Search Results
You are looking at 1 - 10 of 45 items for
- Author or Editor: Pierre Kirstetter x
- Refine by Access: Content accessible to me x
Abstract
High temporal and spatial resolution observations of precipitation occurrence from the NEXRAD-based Multi-Radar Multi-Sensor (MRMS) system are compared to matched observations from CloudSat for 3 years over the contiguous United States (CONUS). Across the CONUS, precipitation is generally reported more frequently by CloudSat (7.8%) than by MRMS (6.3%), with dependence on factors such as the NEXRAD beam height, the near-surface air temperature, and the surface elevation. There is general agreement between ground-based and satellite-derived precipitation events over flat surfaces, especially in widespread precipitation events and when the NEXRAD beam heights are low. Within 100 km of the nearest NEXRAD site, MRMS reports a precipitation frequency of 7.54% while CloudSat reports 7.38%. However, further inspection reveals offsetting biases between the products, where CloudSat reports more snow and MRMS reports more rain. The magnitudes of these discrepancies correlate with elevation, but they are observed in both the complex terrain of the Rocky Mountains and the relatively flat midwestern areas of the CONUS. The findings advocate for caution when using MRMS frequency and accumulations in complex terrain, when temperatures are below freezing, and at ranges greater than 100 km. A multiresolution analysis shows that no more than 1.88% of CloudSat pixels over flat terrain are incorrectly identified as nonprecipitating as a result of shallow showers residing the CloudSat clutter-filled blind zone when near-surface air temperatures are above 15°C.
Abstract
High temporal and spatial resolution observations of precipitation occurrence from the NEXRAD-based Multi-Radar Multi-Sensor (MRMS) system are compared to matched observations from CloudSat for 3 years over the contiguous United States (CONUS). Across the CONUS, precipitation is generally reported more frequently by CloudSat (7.8%) than by MRMS (6.3%), with dependence on factors such as the NEXRAD beam height, the near-surface air temperature, and the surface elevation. There is general agreement between ground-based and satellite-derived precipitation events over flat surfaces, especially in widespread precipitation events and when the NEXRAD beam heights are low. Within 100 km of the nearest NEXRAD site, MRMS reports a precipitation frequency of 7.54% while CloudSat reports 7.38%. However, further inspection reveals offsetting biases between the products, where CloudSat reports more snow and MRMS reports more rain. The magnitudes of these discrepancies correlate with elevation, but they are observed in both the complex terrain of the Rocky Mountains and the relatively flat midwestern areas of the CONUS. The findings advocate for caution when using MRMS frequency and accumulations in complex terrain, when temperatures are below freezing, and at ranges greater than 100 km. A multiresolution analysis shows that no more than 1.88% of CloudSat pixels over flat terrain are incorrectly identified as nonprecipitating as a result of shallow showers residing the CloudSat clutter-filled blind zone when near-surface air temperatures are above 15°C.
Abstract
Precipitation retrievals from passive microwave satellite observations form the basis of many widely used precipitation products, but the performance of the retrievals depends on numerous factors such as surface type and precipitation variability. Previous evaluation efforts have identified bias dependence on precipitation regime, which may reflect the influence on retrievals of recurring factors. In this study, the concept of a regime-based evaluation of precipitation from the Goddard profiling (GPROF) algorithm is extended to cloud regimes. Specifically, GPROF V05 precipitation retrievals under four different cloud regimes are evaluated against ground radars over the United States. GPROF is generally able to accurately retrieve the precipitation associated with both organized convection and less organized storms, which collectively produce a substantial fraction of global precipitation. However, precipitation from stratocumulus systems is underestimated over land and overestimated over water. Similarly, precipitation associated with trade cumulus environments is underestimated over land, while biases over water depend on the sensor’s channel configuration. By extending the evaluation to more sensors and suppressed environments, these results complement insights previously obtained from precipitation regimes, thus demonstrating the potential of cloud regimes in categorizing the global atmosphere into discrete systems.
Significance Statement
To understand how the accuracy of satellite precipitation depends on weather conditions, we compare the satellite estimates of precipitation against ground radars in the United States, using cloud regimes as a proxy for different recurring atmospheric systems. Consistent with previous studies, we found that errors in the satellite precipitation vary under different regimes. Satellite precipitation is, reassuringly, more accurate for storm systems that produce intense precipitation. However, in systems that produce weak or isolated precipitation, the errors are larger due to retrieval limitations. These findings highlight the important role of atmospheric states on the accuracy of satellite precipitation and the potential of cloud regimes for categorizing the global atmosphere.
Abstract
Precipitation retrievals from passive microwave satellite observations form the basis of many widely used precipitation products, but the performance of the retrievals depends on numerous factors such as surface type and precipitation variability. Previous evaluation efforts have identified bias dependence on precipitation regime, which may reflect the influence on retrievals of recurring factors. In this study, the concept of a regime-based evaluation of precipitation from the Goddard profiling (GPROF) algorithm is extended to cloud regimes. Specifically, GPROF V05 precipitation retrievals under four different cloud regimes are evaluated against ground radars over the United States. GPROF is generally able to accurately retrieve the precipitation associated with both organized convection and less organized storms, which collectively produce a substantial fraction of global precipitation. However, precipitation from stratocumulus systems is underestimated over land and overestimated over water. Similarly, precipitation associated with trade cumulus environments is underestimated over land, while biases over water depend on the sensor’s channel configuration. By extending the evaluation to more sensors and suppressed environments, these results complement insights previously obtained from precipitation regimes, thus demonstrating the potential of cloud regimes in categorizing the global atmosphere into discrete systems.
Significance Statement
To understand how the accuracy of satellite precipitation depends on weather conditions, we compare the satellite estimates of precipitation against ground radars in the United States, using cloud regimes as a proxy for different recurring atmospheric systems. Consistent with previous studies, we found that errors in the satellite precipitation vary under different regimes. Satellite precipitation is, reassuringly, more accurate for storm systems that produce intense precipitation. However, in systems that produce weak or isolated precipitation, the errors are larger due to retrieval limitations. These findings highlight the important role of atmospheric states on the accuracy of satellite precipitation and the potential of cloud regimes for categorizing the global atmosphere.
Abstract
The Integrated Multisatellite Retrievals for GPM (IMERG), a global high-resolution gridded precipitation dataset, will enable a wide range of applications, ranging from studies on precipitation characteristics to applications in hydrology to evaluation of weather and climate models. These applications focus on different spatial and temporal scales and thus average the precipitation estimates to coarser resolutions. Such a modification of scale will impact the reliability of IMERG. In this study, the performance of the Final Run of IMERG is evaluated against ground-based measurements as a function of increasing spatial resolution (from 0.1° to 2.5°) and accumulation periods (from 0.5 to 24 h) over a region in the southeastern United States. For ground reference, a product derived from the Multi-Radar/Multi-Sensor suite, a radar- and gauge-based operational precipitation dataset, is used. The TRMM Multisatellite Precipitation Analysis (TMPA) is also included as a benchmark. In general, both IMERG and TMPA improve when scaled up to larger areas and longer time periods, with better identification of rain occurrences and consistent improvements in systematic and random errors of rain rates. Between the two satellite estimates, IMERG is slightly better than TMPA most of the time. These results will inform users on the reliability of IMERG over the scales relevant to their studies.
Abstract
The Integrated Multisatellite Retrievals for GPM (IMERG), a global high-resolution gridded precipitation dataset, will enable a wide range of applications, ranging from studies on precipitation characteristics to applications in hydrology to evaluation of weather and climate models. These applications focus on different spatial and temporal scales and thus average the precipitation estimates to coarser resolutions. Such a modification of scale will impact the reliability of IMERG. In this study, the performance of the Final Run of IMERG is evaluated against ground-based measurements as a function of increasing spatial resolution (from 0.1° to 2.5°) and accumulation periods (from 0.5 to 24 h) over a region in the southeastern United States. For ground reference, a product derived from the Multi-Radar/Multi-Sensor suite, a radar- and gauge-based operational precipitation dataset, is used. The TRMM Multisatellite Precipitation Analysis (TMPA) is also included as a benchmark. In general, both IMERG and TMPA improve when scaled up to larger areas and longer time periods, with better identification of rain occurrences and consistent improvements in systematic and random errors of rain rates. Between the two satellite estimates, IMERG is slightly better than TMPA most of the time. These results will inform users on the reliability of IMERG over the scales relevant to their studies.
Abstract
Satellite-based precipitation estimates are a vital resource for hydrologic applications in data-sparse regions of the world, particularly at daily or longer time scales. With the launch of a new generation of high-resolution imagers on geostationary platforms such as the Geostationary Operational Environmental Satellite series R (GOES-R), an opportunity exists to advance the detection and estimation of flash-flood-scale precipitation events from space beyond what is currently available. Because visible and infrared sensors can only observe cloud-top properties, many visible- and infrared-band-based rainfall algorithms attempt to first classify clouds before deriving a rain rate. This study uses a 2-yr database of cloud-top properties from proxy Advanced Baseline Imager radiances from GOES-R matched to surface precipitation types from the Multi-Radar Multi-Sensor (MRMS) system to develop a naïve Bayesian precipitation type classifier for the four major types of precipitation in MRMS: stratiform, convective, tropical, and hail. Evaluation of the naïve Bayesian precipitation type product showed a bias toward classifying convective and stratiform at the expense of tropical and hail. The tropical and hail classes in MRMS are derived based on the vertical structure and magnitude of radar reflectivity, which may not translate to an obvious signal at cloud top for a satellite-based algorithm. However, the satellite-based product correctly classified the hail areas as being convective in nature for the vast majority of missed hail events.
Abstract
Satellite-based precipitation estimates are a vital resource for hydrologic applications in data-sparse regions of the world, particularly at daily or longer time scales. With the launch of a new generation of high-resolution imagers on geostationary platforms such as the Geostationary Operational Environmental Satellite series R (GOES-R), an opportunity exists to advance the detection and estimation of flash-flood-scale precipitation events from space beyond what is currently available. Because visible and infrared sensors can only observe cloud-top properties, many visible- and infrared-band-based rainfall algorithms attempt to first classify clouds before deriving a rain rate. This study uses a 2-yr database of cloud-top properties from proxy Advanced Baseline Imager radiances from GOES-R matched to surface precipitation types from the Multi-Radar Multi-Sensor (MRMS) system to develop a naïve Bayesian precipitation type classifier for the four major types of precipitation in MRMS: stratiform, convective, tropical, and hail. Evaluation of the naïve Bayesian precipitation type product showed a bias toward classifying convective and stratiform at the expense of tropical and hail. The tropical and hail classes in MRMS are derived based on the vertical structure and magnitude of radar reflectivity, which may not translate to an obvious signal at cloud top for a satellite-based algorithm. However, the satellite-based product correctly classified the hail areas as being convective in nature for the vast majority of missed hail events.
Abstract
Nonuniform beam filling associated with the vertical variation of atmospheric reflectivity is an important source of error in the estimation of rainfall rates by radar. It is, however, possible to correct for this error if the vertical profile of reflectivity (VPR) is known. This paper presents a method for identifying VPRs from volumetric radar data. The method aims at improving an existing algorithm based on the analysis of ratios of radar measurements at multiple elevation angles. By adding a rainfall classification procedure defining more homogeneous precipitation patterns, the issue of VPR homogeneity is specifically addressed. The method is assessed using the dataset from a volume-scanning strategy for radar quantitative precipitation estimation designed in 2002 for the Bollène radar (France). The identified VPR is more representative of the rain field than are other estimated VPRs. It has also a positive impact on radar data processing for precipitation estimation: while scatter remains unchanged, an overall bias reduction at all time steps is noticed (up to 6% for all events) whereas performance varies with type of events considered (mesoscale convective systems, cold fronts, or shallow convection) according to the radar-observation conditions. This is attributed to the better processing of spatial variations of the vertical profile of reflectivity for the stratiform regions. However, adaptation of the VPR identification in the difficult radar measurement context in mountainous areas and to the rainfall classification procedure proved challenging because of data fluctuations.
Abstract
Nonuniform beam filling associated with the vertical variation of atmospheric reflectivity is an important source of error in the estimation of rainfall rates by radar. It is, however, possible to correct for this error if the vertical profile of reflectivity (VPR) is known. This paper presents a method for identifying VPRs from volumetric radar data. The method aims at improving an existing algorithm based on the analysis of ratios of radar measurements at multiple elevation angles. By adding a rainfall classification procedure defining more homogeneous precipitation patterns, the issue of VPR homogeneity is specifically addressed. The method is assessed using the dataset from a volume-scanning strategy for radar quantitative precipitation estimation designed in 2002 for the Bollène radar (France). The identified VPR is more representative of the rain field than are other estimated VPRs. It has also a positive impact on radar data processing for precipitation estimation: while scatter remains unchanged, an overall bias reduction at all time steps is noticed (up to 6% for all events) whereas performance varies with type of events considered (mesoscale convective systems, cold fronts, or shallow convection) according to the radar-observation conditions. This is attributed to the better processing of spatial variations of the vertical profile of reflectivity for the stratiform regions. However, adaptation of the VPR identification in the difficult radar measurement context in mountainous areas and to the rainfall classification procedure proved challenging because of data fluctuations.
Abstract
The vertical profile of reflectivity (VPR) must be identified to correct estimations of rainfall rates by radar for the nonuniform beam filling associated with the vertical variation of radar reflectivity. A method for identifying VPRs from volumetric radar data is presented that takes into account the radar sampling. Physically based constraints on the vertical structure of rainfall are introduced with simple VPR models within a rainfall classification procedure defining more homogeneous precipitation patterns. The model parameters are identified in the framework of an extended Kalman filter to ensure their temporal consistency. The method is assessed using the dataset from a volume-scanning strategy for radar quantitative precipitation estimation designed in 2002 for the Bollène radar (France). The physical consistency of the retrieved VPR is evaluated. Positive results are obtained insofar as the physically based identified VPR (i) presents physically consistent shapes and characteristics considering beam effects, (ii) shows improved robustness in the difficult radar measurement context of the Cévennes–Vivarais region, and (iii) provides consistent physical insight into the rain field.
Abstract
The vertical profile of reflectivity (VPR) must be identified to correct estimations of rainfall rates by radar for the nonuniform beam filling associated with the vertical variation of radar reflectivity. A method for identifying VPRs from volumetric radar data is presented that takes into account the radar sampling. Physically based constraints on the vertical structure of rainfall are introduced with simple VPR models within a rainfall classification procedure defining more homogeneous precipitation patterns. The model parameters are identified in the framework of an extended Kalman filter to ensure their temporal consistency. The method is assessed using the dataset from a volume-scanning strategy for radar quantitative precipitation estimation designed in 2002 for the Bollène radar (France). The physical consistency of the retrieved VPR is evaluated. Positive results are obtained insofar as the physically based identified VPR (i) presents physically consistent shapes and characteristics considering beam effects, (ii) shows improved robustness in the difficult radar measurement context of the Cévennes–Vivarais region, and (iii) provides consistent physical insight into the rain field.
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
Evidence of multiple-scattering-induced pulse stretching for the signal of both frequencies of the Dual-Frequency Precipitation Radar (DPR) on the Global Precipitation Measurement (GPM) mission Core Observatory satellite is presented on the basis of collocated ground-based WSR-88D S-band observations of an extreme case: a tornadic supercell. The ground-based observations clearly show a tilted convective core with a so-called bounded weak-echo region—that is, locations where precipitation is absent or extremely light at the ground while large amounts of liquid or frozen precipitation are present aloft. The satellite observations in this region show reflectivity profiles that extend all the way to the surface despite the absence of near-surface precipitation: these are here referred to as “ghost echoes.” Furthermore, the Ku- and Ka-band profiles exhibit similar slopes, which is a typical sign that the observed power is almost entirely due to multiple scattering. A novel microphysical retrieval that is based on triple-frequency (S–Ku–Ka) observations shows that a dense ice core located between 4 and 14 km with particle sizes exceeding 2.5 cm and integrated ice contents exceeding 7.0 kg m−2 is the source of the ghost echoes of the signal in the lower layers. The level of confidence of this assessment is strengthened by the availability of the S-band data, which provide the necessary additional constraints to the radar retrieval that is based on DPR data. This study shows not only that multiple-scattering contributions may become predominant at Ka already very high up in the atmosphere but also that they play a key role at Ku band within the layers close to the surface. As a result, extreme caution must be paid even in the interpretation of Ku-based retrievals (e.g., the TRMM PR dataset or any DPR retrievals that are based on the assumption that Ku band is not affected by multiple scattering) when examining extreme surface rain rates that occur in the presence of deep dense ice layers.
Abstract
Evidence of multiple-scattering-induced pulse stretching for the signal of both frequencies of the Dual-Frequency Precipitation Radar (DPR) on the Global Precipitation Measurement (GPM) mission Core Observatory satellite is presented on the basis of collocated ground-based WSR-88D S-band observations of an extreme case: a tornadic supercell. The ground-based observations clearly show a tilted convective core with a so-called bounded weak-echo region—that is, locations where precipitation is absent or extremely light at the ground while large amounts of liquid or frozen precipitation are present aloft. The satellite observations in this region show reflectivity profiles that extend all the way to the surface despite the absence of near-surface precipitation: these are here referred to as “ghost echoes.” Furthermore, the Ku- and Ka-band profiles exhibit similar slopes, which is a typical sign that the observed power is almost entirely due to multiple scattering. A novel microphysical retrieval that is based on triple-frequency (S–Ku–Ka) observations shows that a dense ice core located between 4 and 14 km with particle sizes exceeding 2.5 cm and integrated ice contents exceeding 7.0 kg m−2 is the source of the ghost echoes of the signal in the lower layers. The level of confidence of this assessment is strengthened by the availability of the S-band data, which provide the necessary additional constraints to the radar retrieval that is based on DPR data. This study shows not only that multiple-scattering contributions may become predominant at Ka already very high up in the atmosphere but also that they play a key role at Ku band within the layers close to the surface. As a result, extreme caution must be paid even in the interpretation of Ku-based retrievals (e.g., the TRMM PR dataset or any DPR retrievals that are based on the assumption that Ku band is not affected by multiple scattering) when examining extreme surface rain rates that occur in the presence of deep dense ice layers.
Abstract
In the hydrological sciences, the outstanding challenge of regional modeling requires to capture common and event-specific hydrologic behaviors driven by rainfall spatial variability and catchment physiography during floods. The overall objective of this study is to develop robust understanding and predictive capability of how rainfall spatial variability influences flood peak discharge relative to basin physiography. A machine-learning approach is used on a high-resolution dataset of rainfall and flooding events spanning 10 years, with rainfall events and basins of widely varying characteristics selected across the continental United States. It overcomes major limitations in prior studies that were based on limited observations or hydrological model simulations. This study explores first-order dependencies in the relationships between peak discharge, rainfall variability, and basin physiography, and it sheds light on these complex interactions using a multidimensional statistical modeling approach. Among different machine-learning techniques, XGBoost is used to determine the significant physiographical and rainfall characteristics that influence peak discharge through variable importance analysis. A parsimonious model with low bias and variance is created that can be deployed in the future for flash flood forecasting. The results confirm that, although the spatial organization of rainfall within a basin has a major influence on basin response, basin physiography is the primary driver of peak discharge. These findings have unprecedented spatial and temporal representativeness in terms of flood characterization across basins. An improved understanding of subbasin scale rainfall spatial variability will aid in robust flash flood characterization as well as with identifying basins that could most benefit from distributed hydrologic modeling.
Abstract
In the hydrological sciences, the outstanding challenge of regional modeling requires to capture common and event-specific hydrologic behaviors driven by rainfall spatial variability and catchment physiography during floods. The overall objective of this study is to develop robust understanding and predictive capability of how rainfall spatial variability influences flood peak discharge relative to basin physiography. A machine-learning approach is used on a high-resolution dataset of rainfall and flooding events spanning 10 years, with rainfall events and basins of widely varying characteristics selected across the continental United States. It overcomes major limitations in prior studies that were based on limited observations or hydrological model simulations. This study explores first-order dependencies in the relationships between peak discharge, rainfall variability, and basin physiography, and it sheds light on these complex interactions using a multidimensional statistical modeling approach. Among different machine-learning techniques, XGBoost is used to determine the significant physiographical and rainfall characteristics that influence peak discharge through variable importance analysis. A parsimonious model with low bias and variance is created that can be deployed in the future for flash flood forecasting. The results confirm that, although the spatial organization of rainfall within a basin has a major influence on basin response, basin physiography is the primary driver of peak discharge. These findings have unprecedented spatial and temporal representativeness in terms of flood characterization across basins. An improved understanding of subbasin scale rainfall spatial variability will aid in robust flash flood characterization as well as with identifying basins that could most benefit from distributed hydrologic modeling.
Abstract
As a fundamental water flux, quantitative understanding of precipitation is important to understand and manage water systems under a changing climate, especially in transition regions such as the coastal interface between land and ocean. This work aims to assess the uncertainty in precipitation detection over the land–coast–ocean continuum in the Integrated Multisatellite Retrievals for Global Precipitation Measurement (IMERG) V06B product. It is examined over three coastal regions of the United States—the West Coast, the Gulf of Mexico, and the East Coast, all of which are characterized by different topographies and precipitation climatologies. Detection capabilities are contrasted over different surfaces (land, coast, and ocean). A novel and integrated approach traces the IMERG detection performance back to its components (passive microwave, infrared, and morphing-based estimates). The analysis is performed by using high-resolution, high-quality Ground Validation Multi-Radar/Multi-Sensor (GV-MRMS) rainfall estimates as ground reference. The best detection performances are reported with PMW estimates (hit rates in the range [25%–39%]), followed by morphing ([20%–34%]), morphing+IR ([17%–27%]) and IR ([11%–16%]) estimates. Precipitation formation mechanisms play an important role, especially in the West Coast where orographic processes challenge detection. Further, precipitation typology is shown to be a strong driver of IMERG detection. Over the ocean, IMERG detection is generally better but suffers from false alarms ([10%–53%]). Overall, IMERG displays nonhomogeneous precipitation detection capabilities tracing back to its components. Results point toward a similar behavior across various land–coast–ocean continuum regions of the CONUS, which suggests that results can be potentially transferred to other coastal regions of the world.
Abstract
As a fundamental water flux, quantitative understanding of precipitation is important to understand and manage water systems under a changing climate, especially in transition regions such as the coastal interface between land and ocean. This work aims to assess the uncertainty in precipitation detection over the land–coast–ocean continuum in the Integrated Multisatellite Retrievals for Global Precipitation Measurement (IMERG) V06B product. It is examined over three coastal regions of the United States—the West Coast, the Gulf of Mexico, and the East Coast, all of which are characterized by different topographies and precipitation climatologies. Detection capabilities are contrasted over different surfaces (land, coast, and ocean). A novel and integrated approach traces the IMERG detection performance back to its components (passive microwave, infrared, and morphing-based estimates). The analysis is performed by using high-resolution, high-quality Ground Validation Multi-Radar/Multi-Sensor (GV-MRMS) rainfall estimates as ground reference. The best detection performances are reported with PMW estimates (hit rates in the range [25%–39%]), followed by morphing ([20%–34%]), morphing+IR ([17%–27%]) and IR ([11%–16%]) estimates. Precipitation formation mechanisms play an important role, especially in the West Coast where orographic processes challenge detection. Further, precipitation typology is shown to be a strong driver of IMERG detection. Over the ocean, IMERG detection is generally better but suffers from false alarms ([10%–53%]). Overall, IMERG displays nonhomogeneous precipitation detection capabilities tracing back to its components. Results point toward a similar behavior across various land–coast–ocean continuum regions of the CONUS, which suggests that results can be potentially transferred to other coastal regions of the world.
