On the reliability of tube-dip-in-water precipitation collectors in isotope hydrology: A field experiment for low rainfall amounts

The oxygen and hydrogen stable isotope composition in precipitation serves as a benchmark in most isotope atmospheric, (eco-) hydrological, and paleoclimatological applications. Several rain collectors have been designed for collecting monthly, daily, or event-based precipitations aiming to prevent evaporation and associated isotope fractionation. Oil collectors have been the most widely used for many years and only recently they are being replaced by free-oil Tube-dip-in-water collectors, especially after the production of commercial versions. Although several isotopic studies already adopted this precipitation collector, no field experiments were performed to evaluate the isotopic integrity of low-amount precipitation samples. Field testing of precipitation collectors is therefore encouraged under the same environmental conditions of areas where researchers want to undertake their studies. In this work, we tested field performances of different Tube-dip-in-water collectors, compared to an Oil collector and a reference collector with no anti-evaporative system, in preventing evaporation during the collection of low-amount monthly precipitation samples. The field experiment was carried out in the city of Pisa (Italy), in the Mediterranean region, over four different periods in spring and summer 2022, when atmospheric conditions were more likely to promote evaporation. Six precipitation collectors were filled at the beginning of each period with distilled water with a known isotope composition. Four collectors (reference, Oil, Palmex Tube-dip-in-water, Home-made Tube-dip-in-water) were filled to less than 2 % of the total volume, and other two Home-made Tube-dip-in collectors were filled to 5 % and 10 %. Evaporative mass losses were determined gravimetrically at the end of each period, and water aliquots were sampled for isotope analyses. The Oil collector showed the smallest evaporative mass losses and no detectable isotope shifts. The Home-made Tube-dip-in-water collector was also effective in preventing evaporation for low water amounts and it provided acceptable results even for extremely low water volumes, albeit with small isotope shifts. The Palmex collector, by contrast, exhibited worst performance when the collection of very low amount precipitation samples (less than 2 % of the total volume of the bottle) was tested. The isotope shifts were significant and larger than values measured for the Home-made collector filled to same percentage. Among the possible reasons behind the worse performance of this sampler, the outer case material seemed to be the most likely. The Palmex sampler was enclosed in a metal outer case which could have promoted an overheating of the bottle, when exposed to solar radiation, potentially enhancing the day-night temperature fluctuations inside the bottle and between the bottle and the case. This could have led to greater evaporation but potentially also to larger expansion-contraction cycles of the gas inside the bottle, thus promoting vapour exchange with the external atmosphere. However, further experiments will have to be carried out to confirm this hypothesis. Overall, our results indicated Tube-dip-in-water collectors as a reliable and low-cost alternative to oil models in most studies of isotope hydrology, especially in regions with temperate to semi-arid climate, including Mediterranean areas. However, a prudent approach is recommended at the sampling stage whereby the choice of the collector design should be calibrated according to the climatic features of the area and seasonality.