This dataset contains records of planktic foraminiferal δ¹⁸O, bulk sedimentary δ¹⁵N, alkenone concentrations and opal content of sediment core MD01-2380, which was retrieved in the central Banda Sea (5°45.64' S, 126°54.25' E, 3232 m water depth). These records cover the last ⁓800 kyrs. Additionally, the dataset contains a record of bulk sedimentary δ¹⁵N data of MD06-3067, which was retrieved offshore Mindanao, covering the last ⁓160 kyrs and bulk sedimentary δ¹⁵N records of cores MD98-2181, MD98-2177, MD97-2138, MD05-2920, and MD98-2162 that cover the last 25 kyrs. Previously published and new δ¹⁵N records from the western equatorial Pacific were used to generate northern and southern hemisphere stack records for the last 25 kyrs, which are provided as well. The separation of cores into a southern and northern hemisphere stack is based on hydrographic consistency, not latitude per se (for details see Kienast et al. (in revision)). Prior to the calculation of the δ¹⁵N stack records, all time series were interpolated to even time steps of 500 yrs. The northern and southern hemisphere stacks were then calculated as time-dependent mean of the individual records with the corresponding standard deviation.

Stable oxygen isotope analysis on planktic foraminifera Globigerinoides ruber tests (size fraction 250-355 µm) were performed at MARUM, University of Bremen using a Finnigan MAT 251 mass spectrometer with Kiel I devices. The internal carbonate standard is a Solnhofen Limestone, which is calibrated to the National Bureau of Standards (NBS) 19 standard. The long-term analytical precision is better than ±0.07‰. Sedimentary δ¹⁵N was analyzed at the Yale Analytical and Stable Isotopic Center YASIC (Yale University), using a Costech ECS 4010 Elemental Analyzer with Conflo III interface (core MD01-2380) and at the Pacific Centre for Isotopic and Geochemical Research (University of British Columbia), using a Carlo Erba NC 2500 elemental analyzer coupled to a Finnigan Mat Delta Plus mass spectrometer (cores MD98-2181, MD06-3067, MD98-2177, MD05-2920, MD98-2162, MD97-2138). The precision of the isotopic analyses based on in-house standard measurements of ±0.2 ‰.Biogenic opal of MD01-2380 was determined at Dalhousie University following methods developed by Mortlock and Froelich (1989) doi:10.1016/0198-0149(89)90092-7. Alkenone concentrations were measured at the National Taiwan Ocean University, following the method of Yamamoto et al. (2000) doi:10.1016/S0146-6380(00)00080-2.Along with published records of MD01-2386 (Jia et al., (2011) doi:10.1029/2011gl050021; Jian et al. (2020), doi:202010.1073/pnas.1915510117) and U1486 (Lambert et al., (2022) doi:10.1016/j.quascirev.2021.107348), the δ15N records of MD06-3067, MD98-2181, MD98-2177, MD97-2138, MD05-2920, and MD98-2162 were used to generate northern and southern hemisphere stack records, which are provided as well. The separation of cores into a southern and northern hemisphere stack is based on hydrographic consistency, not latitude per se (for details see Kienast et al. (in revision)). Prior to the calculation of the δ¹⁵N stack records, all time series were interpolated to even time steps of 500 yrs. The northern and southern hemisphere stacks were then calculated as time-dependent mean of the individual records with the corresponding standard deviation.