Climatic controls on interannual variability of precipitation delta O-18: Simulated influence of temperature, precipitation amount, and vapor source region
Je. Cole et al., Climatic controls on interannual variability of precipitation delta O-18: Simulated influence of temperature, precipitation amount, and vapor source region, J GEO RES-A, 104(D12), 1999, pp. 14223-14235
We use an atmospheric GCM that incorporates stable isotopes and regional va
por source tracers in the hydrologic cycle to explore the relationship betw
een interannual variability in climate and precipitation delta(18)O globall
y. On the basis of a 12-year simulation forced by observed sea surface temp
eratures (SSTs), we identify changes in the amount of precipitation and in
the contributions of local and nearby vapor sources as the most important d
eterminants of simulated interannual isotopic changes. The model simulates
weak positive correlation between temperature and isotopic variability only
in certain continental regions, mostly in the extratropics. Comparison wit
h long observed records of isotopes and climate indicates that the model si
mulates realistic patterns of temperature-isotope correlation but may overe
stimate the isotopic influence of precipitation amount. perturbations in ci
rculation patterns that alter the transport and mixing of air masses at a s
ite also change the relative contributions of vapor from different source r
egions. Simulated changes in vapor source regions are large, reaching +/- 1
0-15% of the total precipitation, and cause significant isotopic variabilit
y in nearly all grid cells. Our results suggest that shifts among vapor sou
rces may provide an important control on the interannual isotopic variabili
ty observed in modern precipitation and paleoclimatic records. The isotopic
variability simulated in this experiment results from the interaction of s
everal aspects of climate. Interannual temperature variability generally in
volves circulation changes that alter air mass transport, vapor source regi
ons, and condensation history; this advective mechanism may explain the rel
ative weakness of temperature-isotope correlations in both the model and th
e observations.