The 1979–2005 Greenland Ice Sheet Melt Extent From Passive Microwave Data - AGU

[hatrack]

The 1979–2005 Greenland ice sheet melt extent from passive microwave data using an improved version of the melt retrieval XPGR algorithm

X. Fettweis and J.-P. van Ypersele
Institut d'Astronomie et de Géophysique Georges Lemaître, Université Catholique de Louvain,
Louvain-la-Neuve, Belgium

H. Gallée
UMR 5183, Laboratoire de Glaciologie et Géophysique de l'Environnement, CNRS, Université Joseph Fourier,
Saint-Martin dH'ères, France

F. Lefebre
Integrated Environmental Studies, Flemish Institute for Technological Research,
Mol, Belgium

W. Lefebvre
Institut d'Astronomie et de Géophysique Georges Lemaître, Université Catholique de Louvain,
Louvain-la-Neuve, Belgium

Abstract

<1> Analysis of passive microwave satellite observations over the Greenland ice sheet reveals a significant increase in surface melt over the period 1979–2005. Since 1979, the total melt area was found to have increased by +1.22 × 107 km2. An improved version of the cross-polarized gradient ratio (XPGR) technique is used to identify the melt from the brightness temperatures. The improvements in the melt retrieval XPGR algorithm as well as the surface melt acceleration are discussed with results from a coupled atmosphere-snow regional climate model. From 1979 to 2005, the ablation period has been increasing everywhere over the melt zone except in the regions where the model simulates an increased summer snowfall. Indeed, more snowfall in summer decreases the liquid water content of the snowpack, raises the albedo and therefore reduces the melt. Finally, the observed melt acceleration over the Greenland ice sheet is highly correlated with both Greenland and global warming suggesting a continuing surface melt increase in the future.

EDIT


Figure 2. (a) Annually cumulated melt area detected by XPGR (dotted), by ImpXPGR (solid) and simulated by MAR (dashed). (b) Maximum daily melt extent of the ice sheet. The cumulated melt area is defined as the annual total sum of every daily ice sheet melt area and the maximum melt extent is the maximum extent experiencing at least one melt day. Enhanced EPS <317 KB>

<7> Over the periode 1979–2005, the cumulated melt extent retrieved by XPGR and ImpXPGR respectively shows an increase of +0.019 × 107 km2 year−1 and +0.047 × 107 km2 year−1 (See Figure 2a). From 1979, the total increase in 2005 is respectively +0.48 × 107 km2 (i.e. +28%) and +1.22 × 107 km2 (i.e.+54%) with a significance of 92.0% and 99.5%. Following Fettweis et al. <2006>, the total surface meltwater runoff of the Greenland ice sheet also significantly increased since 1979. The significance has been tested using a Monte-Carlo method with 1,000,000 simulations of autocorrelated data series with the same autocorrelation as the (Imp)XPGR time series. The MAR model simulates the same trend as ImpXPGR. The surface melt acceleration is higher using ImpXPGR than using XPGR because rainfall has increased during this period which enhances the probability of perturbations in the microwave melt signal. The ice sheet area on which MAR simulates daily rainfall greater that 1 mm/day has been extending on average by +0.006 × 107 km2 year−1 (significance of 98.1%) since 1979. The trend to heavier rainfall is highly correlated to the warming observed on Greenland as we will see hereafter.

<8> The maximum of both annually cumulated melt area and melt extent occurs in 2005 followed by 2002 in the ImpXPGR melt fields. In the XPGR fields, the maximum annually cumulated melt area occurs in 1991 and the maximum daily melt extent in 2002 according to Steffen et al. <2004> and K. Steffen, Greenland melt extent, 2005, Cooperative Institute for Research in Environmental Sciences, Boulder, Colorado, available at http://cires.colorado.edu/science/groups/steffen/greenland/melt2005/). The trend of the maximum extent experiencing at least one melt day shows an increase of +0.89 × 105 km2 (i.e.+28%) and +1.64 × 105 km2 (i.e.+42%) respectively in XPGR and ImpXPGR since 1979 until now with a significance of 92.9% and 98.0% (See Figure 2b).


Figure 3. (a) Melt trend (in ablation days yr−1) detected by XPGR for the period 1979–2005. (b) Idem by ImpXPGR. (c) Year in which some melt is recorded by ImpXPGR the first time. Negative trends are hatched and surrounded by a black dotted line. Minimum (min) and maximum (max) values are indicated below the figures as well as the ice sheet average (ave) and the standard deviation (sig). Enhanced EPS <630 KB>

<9> This significant snowmelt increase in the microwave satellite data is confirmed in Figure 3 which reveals an increase of the length of the melt season as well as an expansion of the melt area on the Greenland ice sheet throughout the period 1979–2005. The number of melting days increases mainly in the North of Greenland and along the western coast at mid altitude (about near 1500 m – 2000 m). At lower altitude in the western part, no change occurs because the melt period extends already all of the melt season at the beginning of eighties. The changes are smaller along the eastern coast and the trend is even negative in the ImpXPGR fields along the south-eastern mountainous regions of the ice sheet (see Figure 3b). But the microwave brightness temperature could be biased by numerous rock outcrops (boulders) found in this mountainous region . In the XPGR fields, a decrease of the melt period is recorded along the western margin. Figure 3c shows an expansion of the largest melt area reaching since the eighties new zones every year (e.g., the South-Dome at an elevation of 2900 m in 2005 (Steffen, 2005). The expansion of melt area is the most spectacular since the years 2000 in the Northern part of Greenland where no melt has been recorded before above 1500 m.

EDIT

No link, subscription service.