Testing triple oxygen isotope preservation in the new OPEnS totalizer against conventional monthly rainfall collectors

Abstract

Understanding the mechanisms that drive spatial and temporal triple oxygen isotope (Δ′17O) variations in modern precipitation is the first step to expanding the utility of these measurements as an environmental tracer jointly with traditional stable isotope parameters δD, δ18O, and d-excess. However, “totalizers” designed to collect a single precipitation sample pooled over a calendar month and minimize evaporation and associated isotopic fractionation of the sample during that time have not been tested for Δ′17O. We conducted a 30-day laboratory experiment comparing mass losses and isotopic shifts in four totalizers: 1) the OPEnS (Openly Published Environmental Sensing) totalizer, 2) the classic oil-based totalizer, 3) the commercial tube dip-in/pressure equilibration totalizer (Palmex Ltd. RS1), and 4) a reference totalizer (the control, lacking any evaporation reduction mechanism). The OPEnS totalizer was designed as being readily user built with parts costs of about $10, oil-free to facilitate quick and easy sample preparation and risk-free sample analysis, and its collection device expands as it fills to maintain a small gas/water ratio and minimize internal evaporative losses. All totalizers were filled to 12 % of their 3-L volume and placed in a modified laboratory oven with a diurnal temperature change of 23 to 40 °C and an average relative humidity of 9.1 % to simulate extreme evaporative conditions. The OPEnS totalizer experienced the smallest mass loss of water (0.21 %) and smallest isotopic shifts (p < 0.05 for δ18O and d-excess), which were all within measurement error. The oil, tube, and reference totalizers showed larger mass losses (0.41, 1.37, and 1.61 %, respectively) and evaporative enrichment with respect to δD (+0.3, +0.8, and + 2.1 ‰), δ18O (+0.16, +0.23, and + 0.83 ‰), and d-excess (−0.9, −1.0, and − 4.5 ‰). The Δ′17O variations for all totalizers were within measurement error, so we suggest that in less harsh climates their triple oxygen isotope changes during secondary evaporation would be more acceptable. To test the OPEnS totalizer in field settings, we installed it alongside oil totalizers to collect monthly precipitation over three years in the towns of Jolly and San Antonio, Texas, with mean annual precipitation, temperature, and windspeed values of 556 and 563 mm, 18.7 and 21.9 °C, and 5.0 and 3.5 m/s, respectively. Results indicate that the OPEnS and oil totalizers can produce similar isotopic data in the field, but modifications to OPEnS have been implemented to minimize under-catch and stabilize the collection component where high winds are present and additional testing under a variety of environmental conditions is ongoing. OPEnS is scalable according to expected monthly precipitation amounts, providing a cost-effective, high-performance device for quantification of total rainfall and its isotopic composition without oil contamination risks.