Picarro Blog

Identifying the fingerprint of oceanic evaporation from sea surface observations between the North Pole and the coasts of Antarctica

At Picarro, we enjoy living vicariously through the scientists and research groups that use our systems in their projects. The following study takes us through the voyage of the Polarstern ice breaker of the Alfred Wegener Institute in Germany. On behalf of everyone at Picarro, we would like to thank Dr. Jean-Louis Bonne, Dr. Melanie Behrens, Dr. Hanno Meyer, Dr. Sepp Kipfstuhl, Dr. Benjamin Rabe, Dr. Lutz Schönicke, Dr. Hans Christian Steen-Larsen, and Dr. Martin Werner for sharing their journey with us.

Water stable isotopes are commonly exploited in various types of archives for their information on past climate evolutions. Ice cores retrieved from polar ice sheets or high-altitude glaciers are probably the most famous type of climate archives (Jouzel, 2014). In permafrost-covered region, new types of archives based on ice-wedges have been recently developed, extending the spatial coverage of water-isotopes based climate proxies (Meyer et al., 2015, 2010). Ice-wedges are a common type of ground ice conserved within permafrost as vertically oriented foliations of the ice body.  In ice cores as well as ice-wedges, a message about past temperature variations is conserved in the ice, formed from the snow falls whose isotopic compositions vary with the temperature. However, deciphering these past temperature variations is not always straightforward, as the temperature is not the only parameter which can have an imprint on the isotopic composition of the ice. A better understanding of the mechanisms affecting water isotopic composition, such as the changes in origin of the moisture, can be obtained from present-day observations and contribute to the interpretation of such paleoclimate archives. 

Water isotopes are useful tools to understand the water cycle, as they contain information on the phase changes a water mass has undergone. Many studies use water isotopes in complex atmospheric models to refine the analysis of paleoclimatic data. The same models are also used for future climate projections, a domain where large uncertainties are still linked to the prediction of the water cycle and the changes in precipitation patterns. Water isotopes can be useful to benchmark such models and contribute to their improvement.

Polarstern ice breaker

The Polarstern ice breaker in the bay of Longyearbyen, Svalbard, in July 2015.

Our group at the Alfred Wegener Institute in Germany has been focusing on the first step of the atmospheric water cycle: the evaporation at the oceanic surface. To better understand the processes affecting the water isotopic composition during this crucial step, we have installed a Picarro analyzer L2140-i on-board the German research ice breaker Polarstern. This instrument has been continuously recording the water vapour isotopic composition (δ18O and δ2H) since summer 2015 directly above the ocean surface, over all latitudes in the Atlantic sector, from the North Pole up to the coast of Antarctica (Figure 1). We used δ18O and δ2H values to calculate the second order parameter deuterium excess, defined as d-excess = δ2H – 8× δ18O (Dansgaard, 1964). Our observations, newly published in Nature Communications, allowed us to experimentally explore the interactions between the atmospheric moisture and the open sea as well as the sea ice.


Figure 1. Water vapour isotopic observations on-board Polarstern

Figure 1. Location of water vapour isotopic observations recorded on-board Polarstern, from 2015-06-29 to 2017-07-01

According to a commonly cited, 40 year-old theory (Merlivat and Jouzel, 1979),  the meteorological conditions under which the oceanic evaporation takes place leave their fingerprint in the isotopic composition of the vapour. We were able to test this theory for the first time in the field under a large range of climatic conditions. Our observations indeed confirm the expected role of relative humidity and sea surface temperature in the isotopic composition of the evaporated flux. However, contrary to what was expected from this theory, we observed no effect of wind speed on the boundary layer vapour isotopic composition above the oceans (Figures 2 and 3). The water isotope simulations of an atmospheric model (ECHAM5-wiso) showed a better agreement with the observations when no wind effect was considered.

Figure 2. Theoretical Effect of wind speed

Figure 2. Theoretical effect of wind speed on the vapour isotopic signal d-excess, considering the Merlivat and Jouzel 1979 theory (left panel) compared to observations (right panel). Distinct distributions of d-excess against relative humidity at the sea surface are expected from the theory between low and high wind speeds (below or above 7 m/s, respectively orange and blue), but no difference is observed on the field.

In the sea ice covered areas of the polar oceans, the observations were showing very depleted δ18O and δ2H values which could not be reproduced by our atmospheric model. We managed to identify the cause of this discrepancy, by adding a new source of humidity in the model. The sublimation of sea ice as a source of humidity was already considered by the model, but the isotopic composition of the sea ice was assumed equal to the oceanic water from which it was formed. However, snow falls can also be deposited on already formed sea ice and have a much more depleted isotopic composition than the oceanic water. Once integrated in the model as a new potential source of sublimation, this deposited snow drastically changed the simulations of vapour isotopes in polar regions and the simulations are now in much better agreement with the observations in sea ice covered areas (Figure 3). 

Figure 3. Improvement of the simulated vapour isotopic signal by the isotope-enable atmospheric general circulation model

Figure 3. Improvement of the simulated vapour isotopic signal (δ18O and d-excess) by the isotope-enable atmospheric general circulation model ECHAM5-wiso while considering the deposited snow on top of sea ice as a sublimation source and no wind speed effect on the fractionation during oceanic evaporation (ECHAMfinal, dark blue) compared to bare sea ice created from oceanic water only and wind-speed dependent fractionation (ECHAMexp, orange).

Our observations have shown their ability to benchmark atmospheric models of atmospheric water cycle. They highlight different processes having significant consequences on the simulation of water isotopic composition in vapour and precipitation at the global scale, which should be considered in all atmospheric water cycle modelling experiments. They contribute to better understand the creation of the first water isotopic signal during oceanic evaporation. This is particularly important as the oceanic evaporation will later determine the isotopic signal found in subsequent precipitation and therefore in ice-based paleoclimate archives.

The original aim of our project is to combine it with similar observations performed in eastern Siberia (Samoylov station in the Lena river delta, 72°22’N, 126°29’E) also involving a Picarro analyzer L2140-i. Paleoclimatic data based on ice wedges have been retrieved at this location and have depicted a warming during the Holocene, in opposition with other Arctic paleoclimate proxies (Meyer et al., 2015). The different seasonality of the climate evolution has been questioned, as the ice wedges proxies are sensitive to winter temperatures, contrary to most commonly used proxies which are sensitive to summer temperatures. A better understanding of the seasonality of the regional moisture sources and their isotopic signals would significantly contribute to the interpretation of these data.

However, the large spatial coverage of our record makes it also useful for many projects involving water isotopes at any latitude around the whole Atlantic basin. Another project started in 2017 in our group, for which Picarro water vapour observations are continuously conducted in Antarctica at the German station Neumayer-III (70°40´S, 8°16´W). Combining records at sea and on the ice-shelf will provide many information on the regional moisture sources. 


Jean-Louis Bonne

Original article:

Bonne, J.L., Behrens, M., Meyer, H., Kipfstuhl, S., Rabe, B., Schönicke, L., Steen-Larsen, H.C., Werner, M.: Resolving the controls of water vapour isotopes in the Atlantic sector, Nature Communications, doi: 10.1038/s41467-019-09242-6, 2019.



Dansgaard, W., 1964. Stable isotopes in precipitation. Tellus 16, 436–468. https://doi.org/10.3402/tellusa.v16i4.8993

Jouzel, J., 2014. 5.8 - Water Stable Isotopes: Atmospheric Composition and Applications in Polar Ice Core Studies, in: Holland, H.D., Turekian, K.K. (Eds.), Treatise on Geochemistry (Second Edition). Elsevier, Oxford, pp. 213–256. https://doi.org/10.1016/B978-0-08-095975-7.00408-3

Merlivat, L., Jouzel, J., 1979. Global climatic interpretation of the deuterium-oxygen 18 relationship for precipitation. Journal of Geophysical Research: Oceans 84, 5029–5033. https://doi.org/10.1029/JC084iC08p05029

Meyer, H., Opel, T., Laepple, T., Dereviagin, A.Y., Hoffmann, K., Werner, M., 2015. Long-term winter warming trend in the Siberian Arctic during the mid- to late Holocene. Nature Geosci 8, 122–125. https://doi.org/10.1038/ngeo2349

Meyer, H., Schirrmeister, L., Yoshikawa, K., Opel, T., Wetterich, S., Hubberten, H.-W., Brown, J., 2010. Permafrost evidence for severe winter cooling during the Younger Dryas in northern Alaska. Geophysical Research Letters 37. https://doi.org/10.1029/2009GL041013


Source of sublimation

In polar regions, sea ice can be partially covered by a layer of deposited snow, which can later be a source of sublimation. Picture taken in the Fram Strait from the Polarstern ice breaker in July 2015.