Scientific Papers

Abstract: New technologies can help scientists measure and understand Arctic warming, sea ice loss and ecosystem change. NOAA has worked with Saildrone, Inc., to develop an unmanned surface vehicle (USV)—Saildrone—to make ocean surface measurements autonomously, even in challenging high-latitude conditions. USVs augment traditional research ship cruises, mitigate ship risk in high seas and shallow water, and make lower cost measurements. Under remote control, USV sampling strategy can be adapted to meet changing needs. Two Saildrones conducted 97-day missions in the Bering Sea in spring-summer 2015, reliably measuring atmospheric and oceanic parameters. Measurements were validated against shipboard values. Following that, the Saildrone sampling strategies were modified, first to measure the effects of sea-ice melt on surface cooling and freshening, and then to study the Yukon River plume.

In 2021, a novel NOAA-Saildrone project deployed five uncrewed surface vehicle Saildrones (SDs) to monitor regions of the Atlantic Ocean and Caribbean Sea frequented by tropical cyclones. One of the SDs, SD-1045, crossed Hurricane Sam (Category 4) on September 30, providing the first-ever surface-ocean videos of conditions in the core of a major hurricane and reporting near-surface winds as high as 40 m/s. Here, we present a comprehensive analysis and interpretation of the Saildrone ocean surface wind measurements in Hurricane Sam, using the following datasets for direct and indirect comparisons: an NDBC buoy in the path of the storm, radiometer tropical cyclone (TC) winds from SMAP and AMSR2, wind retrievals from the ASCAT scatterometers and SAR (RadarSat2), and HWRF model winds. The SD winds show excellent consistency with the satellite observations and a remarkable ability to detect the strength of the winds at the SD location. We use the HWRF model and satellite data to perform cross-comparisons of the SD with the buoy, which sampled different relative locations within the storm. Finally, we review the collective consistency among these measurements by describing the uncertainty of each wind dataset and discussing potential sources of systematic errors, such as the impact of extreme conditions on the SD measurements and uncertainties in the methodology.