observations that cover wide regions for training convolutional neural networks, especially for the atmosphere above the surface, High-Resolution Rapid Refresh (HRRR) model outputs are used as the truth for training. Statistical results show that using ABI channel 13, 14, and 15 brightness temperatures, skin temperature, solar zenith angle, and GOES total precipitable water product as additional inputs show the lowest MSE. Case study results show that the model developed in this study is good at capturing strong moisture gradients and depicting a dryline.

Increasing aerosol optical depth spatial and temporal availability by merging datasets from geostationary and sun-synchronous satellites

Abstract. This comprehensive study analyzed aerosol products from six low-Earth orbit (LEO) and geostationary Earth orbit (GEO) sensors. LEO sensors like the MODerate resolution Imaging Spectroradiometer (MODIS) and VIsible InfraRed Suite (VIIRS) provide one to two daily global measurements, while GEO sensors (Advanced Himawari Imager: AHI, Advanced Baseline Imager: ABI) offer high-frequency data (∼ 10 min) over specific regions. The combination of LEO and GEO capabilities offers expanded coverage of the global aerosol system if aerosol retrievals are applied consistently across all sensors and packaged in an easy-to-use product. The Dark Target aerosol retrieval algorithm was applied to the six sensors, and the resulting Level 2 aerosol optical depth (AOD) products were gridded and merged into a Level 3 quarter-degree latitude–longitude grid with a 30 min temporal resolution, providing the necessary consistency and packaging. Validation of this packaged Level 3 AOD product against Aerosol Robotics NETwork (AERONET) measurements across global locations showcased the merged product's robustness with a correlation coefficient of 0.83, revealing a global mean bias of approximately ±0.05, with 65.5 % of retrievals falling within an expected uncertainty range, underlining the reliability of the dataset. The new gridded Level 3 dataset significantly improved daily global coverage to nearly 45 %, overcoming the limitations of individual sensors, which typically range from 12 % to 25 %. Furthermore, this merged dataset approximates the diurnal cycle of AOD observed by AERONET, thus offering insights into diurnal signatures retrieved elsewhere. The resulting dataset's high spatiotemporal resolution and improved global coverage, especially in regions covered by GEO sensors (Americas and Asia), make it a valuable tool for diverse applications. Tracking aerosol transport from phenomena like wildfires and dust storms is gaining precision, enabling enhanced air quality forecasting and hindcasting. Additionally, the study positions the merged dataset as a significant asset for evaluating and intercomparing regional or global model simulations, which was previously unattainable in such a gridded format. The dataset and fusion framework layout in this study have the potential to include data from recently (future) launched other GEO (FCI, AMI) and LEO (PACE, VIIRS-JPSS) sensors.

Novel Comparison of Pyrocumulonimbus Updrafts to Volcanic Eruptions and Supercell Thunderstorms Using Optical Flow Techniques

AbstractConvective dynamics in a supercell thunderstorm, a volcanic eruption, and two pyrocumulonimbus (pyroCb) events are compared by computing cloud‐top divergence (CTD) with an optical flow technique called Deepflow. Visible 0.64‐μm imagery sequences from Geostationary Operational Environmental Satellites (GOES)‐R series Advanced Baseline Imager (ABI) are used as input into the optical flow algorithm. CTD is computed after post‐processing of the retrieved motions. Analysis is performed on specific image times, as well as the full time series of each case. Multiple CTD‐based parameters, such as the maximum and the two‐dimensional area exceeding a specified CTD threshold, are examined along with the optical flow‐retrieved wind speed. CTD is shown to accurately and quantitatively represent the behavior and magnitude of different deep convective phenomena, including distinguishing between convective pulses within each individual event. CTD captures updraft intensification as well as differences in convective activity between two pyroCb events and individual updraft pulses occurring within a single pyroCb event. Finally, the characteristics of high‐altitude smoke plumes injected by two separate pyroCb pulses are linked to CTD using ultraviolet aerosol index and satellite imagery. Optical flow‐derived parameters can therefore be applied to individual pyroCbs in real‐time, with potential to characterize pyroCb smoke source inputs for downstream smoke modeling applications and to facilitate future tools supporting air quality modeling and firefighting efforts.

The effects of changes in HITRAN and the water vapor continuum model on infrared radiative transfer calculations and remote sensing applications

This study examines the implications of recent updates to the High Resolution Transmission (HITRAN) database and the MT_CKD water vapor continuum model on infrared radiative transfer calculations and satellite remote sensing applications. Specifically, it assesses the impact of the latest versions of the HITRAN database (i.e. HITRAN2016 and HITRAN2020) and the MT_CKD WV continuum model (i.e. versions 3.2 and 4.1.1) on clear-sky infrared flux calculations and the simulation of satellite infrared channels, specifically those of the Geostationary Operational Environmental Satellite (GOES) Advanced Baseline Imager (ABI). The results indicate that updates to the MT_CKD model tend to reduce atmospheric opacity, while updates to the HITRAN database have the opposite effect. The most significant differences noted are attributed to updates in the MT_CKD, leading to an increase in upward flux of 0.13 Wm−2 for the TRO and a decrease in downward flux of 0.28 Wm−2 for the MLS profile. However, these impacts are not uniform across the spectrum, atmospheric height, or among different atmospheric profiles. Notably, changes in the WV continuum model have significant effects in the far-infrared region. The updates in HITRAN exhibit pronounced differences associated with changes in the spectroscopic parameters of H2O, CO2 and O3. The results of GOES’ ABI channels indicate that changes in HITRAN lead to an average brightness temperature decrease of 0.3 K for channel 12, which exceeds the channel’s sensitivity. This research underscores the importance of continuous updates and evaluations of spectroscopic databases and radiative transfer models to improve the accuracy of atmospheric remote sensing data.

Parameterizing spectral surface reflectance relationships for the Dark Target aerosol algorithm applied to a geostationary imager

Abstract. Originally developed for the moderate resolution imaging spectroradiometer (MODIS) in polar, sun-synchronous low earth orbit (LEO), the Dark Target (DT) aerosol retrieval algorithm relies on the assumption of a surface reflectance parameterization (SRP) over land surfaces. Specifically for vegetated and dark-soiled surfaces, values of surface reflectance in blue and red visible-wavelength bands are assumed to be nearly linearly related to each other and to the value in a shortwave infrared (SWIR) wavelength band. This SRP also includes dependencies on scattering angle and a normalized difference vegetation index computed from two SWIR bands (NDVISWIR). As the DT retrieval algorithm is being ported to new sensors to continue and expand the aerosol data record, we assess whether the MODIS-assumed SRP can be used for these sensors. Here, we specifically assess SRP for the Advanced Baseline Imager (ABI) aboard the Geostationary Operational Environmental Satellite (GOES)-16/East (ABIE). First, we find that using MODIS-based SRP leads to higher biases and artificial diurnal signatures in aerosol optical depth (AOD) retrievals from ABIE. The primary reason appears to be that the geostationary orbit (GEO) encounters an entirely different set of observation geometry than does LEO, primarily with regard to solar angles coupled with fixed-view angles. Therefore, we have developed a new SRP for GEO that draws the angular shape of the surface bidirectional reflectance. We also introduce modifications to the parameterization of both red–SWIR and blue–red spectral relationships to include additional information. The revised red–SWIR SRP includes the solar zenith angle, NDVISWIR, and land-type percentage from an ancillary database. The blue–red SRP adds dependencies on the scattering angle and NDVISWIR. The new SRPs improve the AOD retrieval of ABIE in terms of overall less bias and mitigation of the overestimation around local noon. The average bias of the DT AOD compared to the Aerosol Robotic Network (AERONET) AOD shows a reduction from 0.08 to 0.03, while the bias of local solar noon decreases from 0.12 to 0.03. The agreement between the DT and AERONET AOD is established through a regression slope of 1.06 and a y intercept of 0.01 with a correlation coefficient of 0.74. By using the new SRP, the percentage of data falling within the expected error range (±0.05 % + 15 %) is notably increased from 54 % to 78 %.

Superresolution of GOES-16 ABI Bands to a Common High Resolution with a Convolutional Neural Network

Abstract Superresolution is the general task of artificially increasing the spatial resolution of an image. The recent surge in machine learning (ML) research has yielded many promising ML-based approaches for performing single-image superresolution including applications to satellite remote sensing. We develop a convolutional neural network (CNN) to superresolve the 1- and 2-km bands on the GOES-R series Advanced Baseline Imager (ABI) to a common high resolution of 0.5 km. Access to 0.5-km imagery from ABI band 2 enables the CNN to realistically sharpen lower-resolution bands without significant blurring. We first train the CNN on a proxy task, which allows us to only use ABI imagery, namely, degrading the resolution of ABI bands and training the CNN to restore the original imagery. Comparisons at reduced resolution and at full resolution with Landsat-8/Landsat-9 observations illustrate that the CNN produces images with realistic high-frequency detail that is not present in a bicubic interpolation baseline. Estimating all ABI bands at 0.5-km resolution allows for more easily combining information across bands without reconciling differences in spatial resolution. However, more analysis is needed to determine impacts on derived products or multispectral imagery that use superresolved bands. This approach is extensible to other remote sensing instruments that have bands with different spatial resolutions and requires only a small amount of data and knowledge of each channel’s modulation transfer function. Significance Statement Satellite remote sensing instruments often have bands with different spatial resolutions. This work shows that we can artificially increase the resolution of some lower-resolution bands by taking advantage of the texture of higher-resolution bands on the GOES-16 ABI instrument using a convolutional neural network. This may help reconcile differences in spatial resolution when combining information across bands, but future analysis is needed to precisely determine impacts on derived products that might use superresolved bands.

Geographical coverage analysis and usage suggestions of temporal averaged aerosol optical depth product from GOES-R satellite data

ABSTRACT Geostationary satellites have the capability to offer AOD products with a higher frequency of observation within a given period, thus improving the geographical coverage of averaged AOD products compared to polar satellites. Moreover, the averaged AOD of geostationary satellites is more reflective of the average aerosol load as compared to AOD products derived from polar orbit satellites. Despite this, there is still an absence of comprehensive research on the comparative representativeness of AOD mean products from geostationary and polar orbit satellites, and most current research only focuses on retrieval accuracy. This paper compares the geographical coverage of averaged AOD products released by the Advanced Baseline Imager (ABI) sensor on the Geostationary Operational Environmental Satellite-R (GOES-R) (09:00 UTC-23:00 UTC) with AOD products released by the Moderate-resolution Imaging Spectroradiometer (MODIS) (16:00 UTC and 19:00 UTC) on hourly, daily, semi-monthly and monthly scales. Moreover, the Aerosol Robotic Network datasets were used to evaluate the accuracy of ABI AOD and to propose usage suggestions. In terms of daily AOD products, the AOD daily mean generated by the geostationary satellite GOES-R/ABI consistently outperforms the AOD daily mean generated by MODIS AOD in terms of spatial coverage. However, on monthly scales, the difference is no longer significant. With regard to accuracy, it is proved that when the time scale of averaging is gradually expanded, root mean square error (RMSE) and mean absolute error (MAE) gradually decrease. On the season scale, ABI AOD exhibits the highest level of accuracy during the autumn season (September, October, November); on the spatial scale, ABI AOD exhibits the best accuracy in North America. Therefore, if the daily, semi-monthly or monthly averaged ABI AOD datasets are used, the recommended time is autumn, and the study area is North America.

GOES-R land surface products at Western Hemisphere eddy covariance tower locations

The terrestrial carbon cycle varies dynamically on hourly to weekly scales, making it difficult to observe. Geostationary (“weather”) satellites like the Geostationary Environmental Operational Satellite - R Series (GOES-R) deliver near-hemispheric imagery at a ten-minute cadence. The Advanced Baseline Imager (ABI) aboard GOES-R measures visible and near-infrared spectral bands that can be used to estimate land surface properties and carbon dioxide flux. However, GOES-R data are designed for real-time dissemination and are difficult to link with eddy covariance time series of land-atmosphere carbon dioxide exchange. We compiled three-year time series of GOES-R land surface attributes including visible and near-infrared reflectances, land surface temperature (LST), and downwelling shortwave radiation (DSR) at 314 ABI fixed grid pixels containing eddy covariance towers. We demonstrate how to best combine satellite and in-situ datasets and show how ABI attributes useful for ecosystem monitoring vary across space and time. By connecting observation networks that infer rapid changes to the carbon cycle, we can gain a richer understanding of the processes that control it.

Validating GOES Radar Estimation via Machine Learning to Inform NWP (GREMLIN) Product over CONUS

Abstract Geostationary Operational Environmental Satellites (GOES) Radar Estimation via Machine Learning to Inform NWP (GREMLIN) is a machine learning model that outputs composite reflectivity using GOES-R Series Advanced Baseline Imager (ABI) and Geostationary Lightning Mapper (GLM) input data. GREMLIN is useful for observing severe weather and initializing convection for short-term forecasts, especially over regions without ground-based radars. This study expands the evaluation of GREMLIN’s accuracy against the Multi-Radar Multi-Sensor (MRMS) System to the entire contiguous United States (CONUS) for the entire annual cycle. Regional and temporal variation of validation metrics are examined over CONUS by season, day of year, and time of day. Since GREMLIN was trained with data in spring and summer, root-mean-square difference (RMSD) and bias are lowest in the order of summer, spring, fall, and winter. In summer, diurnal patterns of RMSD follow those of precipitation occurrence. Winter has the highest RMSD because of cold surfaces mistaken as precipitating clouds, but some of these errors can be removed by applying the ABI clear-sky mask product and correcting biases using a lookup table. In GREMLIN, strong echoes are closely related to the existence of lightning and corresponding low brightness temperatures, which result in different error distributions over different regions of CONUS. This leads to negative biases in cold seasons over Washington State, lower 30-dBZ critical success index caused by high misses over the Northeast, and higher false alarms over Florida that are due to higher frequency of lightning.

Analysis of Trapped Small‐Scale Internal Gravity Waves Automatically Detected in Satellite Imagery

AbstractIn water vapor‐sensitive satellite imagery, small‐scale wave‐like perturbations of brightness temperature can be attributed to the presence of trapped internal waves in the troposphere. We present a method for detecting these local perturbations with wavelengths of about 10 km and apply it to imagery from the Advanced Baseline Imager on board the geostationary satellite GOES‐16. The algorithm allows us to analyze 4 years of sub‐hourly data in the southern part of the tropical eastern Pacific, where only a relatively low amount of medium and high clouds obscures the scene. By combining a measure of wave activity/trapping with ERA5 reanalysis data, we connect the occurrence of trapping with the presence of an increased upper‐tropospheric wind shear. This connection is more evident during December, January and February, when upper‐tropospheric jets are more likely. Our work supports existing case and model studies and is a step forward in the statistical and automated analysis of trapped small‐scale internal waves in the atmosphere.

Thermal infrared observations of a western United States biomass burning aerosol plume

Abstract. Biomass burning smoke particles, due to their submicron particle size in relation to the average thermal infrared (TIR) wavelength, theoretically have negligible signals at the TIR channels. However, nearly instantaneous longwave (LW) signatures of thick smoke plumes can be frequently observed at the TIR channels from remotely sensed data, including at 10.6 µm (IR window), as well as in water-vapor-sensitive wavelengths at 7.3, 6.8, and 6.3 µm (e.g., lower, middle, and upper troposphere). We systematically evaluated multiple hypotheses as to causal factors of these IR signatures of biomass burning smoke using a combination of data from the Aqua MODerate resolution Imaging Spectroradiometer (MODIS), Aqua Cloud and the Earth Radiant Energy System (CERES), Geostationary Operational Environmental Satellite 16/17 (GOES-16/17) Advanced Baseline Imager, and Suomi-NPP Visible Infrared Imaging Radiometer Suite (VIIRS) and Cross-track Infrared Sounder (CrIS). The largely clear transmission of light through wildfire smoke in the near infrared indicates that coarse or giant ash particles are unlikely to be the dominant cause. Rather, clear signals in water vapor and TIR channels suggest that both co-transported water vapor injected to the middle to upper troposphere and surface cooling by the reduction of surface radiation by the plume are more significant, with the surface cooling effect of smoke aloft being the most dominant. Giving consideration of the smoke impacts on TIR and longwave, CERES indicates that large wildfire aerosol plumes are more radiatively neutral. Further, this smoke-induced TIR signal may be used to map very optically thick smoke plumes, where traditional aerosol retrieval methods have difficulties.

MAGARA: a Multi-Angle Geostationary Aerosol Retrieval Algorithm

Abstract. For over 40 years, the Geostationary Operational Environmental Satellite (GOES) system has provided frequent snapshots of the Western Hemisphere. The advanced baseline imagers (ABIs) on the GOES-16, GOES-17, and GOES-18 platforms are the first GOES-series imagers that meet the precision requirements for high-quality, aerosol-related research. We present MAGARA, a Multi-Angle Geostationary Aerosol Retrieval Algorithm, that leverages multi-angle ABI imagery to exploit the differences in autocorrelation timescales between surface reflectance, aerosol type, and aerosol loading. MAGARA retrieves pixel-level (up to 1 km) aerosol loading and fine-mode fraction at up to the cadence of the measurements (10 min), fine- and coarse-mode aerosol particle properties at a daily cadence, and surface properties by combining the multi-angle radiances with robust surface characterization inherent to temporally tiled algorithms. We present three case studies, and because GOES-17 was not making observations for one case, we present this as a unique demonstration of the multi-angle algorithm using only a single ABI sensor. We also compare MAGARA retrievals of fine-mode (FM) aerosol optical depth (AOD), coarse-mode (CM) AOD, and single-scattering albedo (SSA) statistically, with coincident AErosol RObotic NETwork (AERONET) spectral deconvolution algorithm (SDA) and inversion retrievals for the same period, and against bias-corrected NOAA GOES-16 and GOES-17 retrieved 550 nm AOD. For MAGARA vs. coincident AERONET over-land 500 nm fine-mode fraction and AOD>0.3, MAE=0.031, RMSE=0.100, and r=0.902, indicating good sensitivity to fine-mode fraction over land, especially for smoky regions. For bias-corrected MAGARA vs. coincident AERONET spectral single-scattering albedo with MAGARA AOD>0.5 (n=116), MAE=0.010, RMSE=0.015, and the correlation is 0.87. MAGARA performs best in regions where surface reflectance varies over long timescales with minimal clouds. This represents a large portion of the western half of the United States, much of north-central Africa and the Middle East, some of central Asia, and much of Australia. For these regions, aerosol type and aerosol loading on timescales as short as 10 min could allow for novel research into aerosol–cloud interactions, improvements to air-quality modeling and forecasting, and tighter constraints on direct aerosol radiative forcing.

A scalable system to measure contrail formation on a per-flight basis

Persistent contrails make up a large fraction of aviation's contribution to global warming. We describe a scalable, automated detection and matching (ADM) system to determine from satellite data whether a flight has made a persistent contrail. The ADM system compares flight segments to contrails detected by a computer vision algorithm running on images from the GOES-16 Advanced Baseline Imager. We develop a flight matching algorithm and use it to label each flight segment as a match or non-match. We perform this analysis on 1.6 million flight segments. The result is an analysis of which flights make persistent contrails several orders of magnitude larger than any previous work. We assess the agreement between our labels and available prediction models based on weather forecasts. Shifting air traffic to avoid regions of contrail formation has been proposed as a possible mitigation with the potential for very low cost/ton-CO2e. Our findings suggest that imperfections in these prediction models increase this cost/ton by about an order of magnitude. Contrail avoidance is a cost-effective climate change mitigation even with this factor taken into account, but our results quantify the need for more accurate contrail prediction methods and establish a benchmark for future development.

Analyzing GOES-R ABI BRDF-adjusted EVI2 time series by comparing with VIIRS observations over the CONUS

The Advanced Baseline Imager (ABI) instruments onboard the Geostationary Operational Environmental Satellite-R series (GOES-R) provide new opportunities to expand their applications in global terrestrial ecology and environmental sciences. It offers a 5–15 min temporal resolution and 0.5–1 km spatial resolution that greatly benefit those environmental applications where spatial resolution can be compromised in favor of more frequent imagery. However, ABI captures reflected radiation at varying sun positions throughout the day, causing significant diurnal variations. The varying off-nadir view across ABI full disk presents further challenges for generating comparable and consistent vegetation observations, such as the two-band Enhanced Vegetation Index (EVI2). To address these challenges, three Bidirectional Reflectance Distribution Function (BRDF) models (RossThick-LiSparse-Reciprocal (RTLSR), Walthall, and Tian models) are examined to generate high-quality BRDF-adjusted EVI2 based on GeoNEX GOES-16 ABI surface reflectance. The ABI BRDF retrievals are then systematically compared to the corresponding VIIRS (Visible Infrared Imaging Radiometer Suite) Nadir BRDF-Adjusted Reflectance (NBAR) products (VNP43 V001) retrieved from RTLSR model. The viewing geometry effects are further analyzed over various land cover types using time series analysis, Pearson Correlation Coefficients (PCC), and Relative Difference (RD) between ABI and VIIRS high-quality time series inversions. The results indicate that the RTLSR model fits well the diurnal ABI observed reflectance and EVI2 values (even for the hot-spot phenomena), and outperforms the Walthall and Tian models for generating ABI high-quality BRDF retrievals with consistently low Root Mean Square Error (RMSE<0.02) and Weight of Determination (WoD < 0.5) across the CONUS. The annual proportion of ABI high-quality BRDF retrievals with a 3-day moving window using the RTLSR model is spatially and statistically comparable to that of VIIRS NBAR with a 16-day retrieval period. Compared to the time series of high-quality VIIRS NBAR EVI2, the BRDF adjusted ABI EVI2 time series still show considerable impacts of view zenith angles over evergreen forest, shrubland, and savanna pixels. However, such impacts are not significant over spatially homogeneous pixels such as cropland and deciduous forests. These results emphasize the need for BRDF correction when considering the impact of off-nadir observations from geostationary satellites on vegetation monitoring. They also demonstrate the advantages of geostationary satellites in capturing rapid changes in greenness trajectories.

A satellite chronology of plumes from the April 2021 eruption of La Soufrière, St Vincent

Abstract. Satellite instruments play a valuable role in detecting, monitoring and characterising emissions of ash and gas into the atmosphere during volcanic eruptions. This study uses two satellite instruments, the Infrared Atmospheric Sounding Interferometer (IASI) and the Advanced Baseline Imager (ABI), to examine the plumes of ash and sulfur dioxide (SO2) from the April 2021 eruption of La Soufrière, St Vincent. The frequent ABI data have been used to construct a 14 d chronology of a series of explosive events at La Soufrière, which is then complemented by measurements of SO2 from IASI, which is able to track the plume as it is transported around the globe. A minimum of 35 eruptive events were identified using true, false and brightness temperature difference maps produced with the ABI data. The high temporal resolution images were used to identify the approximate start and end times, as well as the duration and characteristics of each event. From this analysis, four distinct phases within the 14 d eruption have been defined, each consisting of multiple explosive events with similar characteristics: (1) an initial explosive event, (2) a sustained event lasting over 9 h, (3) a pulsatory phase with 25 explosive events in a 65.3 h period and (4) a waning sequence of explosive events. It is likely that the multiple explosive events during the April 2021 eruption contributed to the highly complex plume structure that can be seen in the IASI measurements of the SO2 column amounts and heights. The bulk of the SO2 from the first three phases of the eruption was transported eastwards, which based on the wind direction at the volcano implies that the SO2 was largely in the upper troposphere. Some of the SO2 was carried to the south and west of the volcano, suggesting a smaller emission of the gas into the stratosphere, there being a shift in wind direction around the height of the tropopause. The retrieved SO2 heights show that the plume had multiple layers but was largely concentrated between 13 and 19 km, with the majority of the SO2 being located in the upper troposphere and around the height of the tropopause, with some emission into the stratosphere. An average e-folding time of 6.07±4.74 d was computed based on the IASI SO2 results: similar to other tropical eruptions of this magnitude and height. The SO2 was trackable for several weeks after the eruption and is shown to have circulated the globe, with parts of it reaching as far as 45∘ S and 45∘ N. Using the IASI SO2 measurements, a time series of the total SO2 mass loading was produced, with this peaking on 13 April (descending orbits) at 0.31±0.09 Tg. Converting these mass values to a temporally varying SO2 flux demonstrated that the greatest emission occurred on 10 April with that measurement incorporating SO2 from the second phase of the eruption (sustained emission) and the beginning of the pulsatory phase. The SO2 flux is then shown to fall during the later stages of the eruption: suggesting a reduction in eruptive energy, something also reflected in ash height estimates obtained with the ABI instrument. A total SO2 emission of 0.63±0.5 Tg of SO2 has been derived, although due to limitations associated with the retrieval, particularly in the first few days after the eruption began, this, the retrieved column amounts and the total SO2 mass on each day should be considered minimum estimates. There are a number of similarities between the 1979 and 2021 eruptions at La Soufrière, with both eruptions consisting of a series of explosive events with varied heights and including some emission into the stratosphere. These similarities highlight the importance of in-depth investigations into eruptions and the valuable contribution of satellite data for this purpose; as these studies aid in learning about a volcano's behaviour, which may allow for better preparation for future eruptive activity.

Deep Learning–Based Summertime Turbulence Intensity Estimation Using Satellite Observations

Abstract Turbulence is what we want to avoid the most during flight. Numerical weather prediction (NWP) model–based methods for diagnosing turbulence have offered valuable guidance for pilots. NWP-based turbulence diagnostics show high accuracy in detecting turbulence in general. However, there is still room for improvements such as capturing convectively induced turbulence. In such cases, observation data can be beneficial to correctly locate convective regions and help provide corresponding turbulence information. Geostationary satellite data are commonly used for upper-level turbulence detection by utilizing its water vapor band information. The Geostationary Operational Environmental Satellite (GOES)-16 carries the Advanced Baseline Imager (ABI), which enables us to observe further down into the atmosphere with improved spatial, temporal, and spectral resolutions. Its three water vapor bands allow us to observe different vertical parts of the atmosphere, and from its infrared window bands, convective activity can be inferred. Such multispectral information from ABI can be helpful in inferring turbulence intensity at different vertical levels. This study develops U-Net based machine learning models that take ABI imagery as inputs to estimate turbulence intensity at three vertical levels: 10–18, 18–24, and above 24 kft (1 kft ≈ 300 m). Among six different U-Net-based models, U-Net3+ model with a filter size of three showed the best performance against the pilot report (PIREP). Two case studies are presented to show the strengths and weaknesses of the U-Net3+ model. The results tend to be overestimated above 24 kft, but estimates of 10–18 and 18–24 kft agree well with the PIREP, especially near convective regions. Significance Statement Turbulence is directly related to aviation safety as well as cost-effective aircraft operation. To avoid turbulence, turbulence diagnostics are calculated from numerical weather prediction (NWP) model outputs and are provided to pilots. The goal of this study is to develop a satellite data–driven machine learning model that estimates turbulence intensity in three different vertical layers to provide additional information along with the NWP-based turbulence diagnostics. Validation results against pilot reports show that the machine learning model performs comparable to NWP-based turbulence diagnostics. Furthermore, results with different channel selections reveal that using multiple water vapor channels can help extract additional information for estimating turbulence intensity at lower levels.

Integrating LEO and GEO Observations: Toward Optimal Summertime Satellite Precipitation Retrieval

Abstract Reliable quantitative precipitation estimation with a rich spatiotemporal resolution is vital for understanding the Earth’s hydrological cycle. Precipitation estimation over land and coastal regions is necessary for addressing the high degree of spatial heterogeneity of water availability and demand, and for resolving the extremes that modulate and amplify hazards such as flooding and landslides. Advancements in computation power along with unique high spatiotemporal and spectral resolution data streams from passive meteorological sensors aboard geosynchronous Earth-orbiting (GEO) and low Earth-orbiting (LEO) satellites offer exciting opportunities to retrieve information about surface precipitation phenomena using data-driven machine learning techniques. In this study, the capabilities of U-Net–like architecture are investigated to map instantaneous, summertime surface precipitation intensity at the spatial resolution of 2 km. The calibrated brightness temperature products from the Global Precipitation Measurement (GPM) Microwave Imager (GMI) radiometer are combined with multispectral images (visible, near-infrared, and infrared bands) from the Advanced Baseline Imager (ABI) aboard the GOES-R satellites as main inputs to the U-Net–like precipitation algorithm. Total precipitable water and 2-m temperature from the Global Forecast System (GFS) model are also used as auxiliary inputs to the model. The results show that the U-Net–like algorithm can capture fine-scale patterns and intensity of surface precipitation at high spatial resolution over stratiform and convective precipitation regimes. The evaluations reveal the potential of extracting relevant, high spatial features over complex surface types such as mountainous regions and coastlines. The algorithm allows users to interpret the inputs’ importance and can serve as a starting point for further exploration of precipitation systems within the field of hydrometeorology.

Forecast Applications of GLM Gridded Products: A Data Fusion Perspective

Abstract The recently deployed GOES-R series Geostationary Lightning Mapper (GLM) provides forecasters with a new, rapidly updating lightning data source to diagnose, forecast, and monitor atmospheric convection. Gridded GLM products have been developed to improve operational forecast applications, with variables including flash extent density (FED), minimum flash area (MFA), and total optical energy (TOE). While these gridded products have been evaluated, there is a continual need to integrate these products with other datasets available to forecasters such as radar, satellite imagery, and ground-based lightning networks. Data from the Advanced Baseline Imager (ABI), Multi-Radar Multi-Sensor (MRMS) system, and one ground-based lightning network were compared against gridded GLM imagery from GOES-East and GOES-West in case studies of two supercell thunderstorms, along with a bulk study from 13 April to 31 May 2019, to provide further validation and applications of gridded GLM products from a data fusion perspective. Increasing FED and decreasing MFA corresponded with increasing thunderstorm intensity from the perspective of ABI infrared imagery and MRMS vertically integrated reflectivity products, and was apparent for more robust and severe convection. Flash areas were also observed to maximize between clean-IR brightness temperatures of 210–230 K and isothermal reflectivity at −10°C of 20–30 dBZ. TOE observations from both GLMs provided additional context of local GLM flash rates in each case study, due to their differing perspectives of convective updrafts. Significance Statement The Geostationary Lightning Mapper (GLM) is a lightning sensor on the current generation of U.S. weather satellites. This research shows how data from the space-based lightning sensor can be combined with radar, satellite imagery, and ground-based lightning networks to improve how forecasters monitor thunderstorms and issue warnings for severe weather. The rate of GLM flashes detected and the area they cover correspond well with radar and satellite signatures, especially in cases of intense and severe thunderstorms. When the GLM observes the same thunderstorm from the GOES-East and GOES-West satellites, the optical energy (brightness) of the flashes may help forecasters interpret the types of flashes observed from each sensor.