Modeling the water isotopes in Greenland precipitation 1959-2001 with the meso-scale model REMO-iso

TY - JOUR

T1 - Modeling the water isotopes in Greenland precipitation 1959-2001 with the meso-scale model REMO-iso

AU - Sjolte, Jesper

AU - Hofmann, G.

AU - Johnsen, Sigfus Johann

AU - Vinther, Bo Møllesøe

AU - Masson-Delmotte, V.

AU - Sturm, C.

PY - 2011/9/21

Y1 - 2011/9/21

N2 - Ice core studies have proved the δ18O in Greenland precipitation to be correlated to the phase of the North Atlantic Oscillation (NAO). This subject has also been investigated in modeling studies. However, these studies have either had severe biases in the δ18O levels, or have not been designed to be compared directly with observations. In this study we nudge a meso-scale climate model fitted with stable water isotope diagnostics (REMO-iso) to follow the actual weather patterns for the period 1959-2001. We evaluate this simulation using meteorological observations from stations along the Greenland coast, and δ18O from several Greenland ice core stacks and Global Network In Precipitation (GNIP) data from Greenland, Iceland and Svalbard. The REMO-iso output explains up to 40% of the interannual δ18O variability observed in ice cores, which is comparable to the model performance for precipitation. In terms of reproducing the observed variability the global model, ECHAM4-iso performs on the same level as REMO-iso. However, REMO-iso has smaller biases in δ18O and improved representation of the observed spatial δ18O- temperature slope compared to ECHAM4-iso. Analysis of the main modes of winter variability of δ18O shows a coherent signal in Central and Western Greenland similar to results from ice cores. The NAO explains 20% of the leading δ18O pattern. Based on the model output we suggest that methods to reconstruct the NAO from Greenland ice cores employ both δ18O and accumulation records.

AB - Ice core studies have proved the δ18O in Greenland precipitation to be correlated to the phase of the North Atlantic Oscillation (NAO). This subject has also been investigated in modeling studies. However, these studies have either had severe biases in the δ18O levels, or have not been designed to be compared directly with observations. In this study we nudge a meso-scale climate model fitted with stable water isotope diagnostics (REMO-iso) to follow the actual weather patterns for the period 1959-2001. We evaluate this simulation using meteorological observations from stations along the Greenland coast, and δ18O from several Greenland ice core stacks and Global Network In Precipitation (GNIP) data from Greenland, Iceland and Svalbard. The REMO-iso output explains up to 40% of the interannual δ18O variability observed in ice cores, which is comparable to the model performance for precipitation. In terms of reproducing the observed variability the global model, ECHAM4-iso performs on the same level as REMO-iso. However, REMO-iso has smaller biases in δ18O and improved representation of the observed spatial δ18O- temperature slope compared to ECHAM4-iso. Analysis of the main modes of winter variability of δ18O shows a coherent signal in Central and Western Greenland similar to results from ice cores. The NAO explains 20% of the leading δ18O pattern. Based on the model output we suggest that methods to reconstruct the NAO from Greenland ice cores employ both δ18O and accumulation records.

U2 - 10.1029/2010JD015287

DO - 10.1029/2010JD015287

M3 - Journal article

SN - 0148-0227

VL - 116

SP - D18105

JO - Journal of Geophysical Research

JF - Journal of Geophysical Research

ER -