We present spatially highly resolved distributions of dissolved seawater REE concentrations ([REE]) along three transects in the zonal (extra-)equatorial current system and the Solomon Strait of the Tropical Western Pacific. Seawater samples were collected using Niskin bottles and filtered through AcroPak500 filter cartridges (0.8/0.45 μm pore size) directly from the Niskin bottles or in the onboard laboratory of R/V L'Atalante cruise CASSIOPEE (GEOTRACES compliant data GPc05) in July-August 2015. We use seawater [REE] in combination with direct physical oceanographic observations to characterize the geochemical composition, origin and pathways of the complex surface and upper layer currents of the Tropical Western Pacific and to quantify the input fluxes of REEs in our study area.

Seawater REEs were purified and pre-concentrated using the automated seaFAST-pico system in offline mode (Elemental Scientific Inc., Nebraska, USA) and measured by isotope dilution (ID) inductively coupled plasma-mass spectrometry (ICP-MS) following the method described in Behrens et al. (2016). Seawater volumes of ~11-55 mL (pH = 2) per sample were spiked with a multi-element spike and passed together with a buffer solution (pH = 6) and MilliQ water through the seaFAST column containing a REE-complexing resin (Nobias PA-1) that allows to wash out the seawater matrix. Total procedural onboard blanks of (onboard) MilliQ water and lab blanks were processed through the seaFAST column and subsequently spiked with a diluted multi-element REE isotope spike for quantification.Dissolved REE concentrations were analyzed using a Thermo Finnigan Element 2 ICP-MS coupled to an autosampler (CETAC ASX-100) and a desolvation introduction system (CETAC Aridus 2) (Behrens et al., 2016). Average instrumental blanks of 2% HNO3 were subtracted for each sample. The accuracy was checked with replicates of the GEOTRACES seawater standard SAFe 3000 m and average [REE] agreed within the 2 RSD confidence interval of the published intercomparison study (Behrens et al., 2016). The external error (≤2.3% for all REEs, except for Ce with 8.2%, 1 RSD) is derived from independently processed seawater replicates from a sample from station 50 at 1001 m water depth (sample 50-2-8, n = 4). If the internal error of a sample was higher than this external error, the internal error is reported. Average total procedural onboard (n = 8) and lab blanks (n = 14) were ≤1.9% for all REEs, except for Ce (4.9%, n = 20, and 28%, n = 2) of the average sample concentration.