Abstract. The surface mass balance (SMB) of the Greenland Ice Sheet is subject to considerable uncertainties that complicate predictions of sea level rise caused by climate change. We examine the SMB of the Greenland Ice Sheet in the 21st century with the surface energy and mass balance model BESSI. To estimate the uncertainty of the SMB, we conduct simulations for four greenhouse gas emission scenarios using the output of a wide range of climate models from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) to force BESSI. In addition, the uncertainty of the SMB simulation is estimated by using 16 different parameter sets in our SMB model. The median SMB across climate models and parameter sets, integrated over the ice sheet, decreases over time for every emission scenario. As expected, the decrease in SMB is stronger for higher greenhouse gas emissions. The regional distribution of the resulting SMB shows the most substantial SMB decrease in western Greenland for all climate models, whereas the differences between the climate models are most pronounced in the north and in the area around the equilibrium line. Temperature and precipitation are the input variables of the snow model that have the largest influence on the SMB and the largest differences between climate models. In our ensemble, the range of uncertainty in the SMB is greater than in other studies that used fewer climate models as forcing. An analysis of the different sources of uncertainty shows that the uncertainty caused by the different climate models for a given scenario is larger than the uncertainty caused by the climate scenarios. In comparison, the uncertainty caused by the snow model parameters is negligible, leaving the uncertainty of the climate models as the main reason for SMB uncertainty.

Abstract. The surface mass balance (SMB) of the Greenland ice sheet is subject to considerable uncertainties that complicate predictions of sea level rise caused by climate change. We examine the SMB of the Greenland ice sheet in the 21st century with the Bergen Snow Simulator (BESSI) surface energy and mass balance model. To estimate the uncertainty of the SMB, we conduct simulations for four greenhouse gas emission scenarios using the output of a wide range of Earth system models (ESMs) from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) to force BESSI. In addition, the uncertainty of the SMB simulation is estimated by using 16 different parameter sets in our SMB model. The median SMB across ESMs and parameter sets, integrated over the ice sheet, decreases over time for every emission scenario. As expected, the decrease in SMB is stronger for higher greenhouse gas emissions. The regional distribution of the resulting SMB shows the most substantial SMB decrease in western Greenland for all ESMs, whereas the differences between the ESMs are most pronounced in the north and around the equilibrium line. Temperature and precipitation are the input variables of the snow model that have the largest influence on the SMB and the largest differences between ESMs. In our ensemble, the range of uncertainty in the SMB is greater than in previous studies that used fewer ESMs as forcing. An analysis of the different sources of uncertainty shows that the uncertainty caused by the different ESMs for a given scenario is larger than the uncertainty caused by the climate scenarios. In comparison, the uncertainty caused by the snow model parameters is negligible, leaving the uncertainty of the ESMs as the main reason for SMB uncertainty.

<p>Wir untersuchen den Einfluss der subtropischen Zirkulation auf tropische Wellen in einem vereinfachten Rahmen und präsentieren Simulationen mit einem nichtlinearen numerischen Modell des tropischen Kanals. Das Modell löst die prognostischen Gleichungen für potentielle Temperatur und Wind für eine einzelne vertikale Mode auf der äquatorialen beta-Ebene. Unser Schwerpunkt liegt auf der Madden-Julian Oszillation (MJO), die ein wichtiger Teil der intrasaisonalen Variabilität ist. Die Anfangsbedingungen und lateralen Randbedingungen sind repräsentativ für die acht Phasen der MJO. Sie werden durch Projektion von Reanalyse-Daten auf Eigenfunktionen der linearisierten Grundgleichungen basierend auf der linearen Wellentheorie und anschließender Bestimmung der Korrelation mit einem MJO-Index erstellt. So wird ermöglicht, dass die Anfangs- und lateralen Randbedingungen der Simulationen alle oder nur eine Auswahl der äquatorialen Wellen enthalten.</p> <p>In unseren Simulationen stellen die lateralen Randbedingungen nacheinander die acht Phasen der MJO dar. Auf kurzen Zeitskalen wird die Zirkulation von den Wellen dominiert, die in den Anfangsbedingungen enthaltenen sind. Auf längeren Zeitskalen wird der Einfluss der Anfangsbedingungen vernachlässigbar klein. Dann werden zusätzlich zu Rossby-Wellen, die sich nach Westen verlagern, auch nach Osten laufende Wellen mit kleiner Amplitude detektiert, welche die Phasengeschwindigkeit der Kelvin-Welle haben. Die Vergleichbarkeit der angeregten ostwärts laufenden Wellen mit der sich ebenfalls nach Osten bewegenden MJO ist zwar noch zu untersuchen, aber wir zeigen, dass solche Wellen durch zeitabhängige laterale Randbedingungen angeregt werden können.</p>

<p>The surface mass balance (SMB) of the Greenland Ice Sheet is subject to considerable uncertainties that complicate predictions of sea-level rise caused by climate change.<br>We examine the SMB of the Greenland Ice Sheet and its uncertainty in the 21st century using a wide ensemble of simulations with the surface energy and mass balance model "BEr<em>ge</em>n Snow SImulator" (BESSI). We conduct simulations for four greenhouse gas emission scenarios using the output of 26 climate models from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) to force BESSI. In addition, the uncertainty of the SMB simulation is estimated by using 16 different parameter sets in our SMB model. The median SMB across climate models, integrated over the ice sheet, decreases for every emission scenario and every parameter set. As expected, the decrease in SMB is stronger for higher greenhouse gas emissions. The uncertainty range in SMB is considerably greater in our ensemble than in other studies that used fewer climate models as forcing. An analysis of the different sources of uncertainty shows that the differences between climate models are the main reason for SMB uncertainty, exceeding even the uncertainty due to the choice of climate scenario. In comparison, the uncertainty caused by the snow model parameters is negligible. The differences between the climate models are most pronounced in the north of Greenland and in the area around the equilibrium line, whereas the ensemble of simulations agrees that the SMB decrease is greatest in the west of Greenland. </p>