Uncertainty in palaeosalinity estimates from foraminiferal geochemical records in the northern Indian Ocean

Palaeosalinity (S*) estimates rely on calculations stemming from the oxygen isotopic composition of foraminiferal shells (δ¹⁸O_F) preserved in ocean sediments. δ¹⁸O_F depends on palaeo in-situ temperature (T*) and the oxygen isotopic composition of seawater (δ¹⁸O_sw), which in turn depends on various physical oceanographic conditions including evaporation, precipitation, advection, and freshwater fluxes. δ¹⁸O_sw and salinity covary linearly owing to similar controls, as they both increase with increases in evaporation and decrease with higher precipitation and continental runoff. Typically, palaeosalinity is estimated assuming a constant spatial and temporal relationship between δ¹⁸O_sw and salinity. However, evaporative and freshwater fluxes exhibit spatiotemporal variability and thus, can change the slope and intercept of the linear δ¹⁸O-salinity relationship. Hence, the use of a constant δ¹⁸O_sw-salinity relationship may produce (systematic) errors in palaeosalinity estimates. Here, we revisit palaeosalinity estimates and provide a reassessment of errors arising from these calculations based on recently reported δ¹⁸O_sw-salinity relationships in the northern Indian Ocean. Our calculations point to errors in palaeosalinity that can be as large as 55% - significant enough to change our understanding of past ocean circulation and palaeomonsoon variability. These results have important implications for reconstructing the palaeoceanographic record and interpreting past ocean-atmosphere variability.