Browse
Removing Biases in Oceanic Turbulent Kinetic Energy Dissipation Rate Estimated from Microstructure Shear DataAbstract
To calculate a turbulent kinetic energy dissipation rate from the microstructure vertical shear of the horizontal velocity via a spectral analysis, shear spectra need first to be cleaned from vibrations of the moving vehicle. Unambiguously, this study shows that the spectral cleaning must be applied all over the frequency range and not only at frequencies larger than 10 Hz, as a recent study suggested. For a Vertical Microstructure Profiler (VMP-6000), not correcting for vehicle vibrations below 10 Hz leads to overestimated dissipation rates from 50% to 700% for usual downcast velocities and for weak dissipation rates (ε < 1 × 10−9 W kg−1). Vibrations concern all vehicles, but the exact vibrational frequency signature depends on the vehicle shape and its downcast velocity. In any case, a spectral cleaning over the whole frequency range is strongly advised. This study also reports on a systematic low bias of inferred dissipation rates induced by the spectral cleaning when too few degrees of freedom are available for the cleaning, which is usually the default of the standard processing. Whatever the dissipation rate level, not correcting for the bias leads to underestimated dissipation rates by a factor 1.4–2.7 (with usual parameters), the exact amplitude of the bias depending on the number of degrees of freedom and on the number of independent accelerometer axes used for the cleaning. It is strongly advised that such a bias be taken into account to recompute dissipation rates of past datasets and for future observations.
Abstract
To calculate a turbulent kinetic energy dissipation rate from the microstructure vertical shear of the horizontal velocity via a spectral analysis, shear spectra need first to be cleaned from vibrations of the moving vehicle. Unambiguously, this study shows that the spectral cleaning must be applied all over the frequency range and not only at frequencies larger than 10 Hz, as a recent study suggested. For a Vertical Microstructure Profiler (VMP-6000), not correcting for vehicle vibrations below 10 Hz leads to overestimated dissipation rates from 50% to 700% for usual downcast velocities and for weak dissipation rates (ε < 1 × 10−9 W kg−1). Vibrations concern all vehicles, but the exact vibrational frequency signature depends on the vehicle shape and its downcast velocity. In any case, a spectral cleaning over the whole frequency range is strongly advised. This study also reports on a systematic low bias of inferred dissipation rates induced by the spectral cleaning when too few degrees of freedom are available for the cleaning, which is usually the default of the standard processing. Whatever the dissipation rate level, not correcting for the bias leads to underestimated dissipation rates by a factor 1.4–2.7 (with usual parameters), the exact amplitude of the bias depending on the number of degrees of freedom and on the number of independent accelerometer axes used for the cleaning. It is strongly advised that such a bias be taken into account to recompute dissipation rates of past datasets and for future observations.
Abstract
The presence of internal waves (IWs) in the ocean alters the isotropic properties of sound speed profiles (SSPs) in the water column. Changes in the SSPs affect underwater acoustics since most of the energy is dissipated into the seabed due to the downward refraction of sound waves. In consequence, variations in the SSP must be considered when modeling acoustic propagation in the ocean. Empirical orthogonal functions (EOFs) are regularly employed to model and represent SSPs using a linear combination of basis functions that capture the sound speed variability. A different approach is to use dictionary learning to obtain a learned dictionary (LD) that generates a nonorthogonal set of basis functions (atoms) that generate a better sparse representation. In this paper, the performance of EOFs and LDs are evaluated for sparse representation of SSPs affected by the passing of IWs. In addition, an LD-based supervised framework is presented for SSP classification and is compared with classical learning models. The algorithms presented in this work are trained and tested on data collected from the Shallow Water Experiment 2006. Results show that LDs yield lower reconstruction error than EOFs when using the same number of bases. In addition, overcomplete LDs are demonstrated to be a robust method to classify SSPs during low, medium, and high IW activity, reporting accuracy that is comparable to and sometimes higher than that of standard supervised classification methods.
Abstract
The presence of internal waves (IWs) in the ocean alters the isotropic properties of sound speed profiles (SSPs) in the water column. Changes in the SSPs affect underwater acoustics since most of the energy is dissipated into the seabed due to the downward refraction of sound waves. In consequence, variations in the SSP must be considered when modeling acoustic propagation in the ocean. Empirical orthogonal functions (EOFs) are regularly employed to model and represent SSPs using a linear combination of basis functions that capture the sound speed variability. A different approach is to use dictionary learning to obtain a learned dictionary (LD) that generates a nonorthogonal set of basis functions (atoms) that generate a better sparse representation. In this paper, the performance of EOFs and LDs are evaluated for sparse representation of SSPs affected by the passing of IWs. In addition, an LD-based supervised framework is presented for SSP classification and is compared with classical learning models. The algorithms presented in this work are trained and tested on data collected from the Shallow Water Experiment 2006. Results show that LDs yield lower reconstruction error than EOFs when using the same number of bases. In addition, overcomplete LDs are demonstrated to be a robust method to classify SSPs during low, medium, and high IW activity, reporting accuracy that is comparable to and sometimes higher than that of standard supervised classification methods.
Abstract
Properties of the surface mixed layer (ML) are critical for understanding and predicting atmosphere–sea ice–ocean interactions in the changing Arctic Ocean. Mooring measurements are typically unable to resolve the ML in the Arctic due to the need for instruments to remain below the surface to avoid contact with sea ice and icebergs. Here, we use measurements from a series of three moorings installed for one year in the Beaufort Sea to demonstrate that upward-looking acoustic Doppler current profilers (ADCPs) installed on subsurface floats can be used to estimate ML properties. A method is developed for combining measured peaks in acoustic backscatter and inertial shear from the ADCPs to estimate the ML depth. Additionally, we use an inverse sound speed model to infer the summer ML temperature based on offsets in ADCP altimeter distance during open-water periods. The ADCP estimates of ML depth and ML temperature compare favorably with measurements made from mooring temperature sensors, satellite SST, and from an autonomous Seaglider. These methods could be applied to other extant mooring records to recover additional information about ML property changes and variability.
Abstract
Properties of the surface mixed layer (ML) are critical for understanding and predicting atmosphere–sea ice–ocean interactions in the changing Arctic Ocean. Mooring measurements are typically unable to resolve the ML in the Arctic due to the need for instruments to remain below the surface to avoid contact with sea ice and icebergs. Here, we use measurements from a series of three moorings installed for one year in the Beaufort Sea to demonstrate that upward-looking acoustic Doppler current profilers (ADCPs) installed on subsurface floats can be used to estimate ML properties. A method is developed for combining measured peaks in acoustic backscatter and inertial shear from the ADCPs to estimate the ML depth. Additionally, we use an inverse sound speed model to infer the summer ML temperature based on offsets in ADCP altimeter distance during open-water periods. The ADCP estimates of ML depth and ML temperature compare favorably with measurements made from mooring temperature sensors, satellite SST, and from an autonomous Seaglider. These methods could be applied to other extant mooring records to recover additional information about ML property changes and variability.
Impacts of Vertical Nonuniform Beam Filling on the Observability of Secondary Ice Production due to SublimationAbstract
Quasi-vertical profiles (QVPs) of polarimetric radar data have emerged as a powerful tool for studying precipitation microphysics. Various studies have found enhancements in specific differential phase K dp in regions of suspected secondary ice production (SIP) due to rime splintering. Similar K dp enhancements have also been found in regions of sublimating snow, another proposed SIP process. This work explores these K dp signatures for two cases of sublimating snow using nearly collocated S- and Ka-band radars. The presence of the signature was inconsistent between the radars, prompting exploration of alternative causes. Idealized simulations are performed using a radar beam-broadening model to explore the impact of nonuniform beam filling (NBF) on the observed reflectivity Z and K dp within the sublimation layer. Rather than an intrinsic increase in ice concentration, the observed K dp enhancements can instead be explained by NBF in the presence of sharp vertical gradients of Z and K dp within the sublimation zone, which results in a K dp bias dipole. The severity of the bias is sensitive to the Z gradient and radar beamwidth and elevation angle, which explains its appearance at only one radar. In addition, differences in scanning strategies and range thresholds during QVP processing can constructively enhance these positive K dp biases by excluding the negative portion of the dipole. These results highlight the need to consider NBF effects in regions not traditionally considered (e.g., in pure snow) due to the increased K dp fidelity afforded by QVPs and the subsequent ramifications this has on the observability of sublimational SIP.
Significance Statement
Many different processes can cause snowflakes to break apart into numerous tiny pieces, including when they evaporate into dry air. Purported evidence of this phenomenon has been seen in data from some weather radars, but we noticed it was not seen in data from others. In this work we use case studies and models to show that this signature may actually be an artifact from the radar beam becoming too big and there being too much variability of the precipitation within it. While this breakup process may actually be occurring in reality, these results suggest we may have trouble observing it with typical weather radars.
Abstract
Quasi-vertical profiles (QVPs) of polarimetric radar data have emerged as a powerful tool for studying precipitation microphysics. Various studies have found enhancements in specific differential phase K dp in regions of suspected secondary ice production (SIP) due to rime splintering. Similar K dp enhancements have also been found in regions of sublimating snow, another proposed SIP process. This work explores these K dp signatures for two cases of sublimating snow using nearly collocated S- and Ka-band radars. The presence of the signature was inconsistent between the radars, prompting exploration of alternative causes. Idealized simulations are performed using a radar beam-broadening model to explore the impact of nonuniform beam filling (NBF) on the observed reflectivity Z and K dp within the sublimation layer. Rather than an intrinsic increase in ice concentration, the observed K dp enhancements can instead be explained by NBF in the presence of sharp vertical gradients of Z and K dp within the sublimation zone, which results in a K dp bias dipole. The severity of the bias is sensitive to the Z gradient and radar beamwidth and elevation angle, which explains its appearance at only one radar. In addition, differences in scanning strategies and range thresholds during QVP processing can constructively enhance these positive K dp biases by excluding the negative portion of the dipole. These results highlight the need to consider NBF effects in regions not traditionally considered (e.g., in pure snow) due to the increased K dp fidelity afforded by QVPs and the subsequent ramifications this has on the observability of sublimational SIP.
Significance Statement
Many different processes can cause snowflakes to break apart into numerous tiny pieces, including when they evaporate into dry air. Purported evidence of this phenomenon has been seen in data from some weather radars, but we noticed it was not seen in data from others. In this work we use case studies and models to show that this signature may actually be an artifact from the radar beam becoming too big and there being too much variability of the precipitation within it. While this breakup process may actually be occurring in reality, these results suggest we may have trouble observing it with typical weather radars.
Abstract
A new full-ocean-depth macroorganisms pressure-retaining sampler (FMPS) was designed to collect pressure-retaining macroorganisms samples from the abyssal seafloor. A mathematical model for pressure compensation in the FMPS recovery process was developed. The effects of FMPS structural parameters, pressure compensator structural parameters, and sampling environment on the pressure retention performance of FMPS were analyzed. Using the developed FMPS engineering prototype, FMPS internal pressure test, high-pressure chamber simulation sampling, and pressure-retaining test was carried out. The test results show that the key components of FMPS can carry 115 MPa pressure, FMPS can complete the sampling action in the high-pressure chamber of 115 MPa, the pressure is maintained at 105.5 MPa, and the pressure drop rate (ratio of pressure drop during FMPS recovery to sampling point pressure) is 9.13%; the experimental results are consistent with the theoretical calculation. The test verified the feasibility of FMPS design and the reliability of pressure retention, providing a theoretical basis and technical support for the design and manufacture of full-ocean-depth sampling devices.
Abstract
A new full-ocean-depth macroorganisms pressure-retaining sampler (FMPS) was designed to collect pressure-retaining macroorganisms samples from the abyssal seafloor. A mathematical model for pressure compensation in the FMPS recovery process was developed. The effects of FMPS structural parameters, pressure compensator structural parameters, and sampling environment on the pressure retention performance of FMPS were analyzed. Using the developed FMPS engineering prototype, FMPS internal pressure test, high-pressure chamber simulation sampling, and pressure-retaining test was carried out. The test results show that the key components of FMPS can carry 115 MPa pressure, FMPS can complete the sampling action in the high-pressure chamber of 115 MPa, the pressure is maintained at 105.5 MPa, and the pressure drop rate (ratio of pressure drop during FMPS recovery to sampling point pressure) is 9.13%; the experimental results are consistent with the theoretical calculation. The test verified the feasibility of FMPS design and the reliability of pressure retention, providing a theoretical basis and technical support for the design and manufacture of full-ocean-depth sampling devices.
Abstract
Global Navigation Satellite System Reflectometry (GNSS-R)-based wind retrieval techniques use the global positioning system (GPS) signals scattered from the ocean surface in the forward direction, and can potentially work in all weather conditions. An overview of recent progress made in the Cyclone Global Navigation Satellite System (CYGNSS) level-2 surface wind products is given. To this end, four publicly released CYGNSS surface wind products—Science Data Record (SDR) v2.1, SDR v3.0, Climate Data Record (CDR) v1.1, and science wind speed product NOAA v1.1—are validated quantitatively against high-quality data from tropical buoy arrays. The latest released CYGNSS wind products (e.g., CDR v1.1, SDR v3.0, NOAA v1.1), as compared with these tropical buoy data, significantly outperform the SDR v2.1. Moreover, the uncertainty among these products is found to be less than 2 m s−1 root-mean-squared difference, meeting the NASA science mission level-1 uncertainty requirement for wind speeds below 20 m s−1. The quality of the CYGNSS wind is further assessed under different precipitation conditions in low winds, and in large-scale convective regions. Results show that the presence of rain appears to cause a slightly positive wind speed bias in all CYGNSS data. Nonetheless, the outcomes are encouraging for the recently released CYGNSS wind products in general, and for CYGNSS data in regions with precipitating deep convection. The overall comparison indicates a significant improvement in wind speed quality and sample size when going from the older version to any of the newer datasets.
Abstract
Global Navigation Satellite System Reflectometry (GNSS-R)-based wind retrieval techniques use the global positioning system (GPS) signals scattered from the ocean surface in the forward direction, and can potentially work in all weather conditions. An overview of recent progress made in the Cyclone Global Navigation Satellite System (CYGNSS) level-2 surface wind products is given. To this end, four publicly released CYGNSS surface wind products—Science Data Record (SDR) v2.1, SDR v3.0, Climate Data Record (CDR) v1.1, and science wind speed product NOAA v1.1—are validated quantitatively against high-quality data from tropical buoy arrays. The latest released CYGNSS wind products (e.g., CDR v1.1, SDR v3.0, NOAA v1.1), as compared with these tropical buoy data, significantly outperform the SDR v2.1. Moreover, the uncertainty among these products is found to be less than 2 m s−1 root-mean-squared difference, meeting the NASA science mission level-1 uncertainty requirement for wind speeds below 20 m s−1. The quality of the CYGNSS wind is further assessed under different precipitation conditions in low winds, and in large-scale convective regions. Results show that the presence of rain appears to cause a slightly positive wind speed bias in all CYGNSS data. Nonetheless, the outcomes are encouraging for the recently released CYGNSS wind products in general, and for CYGNSS data in regions with precipitating deep convection. The overall comparison indicates a significant improvement in wind speed quality and sample size when going from the older version to any of the newer datasets.
Abstract
Acoustically tracked subsurface floats provide insights into ocean complexity and were first deployed over 60 years ago. A standard tracking method uses a least squares algorithm to estimate float trajectories based on acoustic ranging from moored sound sources. However, infrequent or imperfect data challenge such estimates, and least squares algorithms are vulnerable to non-Gaussian errors. Acoustic tracking is currently the only feasible strategy for recovering float positions in the sea ice region, a focus of this study. Acoustic records recovered from underice floats frequently lack continuous sound source coverage. This is because environmental factors such as surface sound channels and rough sea ice attenuate acoustic signals, while operational considerations make polar sound sources expensive and difficult to deploy. Here we present a Kalman smoother approach that, by including some estimates of float behavior, extends tracking to situations with more challenging datasets. The Kalman smoother constructs dynamically constrained, error-minimized float tracks and variance ellipses using all possible position data. This algorithm outperforms the least squares approach and a Kalman filter in numerical experiments. The Kalman smoother is applied to previously tracked floats from the southeast Pacific (DIMES experiment), and the results are compared with existing trajectories constructed using the least squares algorithm. The Kalman smoother is also used to reconstruct the trajectories of a set of previously untracked, acoustically enabled Argo floats in the Weddell Sea.
Abstract
Acoustically tracked subsurface floats provide insights into ocean complexity and were first deployed over 60 years ago. A standard tracking method uses a least squares algorithm to estimate float trajectories based on acoustic ranging from moored sound sources. However, infrequent or imperfect data challenge such estimates, and least squares algorithms are vulnerable to non-Gaussian errors. Acoustic tracking is currently the only feasible strategy for recovering float positions in the sea ice region, a focus of this study. Acoustic records recovered from underice floats frequently lack continuous sound source coverage. This is because environmental factors such as surface sound channels and rough sea ice attenuate acoustic signals, while operational considerations make polar sound sources expensive and difficult to deploy. Here we present a Kalman smoother approach that, by including some estimates of float behavior, extends tracking to situations with more challenging datasets. The Kalman smoother constructs dynamically constrained, error-minimized float tracks and variance ellipses using all possible position data. This algorithm outperforms the least squares approach and a Kalman filter in numerical experiments. The Kalman smoother is applied to previously tracked floats from the southeast Pacific (DIMES experiment), and the results are compared with existing trajectories constructed using the least squares algorithm. The Kalman smoother is also used to reconstruct the trajectories of a set of previously untracked, acoustically enabled Argo floats in the Weddell Sea.
Abstract
In retrievals of sea ice motion vectors (SIMVs) based on passive microwave observations, the use of the high-resolution 89-GHz channel of the Advanced Microwave Scanning Radiometer 2 (AMSR2) has the advantage of enhancing the theoretical precision of correlation-based motion tracking. However, its higher sensitivity to atmospheric moisture than lower-frequency channels links maximum cross-correlation peaks to outlier vectors and obscures signals of valid vectors. This study develops an algorithm to select valid vectors from candidates detected by multiple cross-correlation peaks based on validations with large-scale sea ice displacements extracted from 19- and 37-GHz data after questionable vectors are prefiltered by comparing them with reanalysis surface wind and neighboring vectors. The algorithm selects a vector corresponding to large-scale motion as the optimal vector. The retrieved results from 2013 to 2020 show that by replacing outlier vectors with valid ones detected by second or third cross-correlation peaks, validation with simultaneous observations enables retrieval of more than 60% of the Arctic motion field from 89-GHz data in winter but only 10% in summer; therefore, lower-frequency data are employed for retrievals. The uncertainty assessment using in situ data from acoustic measurements from ocean moorings shows that the algorithm provides daily SIMVs with root-mean-square errors of only 1–2 cm s–1 in idealized winter conditions with the absence of diurnal brightness temperature (Tb) changes that make tracking of the similarity of Tb fields difficult. The analysis also illustrates the applicability limit of the algorithm for summer retrievals.
Significance Statement
An algorithm was developed to validate sea ice motion vectors retrieved from AMSR2 89-GHz data by those from lower-frequency data. The validation via simultaneous observations enabled that valid vectors are sorted from invalid ones resulting from weather effects.
Abstract
In retrievals of sea ice motion vectors (SIMVs) based on passive microwave observations, the use of the high-resolution 89-GHz channel of the Advanced Microwave Scanning Radiometer 2 (AMSR2) has the advantage of enhancing the theoretical precision of correlation-based motion tracking. However, its higher sensitivity to atmospheric moisture than lower-frequency channels links maximum cross-correlation peaks to outlier vectors and obscures signals of valid vectors. This study develops an algorithm to select valid vectors from candidates detected by multiple cross-correlation peaks based on validations with large-scale sea ice displacements extracted from 19- and 37-GHz data after questionable vectors are prefiltered by comparing them with reanalysis surface wind and neighboring vectors. The algorithm selects a vector corresponding to large-scale motion as the optimal vector. The retrieved results from 2013 to 2020 show that by replacing outlier vectors with valid ones detected by second or third cross-correlation peaks, validation with simultaneous observations enables retrieval of more than 60% of the Arctic motion field from 89-GHz data in winter but only 10% in summer; therefore, lower-frequency data are employed for retrievals. The uncertainty assessment using in situ data from acoustic measurements from ocean moorings shows that the algorithm provides daily SIMVs with root-mean-square errors of only 1–2 cm s–1 in idealized winter conditions with the absence of diurnal brightness temperature (Tb) changes that make tracking of the similarity of Tb fields difficult. The analysis also illustrates the applicability limit of the algorithm for summer retrievals.
Significance Statement
An algorithm was developed to validate sea ice motion vectors retrieved from AMSR2 89-GHz data by those from lower-frequency data. The validation via simultaneous observations enabled that valid vectors are sorted from invalid ones resulting from weather effects.
A New Objective Typhoon Location Algorithm Considering a Perturbation Factor Based on FY-4A Brightness Temperature DataAbstract
In this paper, a new objective typhoon positioning algorithm was proposed. The algorithm uses L1 12-channel far-infrared data of the FY-4A geostationary meteorological satellite for objective positioning, verified against best path data provided by the Tropical Cyclone Data Center of the China Meteorological Administration (CMA). By calculating the tangential and radial perturbation values of infrared brightness temperature images, the perturbation factor can be obtained. By adopting the position of the maximum perturbation factor as the center of a circle and considering a radius of no more than 20 km, the position of the minimum perturbation factor was determined; this value represents the central position of the typhoon. Tropical cyclones in 2019 and 2020 were selected for objective positioning, and the objective positioning results were verified against the corresponding time in the best path dataset. The results included centralized verification results for 29 typhoons and optimal path data in 2019. The maximum average error reached 54.67 km, with an annual average typhoon positioning error of 16.15 km. The centralization verification results for 23 typhoons and optimal path data in 2020 indicated a minimum average error of 12.71 km, a maximum average error of 18.56 km, and an annual average typhoon positioning error of 14.82 km. The positioning results for these two years suggest that the longitude obtained with the perturbation factor positioning method is satisfactory, exhibiting an error of less than 20 km.
Significance Statement
The purpose of this study is to help researchers make scientific discoveries and help the development of typhoon center location technology in the future. This is important because accurate positioning of typhoon center can provide effective help for typhoon path prediction and typhoon intensity determination.
Abstract
In this paper, a new objective typhoon positioning algorithm was proposed. The algorithm uses L1 12-channel far-infrared data of the FY-4A geostationary meteorological satellite for objective positioning, verified against best path data provided by the Tropical Cyclone Data Center of the China Meteorological Administration (CMA). By calculating the tangential and radial perturbation values of infrared brightness temperature images, the perturbation factor can be obtained. By adopting the position of the maximum perturbation factor as the center of a circle and considering a radius of no more than 20 km, the position of the minimum perturbation factor was determined; this value represents the central position of the typhoon. Tropical cyclones in 2019 and 2020 were selected for objective positioning, and the objective positioning results were verified against the corresponding time in the best path dataset. The results included centralized verification results for 29 typhoons and optimal path data in 2019. The maximum average error reached 54.67 km, with an annual average typhoon positioning error of 16.15 km. The centralization verification results for 23 typhoons and optimal path data in 2020 indicated a minimum average error of 12.71 km, a maximum average error of 18.56 km, and an annual average typhoon positioning error of 14.82 km. The positioning results for these two years suggest that the longitude obtained with the perturbation factor positioning method is satisfactory, exhibiting an error of less than 20 km.
Significance Statement
The purpose of this study is to help researchers make scientific discoveries and help the development of typhoon center location technology in the future. This is important because accurate positioning of typhoon center can provide effective help for typhoon path prediction and typhoon intensity determination.
Abstract
This study introduces a validation technique for quantitative comparison of algorithms that retrieve winds from passive detection of cloud- and water vapor–drift motions, also known as atmospheric motion vectors (AMVs). The technique leverages airborne wind-profiling lidar data collected in tandem with 1-min refresh-rate geostationary satellite imagery. AMVs derived with different approaches are used with accompanying numerical weather prediction model data to estimate the full profiles of lidar-sampled winds, which enables ranking of feature tracking, quality control, and height-assignment accuracy and encourages mesoscale, multilayer, multiband wind retrieval solutions. The technique is used to compare the performance of two brightness motion, or “optical flow,” retrieval algorithms used within AMVs, 1) patch matching (PM; used within operational AMVs) and 2) an advanced variational optical flow (VOF) method enabled for most atmospheric motions by new-generation imagers. The VOF AMVs produce more accurate wind retrievals than the PM method within the benchmark in all imager bands explored. It is further shown that image regions with low texture and multilayer-cloud scenes in visible and infrared bands are tracked significantly better with the VOF approach, implying VOF produces representative AMVs where PM typically breaks down. It is also demonstrated that VOF AMVs have reduced accuracy where the brightness texture does not advect with the mean wind (e.g., gravity waves), where the image temporal noise exceeds the natural variability, and when the height assignment is poor. Finally, it is found that VOF AMVs have improved performance when using fine-temporal refresh-rate imagery, such as 1- versus 10-min data.
Abstract
This study introduces a validation technique for quantitative comparison of algorithms that retrieve winds from passive detection of cloud- and water vapor–drift motions, also known as atmospheric motion vectors (AMVs). The technique leverages airborne wind-profiling lidar data collected in tandem with 1-min refresh-rate geostationary satellite imagery. AMVs derived with different approaches are used with accompanying numerical weather prediction model data to estimate the full profiles of lidar-sampled winds, which enables ranking of feature tracking, quality control, and height-assignment accuracy and encourages mesoscale, multilayer, multiband wind retrieval solutions. The technique is used to compare the performance of two brightness motion, or “optical flow,” retrieval algorithms used within AMVs, 1) patch matching (PM; used within operational AMVs) and 2) an advanced variational optical flow (VOF) method enabled for most atmospheric motions by new-generation imagers. The VOF AMVs produce more accurate wind retrievals than the PM method within the benchmark in all imager bands explored. It is further shown that image regions with low texture and multilayer-cloud scenes in visible and infrared bands are tracked significantly better with the VOF approach, implying VOF produces representative AMVs where PM typically breaks down. It is also demonstrated that VOF AMVs have reduced accuracy where the brightness texture does not advect with the mean wind (e.g., gravity waves), where the image temporal noise exceeds the natural variability, and when the height assignment is poor. Finally, it is found that VOF AMVs have improved performance when using fine-temporal refresh-rate imagery, such as 1- versus 10-min data.
