Future ocean increasingly transparent to low-frequency sound owing to carbon dioxide emissions

Low-frequency sound in the ocean is produced by natural phenomena such as rain, waves and marine life, as well as by human activities, such as the use of sonar systems, shipping and construction. Sea water absorbs sound mainly owing to the viscosity of the water and the presence of chemical constituents, such as magnesium sulphate, boric acid and carbonate ions. The concentration of dissolved chemicals absorbing sound near 1 kHz depends on the pH of the ocean1, which has declined as a result of increases in acidity due to anthropogenic emissions of carbon dioxide2–4. Here we use a global ocean model5,6 forced with projected carbon dioxide emissions7 to predict regional changes in pH, and thus sound absorption, in the years 1800–2300. According to our projections, ocean pH could fall by up to 0.6 units by 2100. Sound absorption—in the range between ∼100 Hz and ∼10 kHz—could fall by up to 60% in the high latitudes and in areas of deep-water formation over the same time period. We predict that over the twenty-first century, chemical absorption of sound in this frequency range will nearly halve in some of the regions that experience significant radiated noise from industrial activity, such as the North Atlantic Ocean. We suggest that our forecast of reduced sound absorption in acoustic hotspots will help in identifying target regions for future monitoring. Low-frequency ocean sound in the range 1–5,000Hz is produced both by natural phenomena, for example, rain, waves or marine life (Supplementary Table S1), and by human activities such as sonar systems, shipping or construction (Supplementary Table S2). Most low-frequency noise is generated at the surface8. It can enter the deep sound channel either in high latitudes where the sound channel axis depth shoals as a result of latitudinal variations of the sound speed9, or at continental shelves and slopes owing to reflection and scattering of sound waves by downslope conversion10. In the deep sound channel (typically at a depth of 700m in theNorth Pacific Ocean), soundwaves can bend and travel over many thousands of kilometres8. The oceanic penetration of the fossil fuel CO2 signal can now be observed well below such depths, and thus can affect the acoustic properties of the ocean over very large distances. Seawater sound attenuation, beyonddirect geometrical spherical and cylindrical energy dispersal, and bottom losses, is mainly driven by two mechanisms: absorption owing to fluid shear viscosity and owing to chemical resonances of some constituents of sea water, such asmagnesium sulphate and the boric acid/carbonate system11. The contribution of these twomechanisms varies with sound frequency and ambient conditions, for example,