Controls on isotopic gradients in rain

Concentrations of heavy isotopes of hydrogen and oxygen decrease in rain as storms cross land. A model examines the transport of water vapour that causes this effect, and provides insight into past and present climates.

Water vapour is lifted from the sea by evaporation, transported by storms and then released across land as precipitation. As rain falls from clouds, it preferentially removes water containing heavy isotopes, so that both clouds and rain become progressively depleted in deuterium and oxygen-18 as they move across continents. The extent of this 'rainout' dominates all the observed trends of continental isotope gradients, but the gradients can vary by up to 50-fold. These variations reflect complex factors that have long eluded simple quantification. Writing in Earth and Planetary Science Letters, Winnick et al.¹ address this problem with their report of an improved one-dimensional model that follows rain isotopes along a storm track.

The water-cycle processes that cause isotope 'sorting' in rain are well understood, and the resulting isotope gradients in continental precipitation have been extensively documented². At the mid-latitudes, where air temperatures are unstable, atmospheric mixing is particularly pronounced. This lowers isotope gradients as a result of the diffusion-like smoothing of turbulent mixing, a process called eddy diffusion³. And in tropical lowlands such as the Amazon Basin, small humidity gradients across long distances diminish isotope sensitivity to mixing and transport^4,5.

Moisture fluxes from plants and evaporated surface water also affect isotope gradients in precipitation. Transpiration — the evaporation of plant water — occurs through leaf stomata, which act like the tips of small capillaries. The narrow plumbing in leaves prevents back-mixing of evaporated water, and so the water vapour that exits from plants has the same isotopic composition as water taken up by their roots. The flux of moisture returned from plants therefore has a modest influence on downstream rainfall isotope gradients, but it does recharge cloud water-vapour pools, and this reduces the extent of rainout⁶. By contrast, evaporation from the surfaces of lakes and rivers strongly sorts water isotopes, and delivers water vapour depleted in heavy isotopes back to clouds^7,8. The different amounts and mass effects of surface- and plant-water fluxes in different regions cause continental isotope gradients to be sensitive to the geography of vegetation and aridity.

Winnick et al.¹ use their model of water-vapour transport to investigate the dominant factors governing isotope gradients in precipitation. By scaling calculations to the extent of rainout, which they represent using humidity gradients, the authors find that isotopic-gradient variability reflects two things: differing amounts of water vapour returned to clouds (from plants and by evaporation from surface waters, together known as evapotranspiration); and the amount of advection (water-vapour transport by the bulk motion of air) and turbulent mixing in the atmosphere. They conclude that continental isotope gradients correspond to global patterns of air circulation and hydrology, yet are most sensitive to evapotranspiration, transport and mixing in areas near mountains.

Water-isotope records are helpful tools for reconstructing mountain elevation and climate patterns of the past. The isotope patterns of precipitation-derived water that is chemically captured as carbonates in mountain lakes and soils^9,10, as well as in preserved plant waxes¹¹, offer geologists unique insight into how fast and how high mountains rose, and help to constrain tectonic models of mountain building. Hydrologic isotope records also capture the impacts of climate change on the water cycle¹².

But water drawn from different sources along shifting storm tracks can confound isotopic records of elevation and climate processes at any one location (see ref. 11, for example). Winnick et al. point out that isotope gradients based on data from several locations can help to eliminate source effects and thereby improve reconstructions of past environments. By normalizing data to the spatial scale of humidity gradients, the authors have provided researchers studying palaeoaltimetry and palaeoclimate with a potent quantitative tool for unpacking the interlinked influences of vegetation, climate and topography on continental hydrology.

Mountain topography (Fig. 1) causes sharp gradients in both the flux and isotopic character of precipitation. Winnick and co-workers show that these gradients are highly sensitive to factors other than elevation, sounding a cautionary note for palaeoaltimetry applications. But they also find that mountainous localities may capture details of climate impacts that are probably missed in less-sensitive lowland regions — which is good news for climate scientists. Nevertheless, climates and mountain waters are not simple systems. Changing temperatures can shift not only the phase of precipitation (whether it snows or rains, for example), but also the extent of isotope sorting during condensation and evaporation. For example, hot climates reduce isotope gradients in water simply because isotope fractionation is less pronounced at higher temperatures.

Figure 1: Rain over the Denali National Park, Alaska.

Bernd Zoller/Robert Harding Picture Library

Winnick et al.¹ have modelled water-vapour transport along the track of a storm.

Of necessity, Winnick and colleagues' model requires many pre-set parameters, including temperature-linked isotope-mass sensitivities, humidity and the ratio of evaporation to transpiration. But as we venture back through the pageantry of past ecology and climates, parameters based on modern data become increasingly inadequate, because life and its habitats have evolved considerably over long periods of time. Furthermore, short-term variations in vegetation, droughts and temperature — which pose major challenges for society as our planet warms — also occurred throughout Earth's history. The model's requirements therefore bring into focus the need to support isotope records with palaeoclimate and palaeobotanical data. These will help mountains and their isotope gradients to teach us crucial lessons from the past about water, climate and carbon, and how they are linked across continents and the globe.