Page 68 - Summer2022

RAINFALL AT SEA Figure 6. The ambient ocean sound from 0, 165°E. a: Convective rain produces strong sound signals from 1 to 35 kHz. The stratiform rain (drizzle) has a peak signal at 15 kHz. b: Drop size distribution (DSD) using acoustic inversion. c: Rainfall rate is derived using the single-frequency conversion and full-inversion methods. Nystuen (2001) first computed an acoustic inversion for the DSD based on field data collected in a shallow brackish pond in Florida. Applying the same inversion algorithm to open ocean data required a frequency-inde- pendent adjustment of the transfer function (accounting for shallow-water reverberation in the pond and a lower initial acoustic pressure of large bubbles in saltwater com- pared with freshwater). When new methods to measure the DSD directly in the open ocean are developed, a new transfer function can be calculated. The DSD rainfall estimates based on the sound-intensity relationship and inversion diverge as rain changes from convective to stratiform. The simple relationship between sound intensity and rain rate implicitly assumes a DSD shape typical of convective rainfall. Stratiform rainfall has relatively fewer small- and medium-sized raindrops, and thus the sound-intensity relationship overestimates stratiform rainfall rates. Using the DSD inversion method improves the agreement between acoustic DSD estimates and surface rain gauges. The result suggests that the full DSD acoustic inversion for rainfall rate should be used when mixed stratiform/convection rainfall events are being measured acoustically (Figure 6). More Platforms for Collecting Passive Acoustic Data Over the last two decades, advanced observation plat- forms entered service to survey the global ocean, all of which could be considered components of GOOS. The need to remotely sense the air-sea interface from under- water led to the addition of hydrophones to collect ocean ambient-noise data. Besides the moorings that have hosted PALs, various autonomous platforms have been developed for acoustic rain and wind measurements, including Argo floats (Yang et al., 2015), seagliders (Ma et al., 2018), and cabled seafloor observing systems. These platforms provide near-real-time acoustic spectra data and some also record raw acoustic time series, which offers opportunities to extract useful ocean information using passive acoustic methods. Seagliders are underwater autonomous vehicles about 2 m in length that use internal changes in buoyancy to “drive” the vehicle to cycle up and down in the water column. The seaglider has small, fixed wings that cause it to move horizontally as it moves up and down. As a result, a seaglider glides horizontally while zigzagging up and down, usually between the surface and 1,000 m in depth. At the surface, the seaglider uses satellite com- munications to send its data back to shore. In an experiment in the tropical Pacific (Lindstrom et al., 2017), ambient-noise data were acquired using a hydro- phone system on the seagliders. The acoustic data were processed and averaged into 7 frequency bins from 1 to 55 kHz and transmitted to shore. Using the rule-based detection method, various rain and wind conditions were distinguished and classified using ambient-noise spectra shapes. Rain noise was still detected acoustically to 1,000 m, the maximum depth of the seaglider (Figure 7). To test the validity of the seaglider data, simultaneous rain measurements were obtained from a buoy rain gauge and satellite rain rate products. The comparison showed small discrepancies between different measurement methods due to differing spatial and temporal sampling schemes. Although it was difficult to compare rainfall rates and rain events, the seasonal accumulations were in agree- ment. The instantaneous acoustic method has advantages (higher temporal resolution and larger effective surface- sampling area) compared with conventional rain gauges. 68 Acoustics Today • Summer 2022