Evaluating hydrological processes in the Community Atmosphere Model Version 5 (CAM5) using stable isotope ratios of water

Abstract

Water isotope‐enabled climate and earth system models are able to directly simulate paleoclimate proxy records to aid in climate reconstruction. A less used major advantage is that water isotopologues provide an independent constraint on many atmospheric and hydrologic processes, allowing the model to be developed and tuned in a more physically accurate way. This paper describes the new isotope‐enabled CAM5 model, including its isotopic physics routines, and its ability to simulate the modern distribution of water isotopologues in vapor and precipitation. It is found that the model has a negative isotopic bias in precipitation. This bias is partially attributed to model overestimates of deep convection, particularly over the midlatitude oceans during winter. This was determined by examining isotope ratios both in precipitation and vapor, instead of precipitation alone. This enhanced convective activity depletes the isotopic water vapor in the lower troposphere, where the majority of poleward moisture transport occurs, resulting in the insufficient transport of water isotopologue mass poleward and landward. This analysis also demonstrates that large‐scale dynamical or moisture source changes can impact isotopologue values as much as local shifts in temperature or precipitation amount. The diagnosis of limitations in the large‐scale transport characteristics has major implications if one is using isotope‐enabled climate models to examine paleoclimate proxy records, as well as the modern global hydroclimate.

The Connected Isotopic Water Cycle in the Community Earth System Model Version 1

“…The simulated physical state of the atmosphere in iCESM1 compares reasonably well to ERA-Interim (Figures 2 and 3), although there are noticeable biases, such as the presence of a double ITCZ. The model is also generally too cold and too humid relative to ERA-Interim, consistent with iCAM results from Nusbaumer et al (2017). Figure 4 shows the average δ 18 O of precipitation ( Figure 4a) and d-excess of precipitation (Figure 4b), compared with the GNIP.…”

“…Figures 4c and 4d show a scatterplot of the iCESM1 precipitation δ 18 O (Figure 4c) and d-excess (Figure 4d) values for the grid points closest to the GNIP stations, with the black solid line representing a perfect match. iCESM1 captures the general qualitative and quantitative features of isotopes in precipitation, but does exhibit a depleted bias in precipitation δ 18 O (median bias = −2.5‰), which is comparable to that in the uncoupled configuration (median bias = −2.2‰ from Nusbaumer et al (2017)). The deuterium excess, d-excess (defined d-excess = δD − 8 × δ 18 O), values are not simulated well, with less correlation with the observations and too large a mean (median bias = 3.3‰).…”

“…This change was required to allocate detrained condensate to ice and liquid stratiform cloud. A more in-depth description of the isotopic physics, along with an analysis of iCAM5.3 results compared to observations, can be found in Nusbaumer et al (2017).…”

“…These isotopic water fluxes and state variables are remapped and merged for the exchange on the appropriate grid, and sent to each model for use in driving the isotopic hydrologic cycle. Details about the computation of isotopic evaporation over the sea surface can be found in Nusbaumer et al (2017).…”

“…Earlier isotope-enabled simulations typically included only the atmosphere and land surface (Field et al, 2014;Hoffmann et al, 1998;Joussaume et al, 1984;Jouzel et al, 1987;Jouzel et al, 1991;Mathieu et al, 2002;Noone & Simmonds, 2002;Noone & Sturm, 2010;Werner et al, 2011), and have been found to be reliable in reproducing the extensive database of precipitation observations from the Global Network for Isotopes in Precipitation (GNIP) (IAEA/ WMO, 2016). There are fewer more recent atmospheric models with isotopic tracers that have been compared to isotope ratios of tropospheric water vapor from satellite or in situ observations (notably, Schmidt et al, 2005Nusbaumer et al, 2017). Water isotope ratios have also been included in ocean general circulation models (Delaygue et al, 2000;Paul et al, 1999;Schmidt, 1998Schmidt, , 1999 and used for constructing past and present three-dimensional tracer fields (Wadley et al, 2002).…”