Sea Surface Wind Speed Data Research Articles

Optimizing Back-Propagation Neural Network to Retrieve Sea Surface Temperature Based on Improved Sparrow Search Algorithm

Sea surface temperature (SST) constitutes a pivotal physical parameter in the investigation of atmospheric, oceanic, and air–sea exchange processes. The retrieval of SST through satellite passive microwave (PMW) technology effectively mitigates the interference posed by cloud cover, addressing a longstanding challenge. Nevertheless, conventional functional representations often fall short in capturing the intricate interplay of factors influencing SST. Leveraging neural networks (NNs), known for their adeptness in tackling nonlinear and intricate problems, holds great promise in SST retrieval. Nonetheless, NNs exhibit a high sensitivity to initial weights and thresholds, rendering them susceptible to local optimization issues. In this study, we present a novel machine learning (ML) approach for SST retrieval using PMW measurements, drawing from the Sparrow Search Algorithm (SSA) and Back-Propagation neural network (BPNN) methodologies. The core premise involves the optimization of the BP neural network’s initial weights and thresholds through an enhanced SSA algorithm employing various optimization strategies. This optimization aims to provide superior parameters for the training of the BP neural network. Employing AMSR2 brightness temperature data, sea surface wind speed data, and buoy SST measurements, we construct the ISSA-BP model for sea surface temperature retrieval. The validation of the ISSA-BP model against the test data is conducted and compared against the multiple linear regression (MLR) model, an unoptimized BP model, and an unimproved SSA-BP model. The results manifest an impressive R-squared (R2) value of 0.9918 and a root-mean-square error (RMSE) of 0.8268 °C for the ISSA-BP model, attesting to its superior accuracy. Furthermore, the ISSA-BP model was applied to retrieve global sea surface temperatures on 15 July 2022, yielding an R2 of 0.9926 and an RMSE of 0.7673 °C for the OISST product on the same day, underscoring its excellent concordance. The results indicate that SST can be efficiently and accurately retrieved using the model proposed in this paper, based on satellite PMW measurements. This finding underscores the potential of employing machine learning algorithms for SST retrieval and offers a valuable reference for future studies focusing on the retrieval of other sea surface parameters.

Evaluation on sea surface wind speed product of Fengyun-3D satellite microwave radiation imager using ERA-5 dataset

ABSTRACT In this study, we evaluate the two years (from 1 May 2019 to 30 April 2021) global operational sea surface wind speed (SWS) product of Fengyun-3D (FY-3D) polar-orbit meteorological satellite microwave radiation imager (MWRI) using ERA-5 reanalysis data. The total mean bias error (MBE), mean absolute error (MAE) and correlation coefficient of FY-3D/MWRI SWS product are −0.432 ± 3.09 m/s, 1.896 ± 2.50 m/s and 0.76, respectively. The current FY-3D/MWRI SWS data are comparable with the wind speed data of ERA-5 at most of the sea areas, especially the farther offshore sea areas. However, we still find significant overestimations (more than 5 m/s) near the coasts and typical differences in horizontal MBE values in different ascending and descending orbits. Besides, the rainy pixels as another negative impact factor are also common in FY-3D/MWRI SWS product, which can induce manifest overestimations of SWS. Therefore, the future FY-3/MWRI SWS operational algorithm should further develop a rigorous and robust quality-control module and a joint rainy pixel identification algorithm to better eliminate the unreasonable (or anomalous) pixels and rainy pixels. Moreover, the detailed evaluation on the on-orbit calibration performance of MWRI sensor is also necessary for both ascending and descending orbits.

Wind Speed Retrieval Using Global Precipitation Measurement Dual-Frequency Precipitation Radar Ka-Band Data at Low Incidence Angles

In this study, sea surface wind speed was retrieved using the Global Precipitation Measurement (GPM) dual-frequency precipitation radar (DPR) Ka-band data. In order to establish the Ka-band model at low incidence angles, the dependence of the DPR Ka-band normalized radar cross section (NRCS) on the wind speed, incidence angle, sea surface temperature (SST), significant wave height (SWH), and sea surface current speed (CSPD) was analyzed first. We confirmed that the normalized radar cross section depends on the wind speed, incidence angle, and SST. Second, an empirical model at low incidence angles was established. This model links the Ka-band NRCS to the incidence angle, wind speed, and SST. Additionally, the wind speed was retrieved by the model and was validated via the GPM Microwave Imager (GMI) wind product. The validation yielded a root mean square error (RMSE) of 1.45 m/s and the RMSE was better at a lower incidence angle and a higher SST. This model may expand the use of GPM DPR data in enriching the sea surface wind speed data set. It is also helpful for other Ka-band spaceborne radars at low incidence angles to measure wind speed in the future.

Pan-Arctic ocean wind and wave data by spaceborne SAR

ABSTRACT The Arctic is one of the most significant changing areas on the Earth under the climate change scenario. More regions in the Arctic are becoming ice-free oceans in the melting season or through the whole year. Therefore, ocean wind and wave, as the two most important parameters in the air–sea interface, are drawing significant attention to the Arctic Ocean. Scatterometer and radar altimeter are the two traditional remote sensing instruments for ocean wind and wave observations, while the former is limited by coarse spatial resolution and the latter has small spatial coverage. Wind and wave data in high spatial resolution and wide coverage by synthetic aperture radar (SAR) are currently lacking in the Arctic Ocean. We developed an ocean wind and wave dataset by Sentinel-1 SAR in the pan-Arctic Ocean (above 60°N), covering January 2017 to May 2021. By comparing with sea surface wind speed data of scatterometer, the SAR-retrieved wind data achieve an accuracy of 1.23 m/s, in terms of root mean square error (RMSE). Compared with significant wave height data of radar altimeter, the SAR retrievals have an RMSE of 0.66 m. The data records are in the standard NetCDF-4 format. The dataset is publicly available at: http://www.dx.doi.org/10.11922/sciencedb.00834.

New Ku-Band Geophysical Model Function and Wind Speed Retrieval Algorithm Developed for Tiangong-2 Interferometric Imaging Radar Altimeter

A Ku-band geophysical model function for wind speed retrieval named as KuLMOD-H is proposed based on the quasi-specular reflection model for the Chinese Tiangong-2 Interferometric Imaging Radar Altimeter (TG2-InIRA), which can be used to retrieve the wind speed with 2 km resolution. The model is derived by expanding the effective nadir reflection coefficient term and the mean square slope term in the quasi-specular reflection model using quadratic polynomials with wind speed as variable. The model coefficients are obtained by fitting the radar backscattering coefficient data from TG2-InIRA to the collocated sea surface wind speed data from European Center for Medium-Range Weather Forecasts (ECMWF). For solving the problem of potential ambiguous solutions when the incidence angles are relatively large, a regularization approach is further proposed. The retrieved wind speed results have a root-mean-square error (RMSE) of 1.42 m/s compared with the collocated ECMWF wind speed data and at the same time, they are highly consistent with the buoy data. Different from the previous works on TG2-InIRA wind speed retrieval, this article derives a semiphysical model suitable for incidence angles from 1 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^\circ$</tex-math></inline-formula> to 8 <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$^\circ$</tex-math></inline-formula> , by which better retrieval results are achieved. This work not only explores the wind speed retrieval capability of the instrument under low incidence-angle observation, but also provides high-resolution wind speed data for further correction of sea state bias for TG2-InIRA sea surface height measurements.

Characteristics and Applications of the Ground-Based X Band Low Elevation Angle Brightness Temperatures under Low Sea State Based on Measured Data

In ground-based microwave radiometer remote sensing, low-elevation-angle (−3°~3°) radiation data are often discarded because they are considered to be of little value and are often difficult to model due to the complicated mechanism. Based on the observed X-band horizontal polarization low elevation angle microwave radiation data and the meteorological data at the same time, this study investigated the generation mechanism of low elevation angle brightness temperature (LEATB) and its relationship with meteorological data, i.e., temperature, humidity, and wind speed, under low sea state. As a result, one could find that the LEATB was sensitive to the atmosphere at the elevation angle between 1° to 3°, and a diurnal variation of the LEATB reached up to 10 K. This study also found a linear relationship between the LEATB and sea surface wind speed under low sea state at an elevation range from −3° to 0°, i.e., the brightness temperature decreased as the wind speed increased, which was inconsistent with the observations at the elevation angle from −10° to −5°. The variation of the LEATB difference according to the change in the over-the-horizon detection capability (OTHDC) of the shipborne microwave radar was examined to identify the reason for this phenomenon theoretically. The results showed that the LEATB difference was significantly influenced by a change in the OTHDC. Further, this study examined a remote sensing method to extract the sea surface wind speed data from experimental LEATB data under low sea state. The results demonstrated that the X-band horizontal polarization LEATBs were useful to retrieve the sea surface wind speed data at a reasonable accuracy—the root mean square error of 0.02408 m/s. Overall, this study proved the promising potential of the LEATB data for retrieving temperature profiles, humidity profiles, sea surface winds, and the OTHDC.

Optimizing cirrus optical depth retrievals over the ocean from collocated CALIPSO and AMSR-E observations.

Retrievals of particulate optical depths and extinction coefficients from the cloud-aerosol lidar with orthogonal polarization (CALIOP) instrument deployed on the CALIPSO satellite mainly rely on a single global mean extinction-to-backscatter ratio, also known as the lidar ratio. However, the lidar ratio depends on the microphysical properties of particulates. An alternative approach is adopted to infer single-layer semi-transparent cirrus optical depths (CODs) over the open ocean that does not rely on an assumed lidar ratio. Instead, the COD is inferred directly from backscatter measurements obtained from the CALIOP lidar in conjunction with collocated sea surface wind speed data obtained from AMSR-E. This method is based on a Gram-Charlier ocean surface reflectance model relating wind-driven wave slope variances to sea surface wind speeds. To properly apply this method, the impact of multiple scattering between the sea surface and ice clouds should be taken into account. We take advantage of the 532nm cross-polarization feature of CALIOP and introduce an empirical method based on the depolarization change at the sea surface to correct for potential bias in sea surface backscatter caused by whitecaps, bubbles, foam, and multiple scattering. After the correction, the COD can be derived for individual CALIOP retrievals in a single cloud layer over the ocean with this method. The global mean COD was found to be roughly 14% higher than the current values determined by the Version 4 CALIOP extinction retrieval algorithm. This study is relevant to future improvements of CALIOP operational products and is expected to lead to more accurate COD retrievals.

Wind speed trends over the southwestern Atlantic Ocean, between 33° and 50°S

Surface scalar wind speed trends (from 1979 to 2009) at the southwestern Atlantic Ocean, between 33° and 50°S and 55° and 70°W, were estimated from the NCEP/NCAR I (NR1) and the Climate Forecast System Reanalysis (CFSR) using Sen's slope. Trends were tested using the seasonal Mann–Kendall test. Scalar wind speed trends computed from NR1 database were positive throughout the whole study area, with values ranging from +0.01 to +0.03 m s−1 year−1. On the contrary, wind speed trends computed from CFSR database were rather different to trends obtained from NR1 reanalysis. Wind speed and sea surface temperature (SST) trend patterns, both computed from CFSR database, seem to be quite consistent but, in general, no apparent relationship between both patterns was obtained when different global databases were analysed. Sea surface wind speed data compiled from multiple satellite observations were also used to compute wind speed trends. Computed satellite trends resulted in very good agreement with positive values obtained from NR1 database. NR1 data series presented significant inter-annual to multi-decadal oscillations at the Argentinean continental shelf, but they would not seem to be associated either with the Southern Annular Mode or with El Niño Southern Oscillation. Possible impacts of positive speed wind trends on the mean depth of the ocean mixed layer and on the wind wave climate are briefly discussed.

Assessment of Sea Surface Wind from NWP Reanalyses and Satellites in the Southern Ocean

Abstract Reanalysis projects and satellite data analysis have provided surface wind over the global ocean. To assess how well one can reconstruct the variations of surface wind in the data-sparse Southern Ocean, sea surface wind speed data from 1) the National Centers for Environmental Prediction–Department of Energy reanalysis (NCEP–DOE), 2) the European Centre for Medium-Range Weather Forecasts (ECMWF) Interim Re-Analysis (ERA-Interim), 3) National Climate Data Center (NCDC) blended sea winds, and 4) cross-calibrated multiplatform (CCMP) ocean surface velocity are evaluated. First, the accuracy of sea surface wind speed is validated with quality-controlled in situ measurements from research vessels. The results show that the CCMP value is closer to the ship observations than other products, whereas the NCEP–DOE value has the largest systematic positive bias. All four products show large positive biases under weak wind regimes, good agreement with the ship observations under moderate wind regimes, and large negative biases under high wind regimes. Second, the consistency and discrepancy of sea surface wind speed across different products is examined. The intercomparisons suggest that these products show encouraging agreement in the spatial distribution of the annual-mean sea surface wind speed. The largest across-data scatter is found in the central Indian sector of the Antarctic Circumpolar Current, which is comparable to its respective interannual variability. The monthly-mean correlations between pairs of products are high. However, differing from the decadal trends of NCEP–DOE, NCDC, and CCMP that show an increase of sea surface wind speed in the Antarctic Circumpolar region, ERA-Interim has an opposite sign there.

Sea surface wind speed estimation from space-based lidar measurements

Abstract. Global satellite observations of lidar backscatter measurements acquired by the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) mission and collocated sea surface wind speed data from the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E), are used to investigate the relation between wind driven wave slope variance and sea surface wind speed. The new slope variance – wind speed relation established from this study is similar to the linear relation from Cox-Munk (1954) and the log-linear relation from Wu (1990) for wind speed larger than 7 m/s and 13.3 m/s, respectively. For wind speed less than 7 m/s, the slope variance is proportional to the square root of the wind speed, assuming a two dimensional isotropic Gaussian wave slope distribution. This slope variance – wind speed relation becomes linear if a one dimensional Gaussian wave slope distribution and linear slope variance – wind speed relation are assumed. Contributions from whitecaps and subsurface backscattering are effectively removed by using 532 nm lidar depolarization measurements. This new slope variance – wind speed relation is used to derive sea surface wind speed from CALIPSO single shot lidar measurements (70 m spot size), after correcting for atmospheric attenuation. The CALIPSO wind speed result agrees with the collocated AMSR-E wind speed, with 1.2 m/s rms error. Ocean surface with lowest atmospheric loading and moderate wind speed (7–9 m/s) is used as target for lidar calibration correction.

Retrieval of TPW over ocean from locally received AMSU measurements

Total Precipitable Water (TPW) in a column of atmosphere is one of the important parameters useful for a number of meteorological applications. In the present study, a neural network based algorithm has been developed for the retrieval of TPW using NOAA-16 AMSU measurements. The TPW has been derived experimentally using NOAA-16 AMSU measurements locally received from High Resolution Picture Transmission (HRPT) station at India Meteorological Department (IMD) separately over ocean only. The validation of TPW has been carried out against the TPW derived from Radiosonde (RAOB) data. The bias and rms errors against the RAOB derived TPW have been found to about 0.11 mm and 2.98 mm respectively. The inter comparisons of TPW derived using NOAA AMSU data have also been made with that of NOAA/NESDIS derived TPW. Further, case study for the potential use of TPW derived from NOAA AMSU data has been carried out. This case study has revealed that the concentration of maximum precipitable water values in conjunction with high Sea surface wind speed data from Quickscat Scatterometer were found very useful for forecasting the heavy to very heavy rainfall event along the west coast of India. Therefore, AMSU derived TPW could be used as an important parameter for the operational weather forecasting on a real time basis.

The Probability Distribution of Sea Surface Wind Speeds. Part II: Dataset Intercomparison and Seasonal Variability

Abstract The statistical structure of sea surface wind speeds is considered, both in terms of the leading-order moments (mean, standard deviation, and skewness) and in terms of the parameters of a best-fit Weibull distribution. An intercomparison is made of the statistical structure of sea surface wind speed data from four different datasets: SeaWinds scatterometer observations, a blend of Special Sensor Microwave Imager (SSM/I) satellite observations with ECMWF analyses, and two reanalysis products [NCEP–NCAR and 40-yr ECMWF Re-Analysis (ERA-40)]. It is found that while the details of the statistical structure of sea surface wind speeds differs between the datasets, the leading-order features of the distributions are consistent. In particular, it is found in all datasets that the skewness of the wind speed is a concave upward function of the ratio of the mean wind speed to its standard deviation, such that the skewness is positive where the ratio is relatively small (such as over the extratropical Northern Hemisphere), the skewness is close to zero where the ratio is intermediate (such as the Southern Ocean), and the skewness is negative where the ratio is relatively large (such as the equatorward flank of the subtropical highs). This relationship between moments is also found in buoy observations of sea surface winds. In addition, the seasonal evolution of the probability distribution of sea surface wind speeds is characterized. It is found that the statistical structure on seasonal time scales shares the relationships between moments characteristic of the year-round data. Furthermore, the seasonal data are shown to depart from Weibull behavior in the same fashion as the year-round data, indicating that non-Weibull structure in the year-round data does not arise due to seasonal nonstationarity in the parameters of a strictly Weibull time series.