Reconciling Hurricane Hunter, Ascat and Buoy Ocean Wind Observations Under High and Extreme Wind Conditions

2019 Joint Satellite Conference, 28 September - 4 October 2019, BostonExtreme wind references are an essential requirement for assessing the capability of satellite data and for improving wind retrievals in extreme wind conditions. Measuring reliable high and extreme winds in hurricanes from wind scatterometers is challenging as vicarious calibration is needed and calibrated in-situ wind references are limited. The EUMETSAT-funded “C-band High and Extreme-Force Speeds (CHEFS)” project aims at assessing the extreme wind capabilities of the next generation of C-band wind scatterometers onboard Metop-SG. One important step within the EUMETSAT-funded CHEFS project was to provide an appropriate high and extreme-wind reference at scatterometer scales. For very high and extreme winds above 25 m/s, moored buoy measurements are not reliable while the National Oceanic Atmospheric Administration (NOAA) hurricane hunter observations are used as reference by the International Ocean Surface Vector Winds Science Team (IOVWST) community. Moreover, controversy exists in the IOVWST on the quality of moored buoys above 15 m/s rather than 25 m/s. Therefore, in order to build a consolidated wind reference data set in high and extreme wind conditions, an inter-comparison between buoys and hurricane hunter winds would be useful to better understand the existing inconsistencies. However, in this work, since there is a lack of buoys and hurricane hunter collocations, we use ASCAT to bridge the comparison between the former data sets. As such, we use ASCAT collocated with moored buoys on the one hand and with hurricane hunter observations on the other. A thorough comparison of 5 years of moored buoy measurements from the different tropical arrays (i.e., TAO, TRITON, PIRATA, RAMA) and ASCAT wind collocations is first carried out. Moored buoy data that are distributed along the Global Telecommunications System, rather than those directly available from each data provider (the so-called continuous wind data), have been used in this analysis as they have proved to have more high wind events with respect to the continuous wind data sets. Comparisons have been performed by using the Advanced SCATterometer (ASCAT) 10 m surface winds at 12.5 km and 25 km resolution. The results have shown that ASCAT winds are slightly lower than buoy winds at wind speeds higher than 17 m/s, regardless of the anemometer height and/or type. Similar results are obtained when comparing buoys with the European Centre for Medium-Range Weather Forecasts (ECMWF) model winds. Triple collocation analysis of buoy/ASCAT/ECMWF winds has been carried out to assess the performance of buoys and no significant buoy wind degradation is observed at winds ranging between 15 m/s and 25 m/s. Detailed wind patterns as well as in-situ wind profiles in hurricane wind conditions are obtained from the Stepped-Frequency Microwave Radiometer (SFMR) and Global Position System (GPS) dropsondes, respectively. These data are collected by the NOAA and US Air Force hurricane hunters during hurricane seasons and winter campaigns. A statistical analysis of the SFMR and dropsonde differences has been performed in order to detect and filter possible outliers. The results show that the so-called WL150 algorithm used to estimate 10 m surface winds from the dropsonde wind profiles may have an impact in the SFMR-dropsonde statistics, therefore this algorithm has been thoroughly evaluated. In addition, the SFMR spatial representativeness error has been also assessed by aggregating SFMR winds at different spatial (temporal) scales up to 25 km footprint, for SFMR-ASCAT inter-comparison purposes. The results show that the standard deviation of the SFMR-dropsonde difference slightly decrease when averaging SFMR winds along track. Different study cases of extreme wind conditions will be collected and a collocation criterion will be defined in order to analyze 12.5 km ASCAT high winds with respect to storm-motion-centric SFMR and dropsonde 10 m surface wind estimates. ASCAT surface winds will subsequently be used as a bridge allowing an inter-comparison between SFMR/dropsonde and buoys at scatterometer scale. The main results of this study will be presented at the conferencePeer reviewe