Search Results
You are looking at 1 - 10 of 16 items for
- Author or Editor: Veljko Petkovic x
- Refine by Access: All Content x
Abstract
An updated version of the Goddard Profiling Algorithm (GPROF 2014) with a new overland scheme was released with the launch of the Global Precipitation Mission (GPM) core satellite in February 2014. The algorithm is designed to provide consistent precipitation estimates over both ocean and land across diverse satellite platforms. This study tests the performance of the new retrieval, focusing specifically on an extreme rainfall event. Two contrasting 72-h precipitation events over the same area are used to compare the retrieved products against ground measurements. The first event is characterized by persistent and intense precipitation of an unusually strong and widespread system, which caused historical flooding of the central Balkan region of southeastern Europe in May 2014. The second event serves as a baseline case for a more typical midlatitude regime. Rainfall rates and 3-day accumulations given by five conically scanning radiometers (GMI; AMSR2; and SSMIS F16, F17, and F18) in the GPM constellation are compared against ground radar data from the Operational Program for Exchange of Weather Radar Information (OPERA) network and in situ measurements. Satellite products show good agreement with ground radars; the retrieval closely reproduces spatial and temporal characteristics of both events. Strong biases related to precipitation regimes are found when satellite and radar measurements are compared to ground gauges. While the GPM constellation performs well during the nonextreme event, showing ~10% negative bias, it underestimates gauge accumulations of the Balkan flood event by 60%. Analyses show that the biases are caused by the differences between the expected and observed ice-scattering signals, suggesting that better understanding of the environment and its impact on rain profiles is the key for successful retrievals in extreme events.
Abstract
An updated version of the Goddard Profiling Algorithm (GPROF 2014) with a new overland scheme was released with the launch of the Global Precipitation Mission (GPM) core satellite in February 2014. The algorithm is designed to provide consistent precipitation estimates over both ocean and land across diverse satellite platforms. This study tests the performance of the new retrieval, focusing specifically on an extreme rainfall event. Two contrasting 72-h precipitation events over the same area are used to compare the retrieved products against ground measurements. The first event is characterized by persistent and intense precipitation of an unusually strong and widespread system, which caused historical flooding of the central Balkan region of southeastern Europe in May 2014. The second event serves as a baseline case for a more typical midlatitude regime. Rainfall rates and 3-day accumulations given by five conically scanning radiometers (GMI; AMSR2; and SSMIS F16, F17, and F18) in the GPM constellation are compared against ground radar data from the Operational Program for Exchange of Weather Radar Information (OPERA) network and in situ measurements. Satellite products show good agreement with ground radars; the retrieval closely reproduces spatial and temporal characteristics of both events. Strong biases related to precipitation regimes are found when satellite and radar measurements are compared to ground gauges. While the GPM constellation performs well during the nonextreme event, showing ~10% negative bias, it underestimates gauge accumulations of the Balkan flood event by 60%. Analyses show that the biases are caused by the differences between the expected and observed ice-scattering signals, suggesting that better understanding of the environment and its impact on rain profiles is the key for successful retrievals in extreme events.
Abstract
A spatiotemporal correlation technique has been developed to combine satellite rainfall measurements using the spatial and temporal correlation of the rainfall fields to overcome problems of limited and infrequent measurements while accounting for the measurement accuracies. The relationship between the temporal and spatial correlation of the rainfall field is exploited to provide information about rainfall beyond instantaneous measurements. The technique is developed using synthetic radar data. Nine months of Operational Program for the Exchange of Weather Radar (OPERA) data are used on grid sizes of 100, 248, and 500 km with pixel resolutions of 8, 12, and 24 km to simulate satellite fields of view and are then applied to the real satellite data over the Southwest to calculate 3-h rainfall accumulations. The results are compared with the simple averaging technique, which takes a simple mean of the measurements as a constant rainfall rate over the entire accumulation period. Using synthetic data, depending on the time separation of the measurements and their accuracy, a spatiotemporal correlation technique has shown the potential to yield improvements of up to 40% in absolute error and up to 25% in root-mean-square error when compared with the simple averaging technique. When applied to the real satellite data over the Southeast, the technique showed much less skill (general improvement of only 2%–6%).
Abstract
A spatiotemporal correlation technique has been developed to combine satellite rainfall measurements using the spatial and temporal correlation of the rainfall fields to overcome problems of limited and infrequent measurements while accounting for the measurement accuracies. The relationship between the temporal and spatial correlation of the rainfall field is exploited to provide information about rainfall beyond instantaneous measurements. The technique is developed using synthetic radar data. Nine months of Operational Program for the Exchange of Weather Radar (OPERA) data are used on grid sizes of 100, 248, and 500 km with pixel resolutions of 8, 12, and 24 km to simulate satellite fields of view and are then applied to the real satellite data over the Southwest to calculate 3-h rainfall accumulations. The results are compared with the simple averaging technique, which takes a simple mean of the measurements as a constant rainfall rate over the entire accumulation period. Using synthetic data, depending on the time separation of the measurements and their accuracy, a spatiotemporal correlation technique has shown the potential to yield improvements of up to 40% in absolute error and up to 25% in root-mean-square error when compared with the simple averaging technique. When applied to the real satellite data over the Southeast, the technique showed much less skill (general improvement of only 2%–6%).
Abstract
Analyses of the Tropical Rainfall Measuring Mission (TRMM) satellite rainfall estimates reveal a substantial disagreement between its active [Precipitation Radar (PR)] and passive [TRMM Microwave Imager (TMI)] sensors over certain regions. This study focuses on understanding the role of the synoptic state of atmosphere in these discrepancies over land regions where passive microwave (PMW) retrievals are limited to scattering signals. As such the variability in the relationship between the ice-induced scattering signal and the surface rainfall is examined. Using the Amazon River and central Africa regions as a test bed, it is found that the systematic difference seen between PR and TMI rainfall estimates is well correlated with both the precipitating system structure and the level of its organization. Relying on a clustering technique to group raining scenes into three broad but distinct organizational categories, it is found that, relative to the PR, deep-organized systems are typically overestimated by TMI while the shallower ones are underestimated. Results suggest that the storm organization level can explain up to 50% of the regional systematic difference between the two sensors. Because of its potential for retrieval improvement, the ability to forecast the level of systems organization is tested. The state of the atmosphere is found to favor certain storm types when constrained by CAPE, wind shear, dewpoint depression, and vertical humidity distribution. Among other findings, the observations reveal that the ratio between boundary layer and midtropospheric moisture correlates well with the organization level of convection. If adjusted by the observed PR-to-TMI ratio under a given environment, the differences between PMW and PR rainfall estimates are diminished, at maximum, by 30% in RMSE and by 40% in the mean.
Abstract
Analyses of the Tropical Rainfall Measuring Mission (TRMM) satellite rainfall estimates reveal a substantial disagreement between its active [Precipitation Radar (PR)] and passive [TRMM Microwave Imager (TMI)] sensors over certain regions. This study focuses on understanding the role of the synoptic state of atmosphere in these discrepancies over land regions where passive microwave (PMW) retrievals are limited to scattering signals. As such the variability in the relationship between the ice-induced scattering signal and the surface rainfall is examined. Using the Amazon River and central Africa regions as a test bed, it is found that the systematic difference seen between PR and TMI rainfall estimates is well correlated with both the precipitating system structure and the level of its organization. Relying on a clustering technique to group raining scenes into three broad but distinct organizational categories, it is found that, relative to the PR, deep-organized systems are typically overestimated by TMI while the shallower ones are underestimated. Results suggest that the storm organization level can explain up to 50% of the regional systematic difference between the two sensors. Because of its potential for retrieval improvement, the ability to forecast the level of systems organization is tested. The state of the atmosphere is found to favor certain storm types when constrained by CAPE, wind shear, dewpoint depression, and vertical humidity distribution. Among other findings, the observations reveal that the ratio between boundary layer and midtropospheric moisture correlates well with the organization level of convection. If adjusted by the observed PR-to-TMI ratio under a given environment, the differences between PMW and PR rainfall estimates are diminished, at maximum, by 30% in RMSE and by 40% in the mean.
Resolving Surface Rain from GMI High-Frequency Channels: Limits Imposed by the Three-Dimensional Structure of PrecipitationAbstract
The scattering of microwaves at frequencies between 50 and 200 GHz by ice particles in the atmosphere is an essential element in the retrieval of instantaneous surface precipitation from spaceborne passive radiometers. This paper explores how the variable distribution of solid and liquid hydrometeors in the atmospheric column over land surfaces affects the brightness temperature (TB) measured by GMI at 89 GHz through the analysis of Dual-Frequency Precipitation Radar (DPR) reflectivity profiles along the 89-GHz beam. The objective is to refine the statistical relations between observed TBs and surface precipitation over land and to define their limits. As GMI is scanning with a 53° Earth incident angle, the observed atmospheric volume is actually not a vertical column, which may lead to very heterogeneous and seemingly inconsistent distributions of the hydrometeors inside the beam. It is found that the 89-GHz TB is mostly sensitive to the presence of ice hydrometeors several kilometers above the 0°C isotherm, up to 10 km above the 0°C isotherm for the deepest convective systems, but is a modest predictor of the surface precipitation rate. To perform a precise mapping of atmospheric ice, the altitude of the individual ice clusters must be known. Indeed, if variations in the altitude of ice are not accounted for, then the high incident angle of GMI causes a horizontal shift (parallax shift) between the estimated position of the ice clusters and their actual position. We show here that the altitude of ice clusters can be derived from the 89-GHz TB itself, allowing for correction of the parallax shift.
Abstract
The scattering of microwaves at frequencies between 50 and 200 GHz by ice particles in the atmosphere is an essential element in the retrieval of instantaneous surface precipitation from spaceborne passive radiometers. This paper explores how the variable distribution of solid and liquid hydrometeors in the atmospheric column over land surfaces affects the brightness temperature (TB) measured by GMI at 89 GHz through the analysis of Dual-Frequency Precipitation Radar (DPR) reflectivity profiles along the 89-GHz beam. The objective is to refine the statistical relations between observed TBs and surface precipitation over land and to define their limits. As GMI is scanning with a 53° Earth incident angle, the observed atmospheric volume is actually not a vertical column, which may lead to very heterogeneous and seemingly inconsistent distributions of the hydrometeors inside the beam. It is found that the 89-GHz TB is mostly sensitive to the presence of ice hydrometeors several kilometers above the 0°C isotherm, up to 10 km above the 0°C isotherm for the deepest convective systems, but is a modest predictor of the surface precipitation rate. To perform a precise mapping of atmospheric ice, the altitude of the individual ice clusters must be known. Indeed, if variations in the altitude of ice are not accounted for, then the high incident angle of GMI causes a horizontal shift (parallax shift) between the estimated position of the ice clusters and their actual position. We show here that the altitude of ice clusters can be derived from the 89-GHz TB itself, allowing for correction of the parallax shift.
Enhancing PMW Satellite Precipitation Estimation: Detecting Convective ClassAbstract
A decades-long effort in observing precipitation from space has led to continuous improvements of satellite-derived passive microwave (PMW) large-scale precipitation products. However, due to a limited ability to relate observed radiometric signatures to precipitation type (convective and stratiform) and associated precipitation rate variability, PMW retrievals are prone to large systematic errors at instantaneous scales. The present study explores the use of deep learning approach in extracting the information content from PMW observation vectors to help identify precipitation types. A deep learning neural network model (DNN) is developed to retrieve the convective type in precipitating systems from PMW observations. A 12-month period of Global Precipitation Measurement mission Microwave Imager (GMI) observations is used as a dataset for model development and verification. The proposed DNN model is shown to accurately predict precipitation types for 85% of total precipitation volume. The model reduces precipitation rate bias associated with convective and stratiform precipitation in the GPM operational algorithm by a factor of 2 while preserving the correlation with reference precipitation rates, and is insensitive to surface type variability. Based on comparisons against currently used convective schemes, it is concluded that the neural network approach has the potential to address regime-specific PMW satellite precipitation biases affecting GPM operations.
Abstract
A decades-long effort in observing precipitation from space has led to continuous improvements of satellite-derived passive microwave (PMW) large-scale precipitation products. However, due to a limited ability to relate observed radiometric signatures to precipitation type (convective and stratiform) and associated precipitation rate variability, PMW retrievals are prone to large systematic errors at instantaneous scales. The present study explores the use of deep learning approach in extracting the information content from PMW observation vectors to help identify precipitation types. A deep learning neural network model (DNN) is developed to retrieve the convective type in precipitating systems from PMW observations. A 12-month period of Global Precipitation Measurement mission Microwave Imager (GMI) observations is used as a dataset for model development and verification. The proposed DNN model is shown to accurately predict precipitation types for 85% of total precipitation volume. The model reduces precipitation rate bias associated with convective and stratiform precipitation in the GPM operational algorithm by a factor of 2 while preserving the correlation with reference precipitation rates, and is insensitive to surface type variability. Based on comparisons against currently used convective schemes, it is concluded that the neural network approach has the potential to address regime-specific PMW satellite precipitation biases affecting GPM operations.
Abstract
The three-dimensional (3D) structure of precipitation systems is highly dependent on hydrometeor formation processes and microphysics. This study aims to characterize distinct vertical profiles of precipitation regimes by relying on the availability of a high-quality, spatially dense radar network and its capability to observe the 3D structure of the storms. A deep-learning-based framework, coupled with unsupervised clustering methods, is developed to identify types of precipitation structures irrespective of their physical properties. A 6-month period of 3D reflectivity profiles from the Multi-Radar Multi-Sensor (MRMS) network is used to identify different regimes and investigate their properties with respect to the underlying environmental conditions. Dominant features retrieved from radar reflectivity profiles using convolutional neural-network-based autoencoders are employed to identify similar-looking vertical structures using coupled k-means and agglomerative clustering algorithms. The k-means method identifies distinct groups, while the agglomerative clustering visualizes intercluster relationships. The framework identifies 18 clusters that can be broadly combined into five groups of varied echo-top heights. The 18 clusters demonstrate variability with respect to structural features and precipitation rate/type, implying that profiles in each group belong to a physically different precipitation regime. An independent analysis of the regime properties is conducted by matching the MRMS reflectivity profiles with environmental parameters derived from the High-Resolution Rapid Refresh model forecasts. The distribution of the environmental variables confirms cluster-specific feature properties, confirming the physics-based regime separation across the clusters and their dependence on the vertical structure. The identified precipitation regimes can assist in developing physics-guided retrievals and studying precipitation regimes.
Significance Statement
This study proposes a systematic model to identify precipitation profiles of distinct vertical structures and evaluate their dependence on environmental conditions. The model was developed using ground-based radar observations; however, there is potential to extend this model to reflectivity profiles from both ground- and satellite-based sensors. In addition, the identified precipitation regime clusters could be a proxy for the vertical structure of precipitation systems and assist in determining the structural variability within traditional precipitation type classification (e.g., convective versus stratiform). Moreover, identifying the precipitation regimes could also be used to improve satellite-based precipitation retrievals. Finally, a better understanding of precipitation structure would also help improve the initialization of climate models.
Abstract
The three-dimensional (3D) structure of precipitation systems is highly dependent on hydrometeor formation processes and microphysics. This study aims to characterize distinct vertical profiles of precipitation regimes by relying on the availability of a high-quality, spatially dense radar network and its capability to observe the 3D structure of the storms. A deep-learning-based framework, coupled with unsupervised clustering methods, is developed to identify types of precipitation structures irrespective of their physical properties. A 6-month period of 3D reflectivity profiles from the Multi-Radar Multi-Sensor (MRMS) network is used to identify different regimes and investigate their properties with respect to the underlying environmental conditions. Dominant features retrieved from radar reflectivity profiles using convolutional neural-network-based autoencoders are employed to identify similar-looking vertical structures using coupled k-means and agglomerative clustering algorithms. The k-means method identifies distinct groups, while the agglomerative clustering visualizes intercluster relationships. The framework identifies 18 clusters that can be broadly combined into five groups of varied echo-top heights. The 18 clusters demonstrate variability with respect to structural features and precipitation rate/type, implying that profiles in each group belong to a physically different precipitation regime. An independent analysis of the regime properties is conducted by matching the MRMS reflectivity profiles with environmental parameters derived from the High-Resolution Rapid Refresh model forecasts. The distribution of the environmental variables confirms cluster-specific feature properties, confirming the physics-based regime separation across the clusters and their dependence on the vertical structure. The identified precipitation regimes can assist in developing physics-guided retrievals and studying precipitation regimes.
Significance Statement
This study proposes a systematic model to identify precipitation profiles of distinct vertical structures and evaluate their dependence on environmental conditions. The model was developed using ground-based radar observations; however, there is potential to extend this model to reflectivity profiles from both ground- and satellite-based sensors. In addition, the identified precipitation regime clusters could be a proxy for the vertical structure of precipitation systems and assist in determining the structural variability within traditional precipitation type classification (e.g., convective versus stratiform). Moreover, identifying the precipitation regimes could also be used to improve satellite-based precipitation retrievals. Finally, a better understanding of precipitation structure would also help improve the initialization of climate models.
Abstract
Prominent achievements made in addressing global precipitation using satellite passive microwave retrievals are often overshadowed by their performance at finer spatial and temporal scales, where large variability in cloud morphology poses an obstacle for accurate precipitation measurements. This is especially true over land, with precipitation estimates being based on an observed mean relationship between high-frequency (e.g., 89 GHz) brightness temperature depression (i.e., the ice-scattering signature) and surface precipitation rate. This indirect relationship between the observed (brightness temperatures) and state (precipitation) vectors often leads to inaccurate estimates, with more pronounced biases (e.g., −30% over the United States) observed during extreme events. This study seeks to mitigate these errors by employing previously established relationships between cloud structures and large-scale environments such as CAPE, wind shear, humidity distribution, and aerosol concentrations to form a stronger relationship between precipitation and the scattering signal. The GPM passive microwave operational precipitation retrieval (GPROF) for the GMI sensor is modified to offer additional information on atmospheric conditions to its Bayesian-based algorithm. The modified algorithm is allowed to use the large-scale environment to filter out a priori states that do not match the general synoptic condition relevant to the observation and thus reduces the difference between the assumed and observed variability in the ice-to-rain ratio. Using the ground Multi-Radar Multi-Sensor (MRMS) network over the United States, the results demonstrate outstanding potential in improving the accuracy of heavy precipitation over land. It is found that individual synoptic parameters can remove 20%–30% of existing bias and up to 50% when combined, while preserving the overall performance of the algorithm.
Abstract
Prominent achievements made in addressing global precipitation using satellite passive microwave retrievals are often overshadowed by their performance at finer spatial and temporal scales, where large variability in cloud morphology poses an obstacle for accurate precipitation measurements. This is especially true over land, with precipitation estimates being based on an observed mean relationship between high-frequency (e.g., 89 GHz) brightness temperature depression (i.e., the ice-scattering signature) and surface precipitation rate. This indirect relationship between the observed (brightness temperatures) and state (precipitation) vectors often leads to inaccurate estimates, with more pronounced biases (e.g., −30% over the United States) observed during extreme events. This study seeks to mitigate these errors by employing previously established relationships between cloud structures and large-scale environments such as CAPE, wind shear, humidity distribution, and aerosol concentrations to form a stronger relationship between precipitation and the scattering signal. The GPM passive microwave operational precipitation retrieval (GPROF) for the GMI sensor is modified to offer additional information on atmospheric conditions to its Bayesian-based algorithm. The modified algorithm is allowed to use the large-scale environment to filter out a priori states that do not match the general synoptic condition relevant to the observation and thus reduces the difference between the assumed and observed variability in the ice-to-rain ratio. Using the ground Multi-Radar Multi-Sensor (MRMS) network over the United States, the results demonstrate outstanding potential in improving the accuracy of heavy precipitation over land. It is found that individual synoptic parameters can remove 20%–30% of existing bias and up to 50% when combined, while preserving the overall performance of the algorithm.
Abstract
Spaceborne microwave radiometers represent an important component of the Global Precipitation Measurement (GPM) mission due to their frequent sampling of rain systems. Microwave radiometers measure microwave radiation (brightness temperatures Tb), which can be converted into precipitation estimates with appropriate assumptions. However, detecting shallow precipitation systems using spaceborne radiometers is challenging, especially over land, as their weak signals are hard to differentiate from those associated with dry conditions. This study uses a random forest (RF) model to classify microwave radiometer observations as dry, shallow, or nonshallow over the Netherlands—a region with varying surface conditions and frequent occurrence of shallow precipitation. The RF model is trained on five years of data (2016–20) and tested with two independent years (2015 and 2021). The observations are classified using ground-based weather radar echo top heights. Various RF models are assessed, such as using only GPM Microwave Imager (GMI) Tb values as input features or including spatially aligned ERA5 2-m temperature and freezing level reanalysis and/or Dual-Frequency Precipitation Radar (DPR) observations. Independent of the input features, the model performs best in summer and worst in winter. The model classifies observations from high-frequency channels (≥85 GHz) with lower Tb values as nonshallow, higher values as dry, and those in between as shallow. Misclassified footprints exhibit radiometric characteristics corresponding to their assigned class. Case studies reveal dry observations misclassified as shallow are associated with lower Tb values, likely resulting from the presence of ice particles in nonprecipitating clouds. Shallow footprints misclassified as dry are likely related to the absence of ice particles.
Significance Statement
Published research concerning rainfall retrieval algorithms from microwave radiometers is often focused on the accuracy of these algorithms. While shallow precipitation over land is often characterized as problematic in these studies, little progress has been made with these systems. In particular, precipitation formed by shallow clouds, where shallow refers to the clouds being close to Earth’s surface, is often missed. This study is focused on detecting shallow precipitation and its physical characteristics to further improve its detection from spaceborne sensors. As such, it contributes to understanding which shallow precipitation scenes are challenging to detect from microwave radiometers, suggesting possible ways for algorithm improvement.
Abstract
Spaceborne microwave radiometers represent an important component of the Global Precipitation Measurement (GPM) mission due to their frequent sampling of rain systems. Microwave radiometers measure microwave radiation (brightness temperatures Tb), which can be converted into precipitation estimates with appropriate assumptions. However, detecting shallow precipitation systems using spaceborne radiometers is challenging, especially over land, as their weak signals are hard to differentiate from those associated with dry conditions. This study uses a random forest (RF) model to classify microwave radiometer observations as dry, shallow, or nonshallow over the Netherlands—a region with varying surface conditions and frequent occurrence of shallow precipitation. The RF model is trained on five years of data (2016–20) and tested with two independent years (2015 and 2021). The observations are classified using ground-based weather radar echo top heights. Various RF models are assessed, such as using only GPM Microwave Imager (GMI) Tb values as input features or including spatially aligned ERA5 2-m temperature and freezing level reanalysis and/or Dual-Frequency Precipitation Radar (DPR) observations. Independent of the input features, the model performs best in summer and worst in winter. The model classifies observations from high-frequency channels (≥85 GHz) with lower Tb values as nonshallow, higher values as dry, and those in between as shallow. Misclassified footprints exhibit radiometric characteristics corresponding to their assigned class. Case studies reveal dry observations misclassified as shallow are associated with lower Tb values, likely resulting from the presence of ice particles in nonprecipitating clouds. Shallow footprints misclassified as dry are likely related to the absence of ice particles.
Significance Statement
Published research concerning rainfall retrieval algorithms from microwave radiometers is often focused on the accuracy of these algorithms. While shallow precipitation over land is often characterized as problematic in these studies, little progress has been made with these systems. In particular, precipitation formed by shallow clouds, where shallow refers to the clouds being close to Earth’s surface, is often missed. This study is focused on detecting shallow precipitation and its physical characteristics to further improve its detection from spaceborne sensors. As such, it contributes to understanding which shallow precipitation scenes are challenging to detect from microwave radiometers, suggesting possible ways for algorithm improvement.
Abstract
Visible and infrared radiance products of geostationary orbiting platforms provide virtually continuous observations of Earth. In contrast, low-Earth orbiters observe passive microwave (PMW) radiances at any location much less frequently. Prior literature demonstrates the ability of a machine learning (ML) approach to build a link between these two complementary radiance spectra by predicting PMW observations using infrared and visible products collected from geostationary instruments, which could potentially deliver a highly desirable synthetic PMW product with nearly continuous spatiotemporal coverage. However, current ML models lack the ability to provide a measure of uncertainty of such a product, significantly limiting its applications. In this work, Bayesian deep learning is employed to generate synthetic Global Precipitation Measurement (GPM) Microwave Imager (GMI) data from Advanced Baseline Imager (ABI) observations with attached uncertainties over the ocean. The study first uses deterministic residual networks (ResNets) to generate synthetic GMI brightness temperatures with as little mean absolute error as 1.72 K at the ABI spatiotemporal resolution. Then, for the same task, we use three Bayesian ResNet models to produce a comparable amount of error while providing previously unavailable predictive variance (i.e., uncertainty) for each synthetic data point. We find that the Flipout configuration provides the most robust calibration between uncertainty and error across GMI frequencies, and then demonstrate how this additional information is useful for discarding high-error synthetic data points prior to use by downstream applications.
Abstract
Visible and infrared radiance products of geostationary orbiting platforms provide virtually continuous observations of Earth. In contrast, low-Earth orbiters observe passive microwave (PMW) radiances at any location much less frequently. Prior literature demonstrates the ability of a machine learning (ML) approach to build a link between these two complementary radiance spectra by predicting PMW observations using infrared and visible products collected from geostationary instruments, which could potentially deliver a highly desirable synthetic PMW product with nearly continuous spatiotemporal coverage. However, current ML models lack the ability to provide a measure of uncertainty of such a product, significantly limiting its applications. In this work, Bayesian deep learning is employed to generate synthetic Global Precipitation Measurement (GPM) Microwave Imager (GMI) data from Advanced Baseline Imager (ABI) observations with attached uncertainties over the ocean. The study first uses deterministic residual networks (ResNets) to generate synthetic GMI brightness temperatures with as little mean absolute error as 1.72 K at the ABI spatiotemporal resolution. Then, for the same task, we use three Bayesian ResNet models to produce a comparable amount of error while providing previously unavailable predictive variance (i.e., uncertainty) for each synthetic data point. We find that the Flipout configuration provides the most robust calibration between uncertainty and error across GMI frequencies, and then demonstrate how this additional information is useful for discarding high-error synthetic data points prior to use by downstream applications.
Abstract
This study assesses the level-2 precipitation estimates from 10 radiometers relative to Global Precipitation Measurement (GPM) Ku-band precipitation radar (KuPR) in two parts. First, nine sensors—four imagers [Advanced Microwave Scanning Radiometer 2 (AMSR2) and three Special Sensor Microwave Imager/Sounders (SSMISs)] and five sounders [Advanced Technology Microwave Sounder (ATMS) and four Microwave Humidity Sounders (MHSs)]—are evaluated over the 65°S–65°N region. Over ocean, imagers outperform sounders, primarily due to the usage of low-frequency channels. Furthermore, AMSR2 is clearly superior to SSMISs, likely due to the finer footprint size. Over land all sensors perform similarly except the noticeably worse performance from ATMS and SSMIS-F17. Second, we include the Sondeur Atmospherique du Profil d’Humidite Intertropicale par Radiometrie (SAPHIR) into the evaluation process, contrasting it against other sensors in the SAPHIR latitudes (30°S–30°N). SAPHIR has a slightly worse detection capability than other sounders over ocean but comparable detection performance to MHSs over land. The intensity estimates from SAPHIR show a larger normalized root-mean-square-error over both land and ocean, likely because only 183.3-GHz channels are available. Currently, imagers are preferred to sounders when level-2 estimates are incorporated into level-3 products. Our results suggest a sensor-specific priority order. Over ocean, this study indicates a priority order of AMSR2, SSMISs, MHSs and ATMS, and SAPHIR. Over land, SSMIS-F17, ATMS and SAPHIR should be given a lower priority than the other sensors.
Abstract
This study assesses the level-2 precipitation estimates from 10 radiometers relative to Global Precipitation Measurement (GPM) Ku-band precipitation radar (KuPR) in two parts. First, nine sensors—four imagers [Advanced Microwave Scanning Radiometer 2 (AMSR2) and three Special Sensor Microwave Imager/Sounders (SSMISs)] and five sounders [Advanced Technology Microwave Sounder (ATMS) and four Microwave Humidity Sounders (MHSs)]—are evaluated over the 65°S–65°N region. Over ocean, imagers outperform sounders, primarily due to the usage of low-frequency channels. Furthermore, AMSR2 is clearly superior to SSMISs, likely due to the finer footprint size. Over land all sensors perform similarly except the noticeably worse performance from ATMS and SSMIS-F17. Second, we include the Sondeur Atmospherique du Profil d’Humidite Intertropicale par Radiometrie (SAPHIR) into the evaluation process, contrasting it against other sensors in the SAPHIR latitudes (30°S–30°N). SAPHIR has a slightly worse detection capability than other sounders over ocean but comparable detection performance to MHSs over land. The intensity estimates from SAPHIR show a larger normalized root-mean-square-error over both land and ocean, likely because only 183.3-GHz channels are available. Currently, imagers are preferred to sounders when level-2 estimates are incorporated into level-3 products. Our results suggest a sensor-specific priority order. Over ocean, this study indicates a priority order of AMSR2, SSMISs, MHSs and ATMS, and SAPHIR. Over land, SSMIS-F17, ATMS and SAPHIR should be given a lower priority than the other sensors.
