Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1225-6692(Print)
ISSN : 2287-4518(Online)
Journal of the Korean earth science society Vol.39 No.4 pp.305-316

Status and Prospects of Marine Wind Observations from Geostationary and Polar-Orbiting Satellites for Tropical Cyclone Studies

SungHyun Nam1,2, Kyung-Ae Park2,3*
1School of Earth and Environmental Sciences, Seoul National University, Seoul 08826, Korea
2Research Institute of Oceanography, Seoul National University, Seoul 08826, Korea
3Department of Earth Science Education, Seoul National University, Seoul 08826, Korea
Corresponding author: Tel: +82-2-880-7780 Fax: +82-2-874-3289
August 1, 2018 August 22, 2018 August 23, 2018


Satellite-derived sea surface winds (SSWs) and atmospheric motion vectors (AMVs) over the global ocean, particularly including the areas in and around tropical cyclones (TCs), have been provided in a real-time and continuous manner. More and better information is now derived from technologically improved multiple satellite missions and wind retrieving techniques. The status and prospects of key SSW products retrieved from scatterometers, passive microwave radiometers, synthetic aperture radar, and altimeters as well as AMVs derived by tracking features from multiple geostationary satellites are reviewed here. The quality and error characteristics, limitations, and challenges of satellite wind observations described in the literature, which need to be carefully considered to apply the observations for both operational and scientific uses, i.e., assimilation in numerical weather forecasting, are also described. Additionally, ongoing efforts toward merging them, particularly for monitoring three-dimensional TC wind fields in a real-time and continuous manner and for providing global profiles of high-quality wind observations with the new mission are introduced. Future research is recommended to develop plans for providing more and better SSW and AMV products in a real-time and continuous manner from existing and new missions.



    Tropical cyclones (TCs), one of the most devastating natural disasters, cause intensive damage and have resulted in more than one-half million deaths over the world in the last five decades (Powell et al., 2007; Zhang et al., 2009). Most frequently, the damage to life and property associated with TCs is caused by their high winds accompanied by excessively heavy rainfall and flooding (Metcalfe et al., 2008). Many techniques and technologies have been developed to use satellite sensors to monitor the high winds associated with TCs, because it is difficult to collect in-situ wind measurement data in and around TCs, particularly in a real-time and continuous manner. Unprecedentedly, enhanced information derived from multiple satellite platforms, with more and better data sampled at higher spatial, temporal, and channel resolutions, and markedly improved data processing techniques, have enabled us to retrieve real-time or near-real-time sea surface winds (SSWs) and atmospheric motion vectors (AMVs) at upper and lower levels in the troposphere over the global ocean, including areas in and around TCs.

    Satellite-derived SSWs and AMVs have been validated and their quality and error characteristics have been examined by comparing with in-situ measurements and reanalyzed data. Careful and continuous efforts are necessary to apply, merge, and assimilate these into numerical weather prediction models considering the error characteristics of individual SSW and AMV products and to produce significant benefits/synergies for operational usages. In this paper, the status and prospects of SSWs and AMVs retrieved via satellite missions are reviewed, and efforts to merge multiple platform wind observations and to provide three-dimensional winds, emphasizing the winds associated with TCs, are introduced in a broader context. SSWs and AMVs derived from multiple satellite sensors are reviewed in Sections 2 and 3, respectively. In Section 4, methods for estimating the three-dimensional TC wind field from multiple satellite platforms and a new mission to provide global profiles of wind observations are introduced. The paper is concluded and future research is suggested in Section 5.

    Sea Surface Winds from Microwave Remote Sensing

    Wind Measurements from Satellite Scatterometers

    A scientific instrument named the ‘scatterometer’ carried aboard an Earth-orbiting satellite (space-borne scatterometry) has been used to estimate SSW, i.e., its speed and direction, from surface roughness, a method often termed ‘wind retrieval.’ Radar scatterometers, which use radio or microwaves to determine the normalized radar cross section of a sea surface, operate by transmitting a pulse of microwave energy toward the surface of the Earth and measuring the reflected energy (backscatter signal power). The instruments are precisely calibrated to make accurate backscatter measurements and used to retrieve SSW via the scattering mechanism known as Bragg scattering, which occurs from wind-generated capillarygravity waves that are generally in equilibrium with the SSW and in resonance with the microwaves. The SSW retrieval from variations of the observed backscatter power, depending on the speed and direction of SSW, incorporates a nonlinear inversion procedure associated with variations of these waves from different azimuth angles.

    Since the first operational application of the wind scatterometer, the Seasat Scatterometer (SASS, operated at Ku-band or 14 GHz) launched in 1978, there have been several missions for operational SSW observation from space by national agencies. The European Space Agency (ESA) launched the European Remote Sensing Satellite (ERS-1) Advanced Microwave Instrument (AMI) scatterometer in 1992 (ESCAT), which was followed by the ERS-2 AMI scatterometer in 1995 (both AMI fan-beam systems operated at C-band; 5.6 GHz). In 1996, the National Aeronautics and Space Administration (NASA) launched the NASA Scatterometer (NSCAT), which is a Ku-band fan-beam system aboard the Japanese Agency NASDA’s Advanced Earth Observing Satellite-I (ADEOS-I). The first scanning scatterometer (operated at Ku-band), known as ‘SeaWinds’ on QuikSCAT, was launched by NASA in 1999. A second SeaWinds instrument was flown on the NASDA ADEOS-II in 2002. The ESA and the European Organisation for Exploitation of Meteorological Satellites (EUMETSAT) launched the first C-band advanced scatterometer (ASCAT) in 2006 onboard Metop-A (Figa-Saldaña et al., 2002), which is currently providing SSWs with accuracies of 2 m s−1 and 20 o in speed and direction, respectively, typically every 1 or 2 days. A snapshot of Typhoon Soulik while at Category 4 intensity (July 10, 2013) captured by the ASCAT is shown as an example (Fig. 1). Fig. 2 shows another example of outstanding capability of QuikSCAT to observe wind speed and direction during the typhoon Rusa period in 2002. The wind vectors, enlarged from Fig. 2b (the red box), present cyclonic rotation near the center of the typhoon (Fig. 2c).

    Despite the powerful SSW applications using wind scatterometers with relatively small error, this approach has clear limitations, so an assessment of the error characteristics of global and regional SSW observations by the scatterometers remains an important task to better utilize the satellite products. The accuracy of the ASCAT wind vector around TCs was assessed by using in-situ data from dropwindsonde deployed by surveillance and reconnaissance flights during 2007- 2010 (Chou et al., 2013). The results of the comparisons between ASCAT and dropwindsonde data for 987 matching samples showed 1) bias and root-meansquare differences of 1.7 and 5.3 m s−1 , respectively, 2) large wind speed differences occurring in the high wind speed regime, such as in the case of TC, and large wind direction differences in the low wind speed regime, and 3) significantly reduced accuracy of wind vector in both low and high wind speed regimes in moisture saturated regions, implying that the rain contamination issue affects ASCAT accuracy (Table 1; Chou et al., 2013). There are on-going efforts to improve the accuracy of estimating strong winds from ASCAT (i.e., Soisuvarn et al., 2013).

    Wind Measurements from Passive Microwave Radiometers

    Another approach to retrieving SSW speed is using the microwave radiometer, such as the Special Sensor Microwave/Imager (SSM/I), which is a series of earthorbiting multi-band microwave radiometers consisting of 7 radiometers and operating on Defense Meteorological Satellite Program (DMSP) polar orbiting satellites since July 1987 (several subsequent SSM/I instruments have been launched and operated since the DMSP F08 spacecraft). The SSM/I observations correspond to a 1400-km swath on the Earth’s surface and complete global coverage is provided every 2 or 3 days, except for small patches near the poles. SSW is one of the important geophysical parameters retrieved from SSM/ I observations based on the absorption and scattering of microwaves by water in the atmosphere, as well as the roughening of the ocean surface by wind stress, which changes its emission and reflection properties. The measured SSW stress is converted to a wind speed assuming that the boundary layer over the ocean is neutrally stable. The speed of the SSW derived from SSM/I measurements was compared to the speed measured directly by buoy-mounted anemometers, yielding mean difference and standard deviation values typically less than 0.4 and 1.4 m s−1 , respectively (Table 1; Mears et al., 2001).

    Microwave sensors have an advantage in estimating SSW speed because the increase of sea surface emissivity, due to roughness and foam effects driven by SSW, is related physically to the observed brightness temperature. Many well-calibrated ocean emissivity models have been developed for passive microwave radiometers and applied to a number of passive satellite sensors other than the SSM/I, including the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI), Special Sensor Microwave Imager/Sounder (SSMIS), Advanced Microwave Scanning Radiometer for Earth Observing System (AMSR-E) sensor onboard the Aqua satellite, and Advanced Microwave Scanning Radiometer 2 (AMSR-2). Several algorithms have been developed for estimating SSW speed from these passive microwave sensors. For example, the Hong WS algorithm developed for retrieving SSW speed from passive microwave radiometers with an AMSR-2 sensor under both rainy and rain-free conditions shows good agreement with in-situ buoy measurements by anemometers under rain-free conditions, yielding typical bias and rootmean- squared error (RMSE) less than 1.0 and 1.5ms−1 , respectively (Table 1; Hong et al., 2016).

    Wind Measurements from Synthetic Aperture Radars

    In addition to operational use of passive microwave radiometers, synthetic aperture radar (SAR) has also been used to retrieve high-resolution SSW speeds in and around a TC or coastal zone, where the latter has been substantially unavailable from satellite scatterometers. Since the first wind speed observation of Seasat in 1979, high-resolution wind speeds have been obtained from an increasing number of satellite SAR observations such as ERS-1/2, Envisat Advanced Synthetic Aperture Radar (ASAR), Advanced Land Observing Satellite- Band Synthetic Aperture Radar (ALOS-1/2 PALSAR), Radarsat-1/2, TerraSAR-X, COnstellation of small Satellites for the Mediterranean Basin Observation (COSMO-SkyMed), Sentinel-1A/B, and so on. C-band SAR wind speeds can be retrieved from Normalized Rada Cross Section (NRCS) values using empirical Geophysical model functions (GMF) such as CMOD, CMOD4 (Stoffelen and Anderson, 1997), CMOD_IFR2, CMOD5, CMOD5.N (Hersbach, 2010). L-band and X-band model have also developed and widely utilized using the L-band GMF by Isoguchi and Shimada (2009) and XMOD model by Lehner et al. (2012).

    For example, because the RADARSAT SAR image taken in and around typhoon Man-Yi captures the distribution of the surface microwave properties, i.e., C-band (5.3 GHz frequency and HH polarization), in great detail, the SSW field associated with the typhoon could be retrieved by geophysical model function and polarization ratio (Fig. 3; Nam et al., 2012). Another example is the SSW off the east coast of Korea retrieved using L-band ALOS (Advanced Land Observing Satellite) PALSAR (Phased Array type L-band Synthetic Aperture Radar), where a comparison between SAR-retrieved SSW speed and in-situ buoy measurements off the coast showed a relatively small RMSE of 0.67 m s−1 (Kim et al., 2012). The error systematically depended on the SSW direction and the incident angle (large error at small incident angles), and the quality of the SSW speed was enhanced greatly from 6.80 to 1.14 m s−1 at small incident angles by using an improved geophysical model function (Kim et al., 2012).

    Wind Measurements from Satellite Altimeters

    Although the operational SSW field has been obtained mostly from the scatterometers or passive microwave radiometers described above, SSW speed has also been derived from satellite altimeters such as GFO (Geosat Follow-On), Jason-1, Envisat, Jason-2, Cryosat-2, and SARAL (Satellite with ARgos and ALtika), among others, which require accurate SSW speed to correct sea state bias and to obtain sea surface height with enhanced quality. The quality and error characteristics of satellite altimeter-derived SSW have been examined by comparing it with in-situ measurements and reanalyzed data in world seas (i.e., Dobson et al., 1987; Gourrion et al., 2002; Abdalla and Chiara, 2017), recently including the seas around the Korean Peninsula (Choi et al., 2018). In particular, the altimeter SSW speed was compared with in-situ measurements at the Ieodo Ocean Research Station (IORS) and from buoys around the Korean Peninsula for the 10 years from December 2007 to May 2016, and the results showed a bias and RMSE of 0.35 and 1.59 m s−1 and characteristic positive (negative) biases and higher (lower) altimeter SSW speed than in-situ speed at the low (high) speed regime (Table 1; Choi et al., 2018).

    Atmospheric Motion Vectors from Feature Tracking

    Atmospheric Motion Vectors (AMVs) are wind observations derived from satellite images by identifying a feature, i.e., clouds or gradients of water vapor, and later tracking it in consecutive images. The first AMV derivation algorithms were developed and AMVs were first produced in real-time in the 1960s-1970s, and AMVs have been assimilated operationally and used in numerical weather prediction since the 1980s- 1990s. The assimilation often has a positive impact on the predictability of operational forecast systems. The AMV derivation process consists of three main steps: selection of a suitable feature, calculation of the wind vector by measuring the displacement of the tracked feature in consecutive images, and assignment of height by converting the brightness temperature to pressure (Fig. 4). For example, an algorithm for deriving AMVs from the Geostationary Operational Environmental Satellite (GOES) R series (GOES-R) Advanced Baseline Imager (ABI) is detailed in Daniels et al. (2012).

    Presently, there are several operational AMV products provided by National Oceanic and Atmospheric Administration (NOAA), EUMETSAT, the Japan Meteorological Agency (JMA), and the Korea Meteorological Administration (KMA). For example, the operational AMVs include those from several channels (longwave infrared, shortwave infrared, water vapor, and visible channels) of GOES-16 (east), GOES-15 (west), and the S-NPP Visible Infrared Imaging Radiometer Suite (VIIRS). Similarly, EUMETSAT, JMA, and KMA provide various sets of AMV products. KMA’s follow-on geostationary meteorological satellite (Geostationary-Korea Multi-Purpose Satellite 2A or GEO-KOMPSAT-2A, GK-2A) will be launched in late 2018 and will have higher spatial and temporal resolution and a higher number of channels than its predecessor, the Communication, Ocean, and Meteorological Satellite (COMS) or GK-1. JMA’s next geostationary mission, Himawari-9, as a successor to Himawari-8, which has been operational since July 2015, will be operational until 2029. The AMVs from Himawari-8 in the European Center for Medium-range Weather Forecast (ECMWF) system were evaluated and compared to findings for its predecessor, MTSAT- 2 (Multifunction Transport Satellite-2) (Lean et al., 2017).

    The AMV derivation process has the potential to introduce errors into the observations. The two main contributors to the total AMV error are 1) tracking error-tracking the wrong feature during the recognition of consecutive images leads to an error in the displacement vector and 2) height assignment errorerrors in the model brightness temperature derivation and in the brightness temperature to pressure/altitude conversion lead to an error in the feature’s height. A more detailed discussion can be found in Forsythe and Doutriaux-Boucher (2005), Cotton and Forsythe (2010), and Lean et al. (2015). One of the difficulties in utilizing the AMVs is that the errors are difficult to characterize because they are complicated and spatially correlated. As the volume, coverage, and quality of AMV data have increased markedly, it is becoming increasingly important to improve the quality and the assimilation strategy.

    As the horizontal resolution of sensors onboard satellites becomes higher, it becomes possible to identify atmospheric motions induced by convective clouds with rapid scan modes (Kim et al., 2012). A high-resolution visible AMV algorithm for detecting mesoscale atmospheric motions, including ageostrophic flows, has been developed by the National Institute of Meteorological Research (NIMR), Republic of Korea as shown in Fig. 5 (Kim et al., 2012). To retrieve atmospheric motions smaller than meso-β scale effectively, the target size is reduced and the visible channel imagery of the geostationary satellite with 1 km resolution is used. For accurate AMVs, optimal conditions are determined by investigating the sensitivity of the algorithm to target selection and the correction method of height assignment. The results of such work have shown that the optimal conditions are target size of 32 km×32 km, grid interval that is the same as the target size, and the optimal target selection method. The AMVs derived with these conditions depict TCs more effectively, and the optimized mesoscale AMVs have been validated with radiosonde observations, indicating that the algorithm can successfully retrieve mesoscale atmospheric motions (Kim et al., 2012).

    The current operational AMVs have clear limitations, which are 1) the AMVs provide information at a single level of the troposphere, 2) height assignment is known to be an important problem, 3) there are only few vectors at mid-level and even fewer at lower level, and 4) recurrent AMV problems occur in the Tropics (fast speed biases), where important mesoscale phenomena impact the medium-range forecast. The challenges are 1) ensuring production, 2) providing better error characterization, 3) providing better information in the Tropics and for areas in and around TCs, and 4) providing vertical wind profiles. With limited resources at any one center, it is important that data providers and assimilators continue to work together, i.e., AMV intercomparison to develop a common quality index. The combination of a theoretical analysis of the limitations of the AMV derivation and further statistical comparisons of the AMV data with model backgrounds and other observations could help highlight parts of the AMV derivation that can be improved and could also be used to develop estimates of the vector and height errors. These errors can be used to improve the representation of the AMV errors in numerical weather prediction. Additionally, upcoming hyperspectral Infrared (IR) instruments on geostationary satellites, the Aeolus mission (introduced below in the next section), and the IR sounder Lidar mission will provide smallerscale information with high vertical and temporal resolution to feed regional models and threedimensional wind profiles in the near future. Looking toward the future, wind observations will remain an important part of the global observing system and AMVs are expected to remain an important source of wind data for numerical weather prediction for many years.

    Multiple-Platform Approaches for Tropical Cyclone Wind Retrievals

    Due to the fundamental limitations posed by sampling, the SSWs and AMVs in and around TCs are hardly monitored in a real-time and continuous manner using polar-orbit satellite sensors such as scatterometers, microwave radiometers, and altimeters. On the other hand, AMVs derived from geostationary satellites are still very sparse, limited to upper levels, and have low quality. An effort to develop a method to estimate objectively the wind fields associated with TCs or TC winds from multiple satellite platforms has been made, particularly considering the operational application. Global TC SSWs are produced every 6 h using only data from multiple satellite platforms and satellite-based wind retrieval techniques as a product of the Multiplatform Tropical Cyclone Surface Wind Analysis (MTCSWA) (Knaff et al., 2011). In addition, a new and improved method for estimating TC winds using globally and routinely available TC information and infrared satellite imagery has been developed by analyzing the aircraft reconnaissance data (1995-2012) collected within 165 km of a TC center and global model analyses data with a single-field principal component approach (Knaff et al., 2015). The MTCSWA currently provides satellite–based, operational, two-dimensional SSW estimates associated with TC on a real-time sense. For example, the SSW fields in and around typhoon Muifa on April 29, 2017 are derived from microwave sounder, geostationary satellite (AMVs and IR-based SSWs), and scatterometer (ASCAT) as shown in Fig. 6.

    More recently, a new method for estimating threedimensional TC wind fields from multiple satellite platforms has been developing with consideration of the upcoming operational GK-2A products. These ongoing efforts intend to merge four kinds of wind field derived from infrared sensor (GK-2A), microwave sounder (AMSU), scatterometer (ASCAT), and feature track AMVs (GK-2A), basically similar to the MTCSWA (Fig. 6), using weighting functions optimized to minimize the RMSE against reanalysis wind (ECMWF Year of Tropical Convection, YOTC) as a function of the distance from the TC center and altitude (Fig. 7). In Fig. 7, weight functions according to distance from the TC center (Fig. 7a) and altitude (Fig. 7b) for four different satellite platforms are shown as an example, i.e., potential application of GK-2A system. One of the additions to the MTCSWA is to incorporate vertical changes of maximum TC wind speed, radius of maximum TC wind, and size of TC warm core.

    Forthcoming Wind Measurement Sensor

    One of the science missions of the ESA within the Atmospheric Dynamics Mission (ADM), the Doppler wind Lidar (DWL) mission Aeolus, was launched in August 2018 and be operational in 2019, enabling the instrument named Atmospheric LAser Doppler INstrument (ALADIN), the only one to provide global and direct three-dimensional wind measurements from space (Fig. 8). The UV DWL operates at 355 nm, with a pulse repetition frequency of 50 Hz in continuous mode, and with 2 receiver channels. The testing of the ALADIN airborne demonstrator was conducted to demonstrate the potential performance of the spacecraft instrument. The ADM Aeolus mission is intended to have a minimum lifetime of three years. Its scientific objectives are to improve the quality of weather forecasts and to advance our understanding of atmospheric dynamics and climate processes. The explorer objective is to demonstrate the space-based DWL potential for operational use. The Aeolus mission will provide global profiles of high-quality wind observations in the troposphere and lower stratosphere in a near polar sun-synchronous orbit, providing great potential to improve numerical model prediction.

    ECMWF has been contracted by ESA to develop, in collaboration with the Royal Netherlands Meteorological Institute (KNMI), the Aeolus Level 2B/2C processing software, i.e., the wind retrieval, and intends to assimilate the Aeolus (Level 2B) L2B wind observations in ECMWF's global numerical weather prediction model as far as they are able to improve forecast skill. ECMWF has contributed to ESA-funded observation impact studies to assess the potential impact of the Aeolus data. The Aeolus mission is expected to have a positive impact on analysis and forecast quality, particularly tropical AMVs. Given that ALADIN is a high spectral resolution lidar, it is also possible to derive information useful for atmospheric composition modeling. Aeolus will also provide information on the atmosphere's cloud and aerosol optical properties (backscatter and extinction coefficients) via the Level 2A/2B/2C products.

    Concluding Remarks and Recommendations

    Satellite-derived SSWs and AMVs over the global ocean, particularly including the areas in and around TCs, have been provided in a real-time and continuous manner using more and better information derived from technologically improved multiple satellite missions and wind retrieving techniques. Examples include SSWs retrieved from scatterometers, passive microwave radiometers, SAR, and altimeters, as well as AMVs derived by tracking features from geostationary satellites. The data quality and error characteristics of individual wind observations have been examined by comparing them with in-situ measurements and reanalyzed products, and these need to be considered carefully to apply observations for both operational and scientific uses, i.e., assimilation in numerical weather forecasting. Here, the status and prospects of several key SSW and AMV products were reviewed and on-going efforts were introduced from the perspective of merging them, particularly for monitoring three-dimensional TC wind fields in a real-time and continuous manner and for providing global profiles of high-quality wind observations with the new ADM Aeolus mission.

    Unceasing and coordinated efforts should be devoted to creating more and better synergies among efforts and resources, which are limited at any one center. The AMV intercomparison to develop a common quality index is one such example. Both theoretical analyses of the limitations of AMV derivation and statistical comparisons of AMV data with other observations and model backgrounds are important for developing estimates of vector and height errors. The new ADM Aeolus mission and upcoming geostationary missions (e.g., GK-2A, Himawari- 9) provide improved information for retrieving more and better SSW and AMV products in a real-time and continuous manner. Future research needs to follow up by providing 1) ensured quality and error characterization of SSWs and AMVs, 2) better information in the tropics and for areas in and around TCs, 3) vertical wind profiles or three-dimensional wind fields, and 4) smaller-scale information on wind fields with high vertical and temporal resolution.


    This study was supported by National Meteorological Satellite Center (NMSC) of Korea Meteorological Administration (KMA) via “Development of Typhoon and Ocean Applications” program (NMSC-2018-01). Partial supports are also from the project titled “Deep Water Circulation and Material Cycling in the East Sea” funded by the Ministry of Oceans and Fisheries, Korea.



    A snapshot of sea surface winds (SSWs) of typhoon Soulik while at Category 4 intensity captured by NOAA’s advanced scatterometer (ASCAT) program. Source: Wikipedia, Scatterometer_Ascending_Pass.png.


    Spatial distribution of (a) wind speeds (m/s), (b) wind vectors from QuikSCAT data during the period of typhoon Rusa in 2002, and (c) wind vectors enlarged from (b) near the center of the typhoon.


    (a) A RADARSAT-1 SAR image taken on July 11, 2007 in which the location of typhoon eye (circle in color with wind scale in Saffir-Simpson scale) and profiling floats (W1, W2, and E1) before (red rectangles) and after (blue triangles) a typhoon’s passage are superimposed with UTC data (and time for eye position). (b) SSW vectors (arrows) and speed in EW direction (color) estimated from the RADARSAT-1 SAR image. Source: Nam et al. (2012).


    A generalized flow chart of the geostationary satellite imager algorithm for deriving AMVs.


    Comparison of visible channel (a) synoptic AMVs and (b) mesoscale AMVs for typhoon OMAIS, 2315 UTC on March 23, 2010. The color of AMVs indicates the height, as explained by the color scale shown at the bottom. Source: Kim et al. (2012).


    Two-dimensional wind fields in and around typhoon Muifa, 1200 UTC on April 29, 2017. (a, b) Multi-platform wind fields, and wind fields derived from individual platforms: (c) microwave sounder (AMSU), (d) geostationary satellite – featuretrack AMVs, (e) geostationary satellite – infrared-based SSWs, and (f) scatterometer (ASCAT). Here, red dots in (a) and (b) denote the location of typhoon center. Areas where the SSW speed is higher than 20 m/s are remarked with green arrows in (a) and (b). Source: RAMMB, CIRA, and NESDI, NOAA (


    Weighting functions according to (a) distance from the TC center and (b) altitude for four different platforms: infrared (IR, red), microwave sounder (MW, cyan), geostationary satellite feature-track wind (AMV, green), and scatterometer-based SSWs (ASCAT, blue).


    (a) Aeolus measurement geometry. The wind is observed orthogonal to the satellite ground-track, pointing 35 o off-nadir, away from the Sun. (b) Lidar concept. The state-of-the-art ALADIN instrument incorporates two powerful lasers, a large telescope, and very sensitive receivers. The fraction of light that is scattered back toward the satellite is collected by ALADIN’s telescope and measured. Source: ESA,


    Bias and root-mean-squared (RMS) difference or RMS error (RMSE) of satellite SSW data retrieved from scatterometer (ASCAT), microwave radiometers, such as SSM/I and Advanced Microwave Scanning Radiometer (AMSE)-2, and several altimeters. Source data for ASCAT, SSM/I, AMSE-2, and altimetry are from Chou et al. (2013), Mears et al. (2012), Hong et al. (2016), and Choi et al. (2018), respectively. Wind speed is given in m s−1 . Direction is given in degrees clockwise (value in parentheses). IORS denotes the Ieodo Ocean Research Station


    1. Choi, D.Y. , Woo, H.-J. , Park, K.-A. , Byun, D.-S. , and Lee, E. , 2018, Validation of sea surface wind speeds from satellite altimeters and relation to sea state bias . Journalof the Korean Earth Science Society, 39 (2), 139-153.
    2. Chou, K.-H. , Wu, C.-C. , and Lin, S.-Z. , 2013, Assessment of the ASCAT wind error characteristics by global dropwindsonde observations . J. Geophys. Res., 118, 9011-9021.
    3. Cotton, J. and Forsythe, M. , 2010. Fourth analysis of the data displayed on the NWP SAF AMV monitoring website, NWP SAF Technical Report 24.
    4. Daniels, J. , Bresky, W. , Wanzong, S. , Velden, C. , and Berger, H. , 2012, GOES-R Advanced Baseline Imager (ABI) algorithm theoretical basis document for derived motion winds, ver. 2.5. GOES-R Program Office, 96 pp.
    5. Figa-Saldaña, J. , Wilson, J.J.W. , Attema, E. , Gelsthorpe, R. , Drinkwater, M.R. , and Stoffelen, A. , 2014, The advanced scatterometer (ASCAT) on the meteorological operational (MetOp) platform: A follow on for European wind scatterometers . Canadian Journal of Remote Sensing, 28 (3), 404-412.
    6. Forsythe, M. and Doutriaux-Boucher, M. , 2005. Second analysis of the data displayed on the NWP SAF AMV monitoring website. NWP SAF Technical Report 20. Hersbach, H., 2010, Comparison of C-band scatterometer CMOD5. N equivalent neutral winds with ECMWF . Journal of Atmospheric and Oceanic Technology, 27 (4), 721-736.
    7. Hong, S. , Seo, H.-J. , and Kown, Y.-J. , 2016, A unique satellite-based sea surface wind speed algorithm and its application in tropical cyclone intensity analysis . J. Atmos. Oceanic Tech., 33, 1363-1375.
    8. Isoguchi, O. and M. Shimada, 2009. An L-band ocean geophysical model function derived from PALSAR , IEEE Transactions on Geoscience and Remote Sensing, 47(7): 1925-1936.
    9. Kim, T.-S. , Park, K.-A. , Choi, W.-M. , Hong, S. , Choi, B.- C. , Shin, I. , and Kim, K.-R. , 2012, L-band SAR-derived sea surface wind retrieval off the east coast of Koreaand error characteristics . Korean Journal of Remote Sensing, 28 (5), 477-487.
    10. Knaff, J.A. , DeMaria, M. , Molenar, D.A. , Sampson, C.R. , and Seybold, M.G. , 2011, An automated, objective, multiple-satellite-platform tropical cyclone surface wind analysis . J. Meteorol. Climatol., 50, 2149-2166.
    11. Knaff, J.A. , Longmore, S.P. , DeMaria, R.T. , and Molenar, D.A. , 2015, Improved tropical-cyclone flight-level wind estimates using routine infrared satellite reconnaissance . J. Appl. Meteorol. Climatol., 50, 2149-2166.
    12. Lean, K. , Bormann, N. , and Salonen, K. , 2017, Assessment of Himawari-8 AMV data in the ECMWF system. EUMETSAT/ECMWF Fellowship Programme Research Report 42.
    13. Lean, P. , Migliorini, S. , and Kelly, G. , 2015, Understanding atmospheric motion vector vertical representativity using a simulation study and first-guess departure statistics . J. Appl. Meteorol. Climatol.54, 2479-2500.
    14. LehnerS. , PleskachevskyA. , and BruckM. , 2012, Highresolution satellite measurements of coastal wind field and sea state , International Journal of Remote Sensing,33 (23), 7337-7360.
    15. Mears, C.A. , Smith, D.K. , and Wentz, F.J. , 2001, Comparison of Special Sensor Microwave Imager and buoy-measured wind speeds from 1987 to 1997 . J.Geophys. Res., 106, 11,719-11,729.
    16. Metcalfe, D.J. , Bradford, M.G. , and Ford, A.J. , 2008, Cyclone damage to tropical rain forests: Species- and community-level impacts . Austral Ecol., 33, 432-441.
    17. Nam, S.H. , Kim, D.-J. , and Moon, W.M. , 2012, Observed impact of mesoscale circulation on oceanic response to typhoon Man-Yi (2007) . Ocean Dynamics, 62 (1), 1-12.
    18. Powell, M.D. and Reinhold, T.A. , 2007, Tropical cyclone destructive potential by integrated kinetic energy . Bull.Am. Meteorol. Soc., 88, 513-526.
    19. Soisuvarn, S. , Jelenak, Z. , Chang, P. S. , and Alsweiss, S. O. , 2013, A high wind Geophysical Model Function for C-band vertically polarized satellite scatterometermeasurements . IEEE Trans. Geo. Rem. Sen., 51, 3744-3760.
    20. Stoffelen, A. , and Anderson, D. , 1997, Scatterometer data interpretation: Estimation and validation of the transfer function CMOD4 . Journal of Geophysical Research:Oceans, 102 (C3), 5767-5780.
    21. Zhang, Q. , Wu, L. , and Liu, L. , 2009, Tropical cyclone damage in China 1983-2006 . Bull. Am. Meteorol. Soc.,99, 489-495.