A new glacier inventory for 2009 reveals spatial and temporal variability in glacier response to atmospheric warming in the Northern Antarctic Peninsula, 1988–2009

Davies, B. J., Carrivick, J. L., Glasser, N. F., Hambrey, M. J., Smellie, J. S. (2012). A new glacier inventory for 2009 reveals spatial and temporal variability in glacier response to atmospheric warming in the northern Antarctic Peninsula, 1988-2009. The Cryosphere Discuss, 5, 3541-3595.

length or altitude. This paper firstly uses ASTER images from 2009 and a SPIRIT DEM from 2006 to classify the area, length, altitude, slope, aspect, geomorphology, type and hypsometry of 194 glaciers on Trinity Peninsula, Vega Island and James Ross Island. Secondly, this paper uses LANDSAT-4 and ASTER images from 1988 and 2001 and data from the Antarctic Digital Database (ADD) from 1997 to document glacier change 10 198810 -200910 . From 198810 -200110 , 90 % of glaciers receded, and from 200110 -2009, 79 % receded. Glaciers on the western side of Trinity Peninsula retreated relatively little. On the eastern side of Trinity Peninsula, the rate of recession of ice-shelf tributary glaciers has slowed from 12.9 km 2 a −1 (1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)  water glacier recession rates may result from the influence of glacier length, altitude, slope and hypsometry on glacier mass balance. High snowfall means that the glaciers on the Western Peninsula are not currently rapidly receding. Recession rates on the eastern side of Trinity Peninsula are slowing as the floating ice tongues retreat into the fjords and the glaciers reach a new dynamic equilibrium. The rapid glacier recession

Introduction
The Northern Antarctic Peninsula Ice Sheet is particularly dynamic and sensitive to climate change because of its relatively small size and northern location (Vaughan et al., 2003;Smith and Anderson, 2010). The Antarctic Peninsula region is also one of the most rapidly warming places on Earth, with mean air temperature increasing by 5 2.5 • C over the last 50 yr (King, 1994;Turner et al., 2005). With continued atmospheric warming predicted for the Antarctic Peninsula region over the next century (Vaughan et al., 2003), the viability of glaciers in this region is questionable. With this warming trend, the −9 • C annual isotherm (the thermal limit of ice shelves (Morris and Vaughan, 2003)) has moved southwards, resulting in 28 000 km 2 being lost from Antarctic Penin- 10 sula ice shelves since 1960 (Vaughan and Doake, 1996;Cook et al., 2005;Cook and Vaughan, 2010). Ice-shelf tributary glaciers accelerated and thinned following the disintegration of Larsen Ice Shelf (De Angelis and Skvarca, 2003), with up to a six-fold increase in centre-line speeds (Scambos et al., 2004). Other tidewater glaciers on the Antarctic Peninsula are accelerating, thinning and retreating in response to increased 15 atmospheric and sea surface temperatures (Pritchard and Vaughan, 2007). Antarctic Peninsula glaciers currently contribute 0.22 ± 0.16 mm a −1 to sea level rise (Hock et al., 2009), and have a total eustatic sea-level equivalent of 0.24 m (Pritchard and Vaughan, 2007). It is clearly important to establish how the on-going mass balance changes and associated thinning, drawdown and ice front retreat will progress, and Peninsula by Prince Gustav Channel, which is 8 to 24 km wide (Fig. 1). Water depths reach 1200 m and shallow southwards to less than 450 m (Evans et al., 2005). The north-south orientated Antarctic Peninsula mountains are an orographic barrier to persistent Southern Ocean westerlies. Cold continental air in the Weddell Sea also flows northwards, barring the warmer maritime air masses of the Bellingshausen Sea (King 10 et al., 2003). The Western Antarctic Peninsula therefore has a polar maritime climate, dominated by the warm Bellingshausen Sea, whilst the Eastern Antarctic Peninsula and James Ross Island have a polar continental climate, dominated by the Weddell Sea (Martin and Peel, 1978;Vaughan et al., 2003). The western coast of the Antarctic Peninsula is therefore typically 7 • C warmer than the eastern coast. These differences 15 are further exacerbated by the different sea-ice regimes in the Bellingshausen and Weddell seas, with the Weddell Sea being sea-ice bound for much of the year (King et al., 2003). Climate records from the South Orkney Islands suggest that regional warming probably began in the 1930s (Vaughan et al., 2003). The greatest warming rates are in 20 the east (at Marambio and Esperanza stations; cf. Fig. 1), and the smallest on the northwest coast. The most pronounced warming is in the winter months, and is related to changes in atmospheric circulation, sea ice extent and ocean processes (Stastna, 2010). The Antarctic Oscillation represents the periodic strengthening and weakening of the belt of tropospheric westerlies that surround Antarctica (van den Broeke and van 25 Lipzig, 2004). A strengthening of this circumpolar vortex results in an asymmetric surface pattern change, with pressure falling over Marie Byrd Land. This pressure pattern causes northerly flow anomalies, which in turn produce cooling over East Antarctica 3545 The Antarctic Peninsula Ice Sheet is typically 500 m thick over Trinity Peninsula, and its central plateau glaciers attain altitudes of over 1600 m a.s.l. along its central spine (Fig. 1). Trinity Peninsula outlet glaciers flow predominantly east and west, perpendicular to the Peninsula spine (Heroy and Anderson, 2005). On the eastern flank many outlet glaciers terminate in floating or partly floating tongues. The North-Western Trinity Peninsula coastline mainly comprises grounded ice cliffs with the floating ice margins of a few small ice shelves and numerous small glaciers (Ferrigno et al., 2006). Currently, 80 % of James Ross Island is ice-covered (Rabassa et al., 1982), with the remainder comprising rock or Quaternary sediments with perennial permafrost (Fukuda et al., 1992). The island is bounded to the Southwest, South and East by 20 high cliffs and valley glaciers, and the ice cap drains over the cliffs into outlet glaciers, principally via avalanches and ice falls (Skvarca et al., 1995). The bedrock structure of resistant volcanic rocks overlying soft Cretaceous sediments (Smellie et al., 2008;Hambrey et al., 2008) fosters the development of elongate, over-deepened cirques, with steep or near vertical back walls up to 800 m high. Vega Island is dominated by 25 two plateau ice caps that feed numerous small tidewater glaciers. The ice-free area (approximately 34 %) is dominated by permafrost processes. Snow Hill Island, Corry Island and Eagle Island ( Fig. 1)  Snow Hill Island has extensive floating margins. Retreating tidewater termini have been mapped for the entire Trinity Peninsula, with outlet glaciers on James Ross Island showing large reductions in area since the 1940s (Ferrigno et al., 2006). PGIS was connected to Larsen Ice Shelf until 1957/58 (Cook andVaughan, 2010). The northernmost ice shelf on the Eastern Antarctic Peninsula, it 5 was the first to show signs of retreat (Ferrigno et al., 2006), and rapidly disintegrated in 1995 (Skvarca et al., 1995;Cooper, 1997), followed by the thinning, acceleration and rapid recession of the former ice shelf feeding glaciers (Rau et al., 2004;Glasser et al., 2011). During the later twentieth century, most glaciers on James Ross Island and North-Eastern Antarctic Peninsula experienced rapid recession and calving, with the strongest evidence of recession in the northernmost parts. This is generally attributed to prevailing climatic warming (Vaughan et al., 2003;Rau et al., 2004). An average retreat rate of the James Ross Island glaciers of 1.8 km 2 a −1 from 1975-1988 (Skvarca et al., 1995) doubled after the disintegration of PGIS, to 3.8 km 2 a −1 between 1988 and 2001 (Rau et al., 2004). Although Skvarca et al. (1995) noticed no dramatic change for 15 small land-based glaciers from 1977 to 1988, retreat was observed from 1988 to 2001 (Skvarca et al., 1995;Skvarca and De Angelis, 2003). 20 This inventory is a snapshot of the glaciers on Trinity Peninsula, James Ross Island, Snow Hill Island and Vega Island in 2009. This data has been made available for download from the GLIMS geodatabase. Our methods conform to those set out by the GLIMS programme (Racoviteanu et al., 2009), and are given in full in the Supplement. Mapping was conducted in ArcGIS 9.3 from interpretation of 2009 from the 2006 SPIRIT Digital Elevation Model (DEM) ( Table 1). Ground-control points were not necessary since accurate geolocation of the SPOT-5 images is possible. Each pixel in a SPOT-5 image can be located on the ground to ±25 m at the 66 % confidence interval, although errors may be larger in areas of flat glacier ice (Reinartz et al., 2006;Berthier et al., 2007;Korona et al., 2009). Glacier names and numbers were taken from previous glacier inventories and published maps (Rabassa et al., 1982;British Antarctic Survey, 1995, 2010Czech Geological Survey, 2009).

Error estimation
Sources of uncertainty and mitigation strategies are summarised in Table 2. The largest errors in glacier drainage basin delineation are derived from interpreter error 15 of ice-divide identification and mapping. DEM quality in areas of flat white ice also limits the accuracy of ice divide mapping. ASTER Level 1B images are pre-processed but were co-registered to the SPOT-5 images to reduce errors. Calculation of area is limited by the pixel resolution of the dataset (i.e. ±15 m for ASTER images). Mapping was estimated to be accurate to within 3 pixels (i.e. 45 m). A buffer of 22.5 m width was 20 therefore placed either side of each glacier polygon, providing a minimum and maximum estimate of size and a total error margin equivalent to 3 pixels. There may also be an error in ASTER co-registration (Granshaw and Fountain, 2006); but this is assumed to be within the mapping limits.

Drainage basin delineation
Automated methods for large-scale inventories can be used to map large regions quickly (e.g., Svoboda and Paul, 2009;Bolch et al., 2010), but they do not allow such high detail in digitisation. For this reason, we used manual digitisation to establish glacier polygons in our high-resolution study of a small region, following guidelines set 5 out by the GLIMS programme (refer to Supplement). A glacier polygon is a coherent body of ice that includes all tributaries and connected feeders, and excludes exposed ground and nunataks (Racoviteanu et al., 2009;Raup and Khalsa, 2010). On James Ross Island many glaciers have well defined cirques, but the plateau above them was included in their polygons (Fig. 1); this is because ice situated above the cliff-backed 10 bergschrund contributes snow and ice through frequent avalanches and creep flow to the glacier.  Rau et al., 2005;Raup et al., 2007a,b;Racoviteanu et al., 2009;Paul et al., 2010;Raup and Khalsa, 2010). Form, Primary Classification, Frontal Characteristics and Remarks are geomorphological descriptors that provide detailed information on the glaciers shape, terminus and classification (e.g., cirque, niche, outlet, valley glaciers) that correspond to GLIMS guidelines (Rau et al., 2005;Paul et al., 2010). Introduction

Tables Figures
Back Close

Full Screen / Esc
Printer-friendly Version

Interactive Discussion
Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | but as their recession is controlled more directly by changes in atmospheric temperature and precipitation than marine-terminating glaciers, they can be a sensitive indicator of climate change (Oerlemans, 2005;Carrivick et al., 2011). The nature of the marine-terminating glacier tongue has important implications for oceanographic and glaciological applications, but the grounding zone can only be accurately located 5 with detailed geophysical surveys (Brunt et al., 2010), which are impractical for glacial inventories. However, the floating tongues of marine-terminating glaciers typically have several visual characteristics in common: a pronounced break in slope at the grounding-line, an irregular, heavily crevassed and convex calving front, a flat long profile, an irregular margin, and occasional large rifts. Alternatively, grounded tidewater glaciers are typically characterised by a concave calving front, a steadily dipping long profile, and have no clear break in slope (Dahl and Nesje, 1992;Scambos et al., 2004;Reinartz et al., 2006;Lambrecht et al., 2007;Fricker et al., 2009). Marine-terminating glaciers that exhibit a combination of these characteristics are defined as partially floating, where either it is difficult to determine whether a glacier is floating or not, or where 15 the glacier may be grounded at the lateral margins but floating in the centre of the calving margin. Data were automatically derived in the GIS for minimum (H MIN ), maximum (H MAX ), mean (H MEAN ) and median (H MEDIAN ) elevation using the 2006 SPIRIT DEM. Elevation data are unavailable for GAP40, GAP43 and glaciers north of GAP34 because of the 20 geographical limits of the SPIRIT DEM. Values for mean slope and aspect were also calculated for each glacier polygon using the 2006 SPIRIT DEM (refer to Supplement). Aspect and slope are useful data for modelling purposes (Paul et al., 2010), and mean slope can be used as a proxy for ice thickness. 25 Glacier hypsometry is the distribution of glacier surface area with altitude, and is a major control on mass balance, as it strongly determines the accumulation area ratio (AAR) of a glacier. It is dependent on valley shape, topographic relief and glacier 3550 Introduction depth. When the ELA of a glacier changes, for example due to climate change, the extent of the impact on the glacier relies on the glacier hypsometry (Furbish and Andrews, 1984). Faster rates of accumulation area loss during climatically-driven ELA rise will result in increased glacier recession. Glacier hypsometric curves were calculated by masking 54 glacier polygons (i.e. those over 40 km 2 ; all tidewater outlet glaciers) with the SPIRIT DEM, and the area in each 100 m elevation bin was calculated. Hypsometric curves of cumulative area (km 2 ) could then be plotted for the largest glaciers.

Glacier hypsometry
A single Hypsometric Index (HI) (equivalent to "altitude skew") was calculated for each glacier polygon using the equation below, originally presented by Jiskoot et al. (2009)

Equilibrium line altitude derivation
In a glacier inventory, an estimation of the mean long-term ELA is considered to be vital, because it divides a glacier into ablation and accumulation areas. If air temperature rises and/or if there is declining effective accumulation, mass balance will decline and 15 the ELA will rise. ELAs have considerable inter-annual variability, with positive and negative mass balances from one year to the next (e.g., Carrivick and Chase, 2011). The long-term ELA is best determined by a programme of mass-balance measurements over several years (Braithwaite and Raper, 2009). However, field programmes are time consuming and expensive and the Northern Antarctic Peninsula region is very difficult 20 to access. Indeed, only one mass-balance investigation has been published for the study area . However, there are a wide range of remote methods for calculating glacier ELA (e.g., Carrivick and Brewer, 2004). In this study, five different long-term ELA derivation methods were applied (summarised in the Supplement and in

Calculation of glacier change 1988 -1997 -2001 -2009
Assuming no migration of ice divides, and using the datasets named in Table 1 and in the Supplement, the extent and length of each glacier was also mapped for 1988 and 2001. The ice front positions available in the ADD (from Cook et al., 2005) were used to map ice extents in 1997. However, these data were only available for a limited 5 number of tidewater glaciers. These time slices effectively capture the periods before, during and after ice-shelf collapse. The difference in area of each glacier polygon for each time slice allowed calculation of area lost and rates of recession. Annual rates of glacier recession were calculated for each period. In subsequent analyses, glacier surface areas that changed less than the calculated mapping error were classified as 10 "stationary". "Shrinkage" indicates a loss of surface area.

Glacier size and elevation
In 2009, the Northern Antarctic Peninsula region had 194 individual glaciers that covered a total area of 8140 ± 262 km 2 ( Fig. 1; Table 5; raw data available for download 15 from GLIMS; www.glims.org). Trinity Peninsula was 95 % glacierised with 62 glaciers covering 5827 ± 154 km 2 (Table 5), which included the largest glaciers on the Northern Antarctic Peninsula with 20 glaciers > 100 km 2 . James Ross Island was 75 % glacierised, which is a reduction of 5 % since the survey in 1977 (Rabassa et al., 1982). Our inventory identified 104 glaciers on James Ross Island, whereas the original 1977 20 survey identified 138. This difference is because of the significantly different approach used to map the glaciers (using ice divides and flow unit boundaries to delimit glacier polygons). Many smaller glaciers or possibly snow patches mapped by Rabassa et al. (1982) are no longer discernible. While we attempted to follow the nomenclature used by Rabassa et al. (1982)  had one large ice cap covering 326 ± 3.6 km 2 , and was 96 % glacierised. The mean elevation of glaciers was skewed towards glaciers in the 200-300 m a.s.l. bin (Fig. 3a). A small number of glaciers over 100 km 2 account for most of the glacierised area (Fig. 3b). As the total area of glaciers in the size class 0.1-0.5 km 2 is only 3 km 2 , the underestimate of glacierised area caused by having a minimum glacier 10 size of 0.1 km 2 is likely to be insignificant and certainly considerably less than 3 km 2 . H MEAN of glaciers on Trinity Peninsula peaked at 400 m a.s.l., reflecting its high mountain chain. James Ross Island had the highest number of glaciers with mean elevations above 200 m. There was a strong correlation of glacier length and maximum elevation (r 2 = 0.8; Fig. 3c), and a rather weaker correlation between log glacier area and ele-15 vation (r 2 = 0.5; Fig. 3d; Table 3). There is a weak relationship with log glacier area, H MEAN and slope (Fig. 3e,f) and between glacier length and slope (Table 3). In general, longer glaciers had a lower slope, with large low-angled ablation areas.

Glacier form and classification
Outlet or valley glaciers drain the large plateau ice caps that rest on the Trinity Penin-20 sula mountain chain (Table 6). Almost all the glaciers on Trinity Peninsula calve into the ocean. Of these calving glaciers, 10 had floating and 20 had partially floating termini. James Ross Island had 16 glaciers with floating termini (Fig. 4a), 48 land-terminating glaciers and one lacustrine glacier (GIJR86;

Equilibrium line altitudes
There is a very strong regressionc between ELA MEDIAN , ELA AAR and ELA THAR ( Fig. 3h; Tables 3, 4 and 5). However, there is considerable scatter in ELA HESS and ELA TSL and they show no correlation with ELA MEDIAN (Table 5; p values of 0.4 and 0.3, respectively). The large standard error indicates little regional consistency in ELA TSL .

5
Mapping snowlines on Trinity Peninsula is likely to produce substantially different results to other methods, as the strong east-west precipitation gradient on Trinity Peninsula results in snow lines close to sea level on the western coast, and snow lines at 300 to 400 m a.s.l. on the Eastern Peninsula. Strong winds also affect the distribution of snow at the end of the ablation season. In addition, it was difficult to achieve 10 total scene coverage for each year mapped, and there are large gaps in the data. These polar glaciers are characterised by complex accumulation zones with patches of accumulation and ablation separated by areas of superimposed ice , making accurate remote mapping of transient snow lines difficult. Furthermore, mapping of snowlines may occur either just after or just before a snowfall event. As 15 snowfalls occur throughout the summer season in polar regions, this is likely to further skew results. Because of these inherent methodological difficulties, large data gaps, large inter-annual scatter and low correlation, ELA TSL was henceforth excluded from the analysis. ELA HESS was also excluded from further analysis as it is likely to only be applicable to 13 land-terminating glaciers (not being generally applicable to marine-20 terminating glaciers, glaciers with ice falls, glaciers with complex or compound cirques, or glaciers terminating in steep cliffs; Jiskoot et al., 2009), and there was again large scatter and low correlation (Fig. 3h). ELA MEAN is therefore the average of ELA MEDIAN , ELA AAR and ELA THAR . However, the calculated ELA MEAN for each glacier must necessarily be treated with caution, as 25 it is entirely derived from the altitudinal range and hypsometric curves of the glaciers, has not been checked against mass-balance data, and does not take into account east-west orographic precipitation and wind variations. Standard deviations range from < 5 m to > 500 m, illustrating the difficulty in obtaining a meaningful parameter from topographical data alone. Our topographical ELA predictions can be compared with previous mass balance studies carried out on GIV09, Vega Island (Glaciar Bahia del Diablo; Skvarca et al., 2004). GIV09 was, at time of publication, experiencing a negative mass balance, thin-5 ning and frontal retreat. Skvarca et al. (2004) found the equilibrium line of GIV09 difficult to determine, as it had an accumulation area with zones of snow accumulation and ablation, separated by zones of superimposed ice. The altitude of zero mass balance was estimated by Skvarca et al. (2004) to be 400 m a.s.l. This is much higher than our remotely mapped ELA MEAN of 245 ± 92 m a.s.l. It is thus possible that our 10 predicted ELA MEAN is an underestimate, but it is impossible to verify without further spatially-distributed mass-balance data. Figure 4b shows the distribution of ELA over the study region. The subdued topography on Ulu Peninsula, James Ross Island, results in low ELA MEAN of around 100 m a.s.l. However, the glaciers that have accumulation areas on the Mount Haddington Ice Cap 15 can have ELA MEAN > 800 m a.s.l. ELA MEAN on Trinity Peninsula are generally around 500 to 600 m a.s.l., with the exception of GAP60 and 61, which have ELA MEAN of 1194 ± 727 and 1265 ± 787 m a.s.l., respectively. The range of values may be less for land-terminating simple-basin glaciers, whose size may be more closely determined by climatic factors. For example, GIJR103, a grounded valley glacier with a simple basin, 20 has an ELA MEAN of 87 ± 1 m a.s.l. The range of errors for simple basin glaciers with floating tongues (such as, for example, GIJR64, ELA MEAN 909 ± 432 m a.s.l.) may be because of their large low-lying ablation areas.

Glacier aspect
Mean glacier aspect is summarised in Fig. 4c,e. There is no correlation with eleva-25 tion ( Fig. 3i; Table 5). The aspect of a glacier controls receipt of solar insolation, as well as influencing the drifting of snow through persistent winds. Most glaciers display asymmetry over their surface area, which results, for example, in a strong northeast to 3555 Introduction west-facing preferred glacier aspect on Trinity Peninsula (Fig. 4c). On James Ross Island, there is a preference for northwest-facing aspects. Aspects on Vega Island trend largely southeast and northwest, reflecting the west-east axis of the island.

Glacier hypsometry
The hypsometric index (HI) of all glaciers was calculated and glaciers were divided into 5 five categories from HI > 1.5 (very bottom heavy) to HI < −1.5 (very top heavy) (Table 7; see Jiskoot et al., 2009). There was considerable inter-catchment variability in glacier shape and elevation distributions. Over half (35) of the large outlet glaciers on Trinity Peninsula were very bottom-heavy, with large low-lying areas below the median altitude ( Fig. 4d,f). Exceptions are a few glaciers that terminate within narrow bays 10 (e.g., GAP13, GAP17). However, on the west coast of Trinity Peninsula, glaciers are more equi-dimensional or top-heavy. Glaciers that were formerly tributaries to PGIS were large with relatively small accumulation areas. The remaining glaciers on Eastern Trinity Peninsula were generally bottom-heavy, for example, GAP34, GAP31 and GAP20.

15
On James Ross Island, the low-lying topography of Ulu Peninsula resulted in bottomheavy outlet glaciers draining Dobson Dome (Fig. 4d,f), but the large Mount Haddington Ice Cap contained largely top-heavy glaciers, such as GIJR115 and GIJR61. The longprofiles of tidewater outlet glaciers draining the Mount Haddington Ice Cap are typified by sharp changes in slope angle in ice falls at their cirque headwalls, followed by large, 20 low-lying and relatively flat partly floating or floating glacier tongues (Fig. 4g).
In Fig. 5, example normalised hypsometric curves derived from the 2006 SPIRIT DEM are presented and the normalised ELA MEAN has been plotted, which generally fell at about 60 % of the accumulation area. ELA MEAN , which is closely related to median elevation (which was used in the ELA calculation; see Sect. 3.1.5), is a control to climate change if the ELA rises and allows a large additional area to be an ablation zone, especially if the ELA rises above the new median altitude. Comparing the hypsometric curves in Fig. 5 with the long profiles presented in Fig. 4g, it is obvious that the low-lying floating tongues of the Mount Haddington Ice Cap outlet glaciers result in large, low-lying ablation areas that may be vulnerable climate change (rises in 5 temperature or decreases in effective precipitation).

Change in glacierised area
On Fig. 1

Land-terminating glaciers
The recession of land-terminating glaciers is of particular interest because their activity is directly related to atmospheric and climatic changes, and these glaciers are scarce 15 on the Antarctic Peninsula (Skvarca and De Angelis, 2003). In the study region, they are generally small; most are less than 1 km 2 , and so would be expected to react fastest to external forcings (Paul et al., 2004), although this can vary regionally (Raper and Braithewaite, 2009). Response times are also determined by slope and mass balance gradient (Oerlemans, 2005); Fig. 3f illustrates that there is a weak negative 20 relationship between glacier area and mean slope, and that smaller glaciers are often steeper (p value of 0.001).
The seven land-terminating glaciers on Vega Island have all retreated since 1988 (Fig. 6a,b). However, there has been little shrinkage in these glaciers since 2001. Most glacier retreat on Vega Island occurred in the period 1988-2001, with the glaciers 25 generally shrinking by 2 to 7 % (Fig. 6a). The majority of the 48 land-terminating glaciers on James Ross Island are < 1 km 2 , but the shrinkage of these glaciers is highly variable. For glaciers < 1 km 2 , the percentage shrinkage varies from 0 to 66 %  (Fig. 6c,d). Glaciers < 3 km 2 also show highly variable patterns of shrinkage, from 0 to 35 % (Fig. 7e,f). Overall, the largest reaction to climate warming in the Antarctic Peninsula is in small land-terminating mountain glaciers that are less than 1 km 2 . Indeed, there is a weak regression between initial glacier size (1988) and total glacier area lost ; r 2 = 0.19; Fig. 6g; Table 5).

5
From previous inventories, it is evident that most of the margins land-terminating glaciers in the Northern Antarctic Peninsula region were stationary until 2001-2002, whereupon shrinkage was widely observed (Rau et al., 2004). Our study shows that glacier recession has continued since this period. However, the largest areal changes were observed in the period 1988-2001. Figure 6h shows that the largest annual rates 10 of recession on James Ross Island were between 1988-2001, with only limited recession of land-terminating glaciers after this period. This supports initial observations by Rau et al. (2004). In the period 1975-1988, five land-terminating glaciers were found to be retreating; in the period 1988-2002, 17 out of 21 land-terminating glaciers were retreating. Of the 48 land-terminating glaciers we mapped on the island, 36 out of the 48 mapped land-terminating glaciers had retreated from 1988-2001. In the period 2001-2009, 15 glaciers had retreated. In both time periods, the remaining glaciers were stationary and none had advanced.

Marine-terminating glaciers
In total, glaciers receded by 1319.5±395.6 km 2 in the Northern Antarctic Peninsula be-  The highest annual rate of recession was reached in the period 1988-1997, when Sjögren Glacier (GAP12; Fig. 1) receded at 17.7 km 2 a −1 . Sjögren Glacier was a major contributor to the ice shelf, and lost 1. Analysing the rates of recession for the glaciers that fed PGIS in more detail, Fig. 7c highlights the difference in GAP15, GAP14 and GAP12. It is apparent that most of the shrinkage of these three glaciers occurred before 1997, immediately preceding iceshelf disintegration, with slower and steadier retreat thereafter. These areal losses are 25 substantially larger than the mapped errors. The recession rate decreases after this, with less areal loss. On the other hand, for glaciers on James Ross Island, in particular GIJR92 and GIJR90, which are situated in Röhss Bay, the recession rate is fastest after 2001 (Fig. 7c).  (Table 8), but after 2001 none have advanced more than the error margin. All advances were small and were subsumed by the overall recession from 1988 to 2009. There are fewer data available for Western Antarctic Peninsula, but shrinkage rates for the period 1988-2009 have remained low in that region (Fig. 9). This contrasts with glaciers on North-Eastern Trinity Peninsula (Fig. 8a).

TCD
Some ice-shelf tributary glaciers on James Ross Island also show small signs of advance during the period (Table 8). The remaining glaciers of Röhss Bay experienced enhanced drawdown and faster rates of annual recession after 2001. GIJR90 declined in area at a rate of 1.9 km 2 a −1 from 2001-2009, when the ice shelf retreated beyond the narrow pinning point at the head of Röhss Bay, but the margin was stationary from 15 1988-2001. This is highlighted in Fig. 8d, which shows the difference in recession between 1988-2001 and 2001-2009. These tidewater glaciers now no longer terminate in a discernible ice shelf, but calve directly into Röhss Bay.

Comparison with previous inventories
Our data support the general recessional trend reported in previous inventories (Rabassa et al., 1982;Skvarca et al., 1995;Rau et al., 2004). Our inventory shows a continued and steadily increasing glacier recession since the first inventory 1977 (cf., 5 Rabassa et al., 1982). Average retreat rates for James Ross Island were 1.8 km 2 a −1 in the period 1975-1988, and 2.1 km 2 a −1 in the period 1988-1993 (Skvarca et al., 1995;Rau et al., 2004 shelves from pinning points (such as Persson Island) can result in enhanced calving and rapid retreat. The acceleration of recession of the ice shelf in Röhss Bay after 2001 was therefore caused by ice-dynamical factors, exacerbated by continued atmospheric warming. Small amounts of growth in some ice-shelf glaciers was observed from 1988 to 2001, which could be because of pinning against Persson Island and the 5 mainland of James Ross Island, and structural and dynamic variations in ice-shelf configuration. However, this advance was subsumed by the overall large glacier shrinkage (cf. Fig. 8). In addition, up-glacier thinning may result in steepening, increased driving stress, faster flow, and short-term advance (cf., Meier and Post, 1987). The recession of marine-terminating glaciers in the Northern Antarctic Peninsula 10 highlights some important trends. For glaciers feeding PGIS, rates of recession were highest during the period of ice-shelf disintegration. Ice-shelf removal can lead to the destabilisation of tributary glaciers, as ice shelves reduce longitudinal stresses and limit glacier motion upstream of the ice shelf (Rignot et al., 2004;Pritchard and Vaughan, 2007;Hulbe et al., 2008). These tributary glaciers began to stabilise and 15 reach a new dynamic equilibrium after 2001 and rates of recession began to reduce once they become stabilised in their narrow fjords. Indeed, fjord geometry has previously been observed to be a major control on tidewater glacier advance and recession rates (Meier and Post, 1987). After 2001, the region therefore entered a period of "normal" glacier recession. The Northern Antarctic Peninsula region has thus had three The majority of the glacier ice lost on James Ross Island is accounted for by the disintegration of PGIS in 1995. For most remaining glaciers, recession rates were highest prior to 2001, with ice-shelf tributary glaciers on Eastern Trinity Peninsula exhibiting particularly rapid recession. Western Trinity Peninsula glaciers had very low rates of 5 recession or have remained unchanged. This differing regional response can probably be attributed to precipitation gradients, exacerbated by climatic warming and differential wind patterns (cf., van Lipzig et al., 2004). As south-westerly winds blow across the Antarctic Peninsula mountains, rising air precipitates moisture as snow in the west, starving the east of precipitation (Aristarain et al., 1987;Vaughan et al., 2003). James Ross Island lies in the precipitation shadow of the Antarctic Peninsula. Higher air temperatures will have resulted in a rise in ELA; smaller accumulation areas mean that glaciers are less able to withstand precipitation starvation.
Previous workers have hypothesised that the disintegration of the ice shelf may have affected the regional climate (Rau et al., 2004). The response of land-terminating 15 glaciers is particularly interesting, because on James Ross Island, these glaciers exhibited their highest annual rates of retreat in the ice-shelf disintegration period. Landterminating glaciers are directly influenced by climatic perturbations and their mass balances are therefore a sensitive indicator of climate variability (Oerlemans, 2005). It therefore appears that the removal of the ice shelf changed the local climate and 20 thus had a strong impact on the island's glaciers. It has been suggested that ice-shelf break up would immediately affect the climate system through the formation of deep water (Hulbe et al., 2004). The removal of the ice shelf would have raised local air temperatures through the availability of more ice-free water in summer. In addition, small glaciers on James Ross Island may be particularly susceptible to changes in 25 local climate because of the large land area, which has a low albedo.

Glacier hypsometry
Topography and hypsometry are important factors in influencing glacier mass balance, and may explain some of the inter-catchment variability in recession rates. Figure 8c shows GAP17 and GAP18 (both outlet tidewater glaciers) receding at different rates. GAP17 remained stationary during the period 2001-2009, whilst the adjacent GAP18 5 retreated at 0.2 ± 0.6 km 2 a −1 during the same interval. However, GAP17 has a very top-heavy HI index of −1.6, compared with 1.8 for GAP18 (cf. Fig. 5d) ). These glaciers have accumulation areas situated at high altitudes, which contributes to their stability. The hypsometric curves of these grounded tidewater glaciers indicate that they will retain large accumulation areas if future recession occurs and thus recession will be slow, rendering them less sensitive to an upwards shift in the ELA. 15 The outlet glaciers draining Mount Haddington Ice Cap typically have a large upland accumulation area on the ice cap, a steep icefall over cirque headwalls, and a very flat and low-lying partly floating or floating tongue. These attributes can be observed in the stepped hypsometric curves of GIJR123 and GIJR115 (Fig. 6). This unusual glacier long profile (Fig. 4g) is bedrock-controlled. Vertical joints in the hard Neogene 20 basalts and hyaloclastites encourage the formation of steep cirque headwalls. The glaciers then erode a deep basin in the softer Cretaceous sandstone and mudstone beneath, resulting in partly floating to floating glacier tongues. Rising sea levels would encourage further grounding-line retreat. 25 If atmospheric temperature rises over the Antarctic Peninsula continue at the current rate of 2.5

Changes in equilibrium line altitude
• C per 50 yr (Vaughan et al., 2003;IPCC, 2007) reached in the Antarctic Peninsula region between the "present" and the year 2040. Given a calculated adiabatic lapse rate of −5.8 • C per 1000 m (Aristarain et al., 1987), and assuming that the topographic ELA values calculated are representative of the current isoline of zero mass balance on these glaciers, an increase of 2 • C would raise ELAs over the Northern Antarctic Peninsula by 345 m. This is illustrated as ELA 2C 5 on the hypsometric curves in Fig. 5, and it is immediately obvious that the differential hypsometry of the glaciers may be an important control on rates of recession in the future. Bottom-heavy glaciers with low-lying ELA MEAN values (such as GAP12) and top-heavy glaciers with high ELA MEAN values (such as GAP13; Fig. 5) are most sensitive to change. In general, however, the tidewater glaciers on Eastern Trinity Peninsula 10 have large, high-elevation accumulation areas. Without a further strong perturbation to the prevailing climate, these large tidewater glaciers will most likely stabilise when they reach their grounding-lines. The ELA 2C calculated for these glaciers (Fig. 5) indicates that even with a 2 • C rise in temperature, glaciers will still retain accumulation areas covering at least 40 % of the glacier surface, assuming that glacier hypsome- 15 tries remain similar. The steady steep slopes on these glaciers also render them less vulnerable; a rise in ELA does not expose a significantly larger area to ablation. Additionally, as these glaciers retreat towards the grounding-line zones, ablation areas will decrease in size, resulting in a more stable mass balance. The tidewater glaciers on James Ross Island are likely to be extremely vulnerable to 20 a 345 m rise in ELA. GIJR27, for example, may be particularly vulnerable as this tidewater glacier has a large, low-lying flat tongue, and ELA 2C plots above the accumulation area. Other glaciers with low-lying tongues will have projected accumulation areas covering only 20-30 % of their area (e.g., GIJR72, GIJR115, GIJR128, GIJR136). These glaciers are likely to continue to retreat very rapidly. Furthermore, the large accumula-

Oceanic temperatures and climate forcing
The calving flux and recession of tidewater glaciers in the study region is likely to be largely controlled by non-linear glacier responses to changes in oceanic temperatures as well as atmospheric temperature and precipitation gradients. Ocean temperatures around the Western Antarctic Peninsula have risen by more than 1 • C since 5 1951 (Gille, 2002(Gille, , 2008Meredith and King, 2005), driven by reduced sea-ice formation and atmospheric warming. Changes in sea-surface temperatures (and sea ice extent) may influence enhance bottom melting and thinning (Holland et al., 2010) and therefore propensity to calve (Benn et al., 2007a), thus accelerating terminus retreat. Tensile strength of glacier ice decreases as ice becomes more temperate, resulting in 10 increased calving rates (Powell, 1991). Furthermore, retreat of the terminus, as a result of increased calving, leads to larger up-glacier stretching rates, greater ice speeds and glacier thinning, further exacerbating increased calving rates for tidewater glaciers (Meier and Post, 1987;van der Veen, 2002), and possibly contributing to short-term small glacier advances. 15 The observed asymmetry and variability in recession in tidewater glaciers may be explained by complex calving processes. The calving flux of tidewater glaciers is controlled by crevasse depth, ice velocity, strain rates, ice thickness and water depth (Benn et al., 2007a). In grounded tidewater glaciers, terminus position may be controlled by the local geometry of the fjord (van der Veen, 2002). Rapid recession of tidewater 20 glaciers in the region is likely to be enhanced by the asymmetry of the calving flux between advancing and retreating glaciers, with increased calving exacerbating the recessional trends (Benn et al., 2007b).

Conclusions
The Antarctic Peninsula is experiencing rapid atmospheric warming, which has re- 25 sulted in the disintegration of many ice shelves (Scambos et al., 2003;Hulbe et al., Scambos, 2008;Cook and Vaughan, 2010). Numerous papers have documented the disintegration of those ice shelves, but this is the first to thoroughly and quantitatively investigate changes in all glaciers in the Trinity Peninsula region. We provide the first detailed inventory of 194 glaciers on Trinity Peninsula and islands to the southeast, with detailed estimates of size, length, elevation ranges, ELA, 5 slope, aspect, hypsometry, morphological descriptors, form and classification. These data will be invaluable to numerical modellers seeking to predict the future behaviour of this climatically sensitive region. The full glacier inventory is available for download from GLIMS (www.glims.org). We measured variability in glacier length and area changes on the Northern Antarc-  Tidewater glaciers on Western Trinity Peninsula remained stable between 1988 and 2009, and to date show only slow rates of retreat. They receive abundant snow from prevailing south-westerly winds, and, unless strong perturbation to this system occurs, will probably remain stable in areal extent. This inference does not take account of glacier thinning, which is observed in other parts of the Antarctic Peninsula, even after 5 frontal stabilisation has occurred (Rott et al., 2011). In contrast, large tidewater glaciers on the Eastern Trinity Peninsula are apparently more vulnerable to continued climate warming, but they may stabilise in the future as they retreat towards their grounding zones.
Glaciers on Trinity Peninsula that fed PGIS have stabilised since 2001, with slower 10 annual recession rates since then. On James Ross Island, the largest areal changes have been from low-lying tidewater glaciers. Annual tidewater glaciers recession rates are likely to continue in response to continued atmospheric warming. The primary control for the widespread glacier retreat in Northeast Antarctic Peninsula is apparently the observed climatic warming, with glacier size, length, slope, type, ELA and altitude 15 exerting a strong mitigating or enhancing role. Despite fears of continued run-away retreat and terminal destabilisation of tidewater glaciers, this study shows that ice-shelf tributary glaciers are more likely to undergo a time-limited period of adjustment. Ultimately, ice-shelf tributary glaciers will find a new dynamic equilibrium, and then retreat slowly in response to warming climates and rising 20 ELAs. We therefore anticipate that the results of this project will be relevant to studies concerned with the more southerly ice shelves that surround the Antarctic Peninsula. The behaviour of PGIS tributary glaciers and their long-term response to ice-shelf removal can be used to model and predict glacier response to contemporary and future ice-shelf disintegration events.

Tables Figures
Back Close

Full Screen / Esc
Printer-friendly Version
Error is less than 3 pixels so is included in the buffer above    Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Table 9. Glacier change, 1988Glacier change, to 2009. Note there is particularly limited data for 1997. Error margin for 2009 includes errors inherent in polygon determination. As analysis of frontal change in 2001 and 1988 assumes no migration of ice divides, error is calculated by using a 3-pixel wide buffer on either side of the ice front only, and is therefore a minimum error of ice front change only. Where data is missing, it is assumed that the glacier did not change in size between that year and the last year for which data was available; therefore, these figures are a minimum estimate. TP = Trinity Peninsula. JRI = James Ross Island. VI = Vega Island. IIC = Island Ice Caps. All = all glaciers. All area is given in km 2 .

Area
Area 2009 Table 5 1  0  V  I  I  G   1  0  V  I  I  G  0  3  V  I  I  G   0  3  V  I  I  G   2  2  V  I  I  G   8  2  V  I  I  G   2  2  V  I  I  G   1  2  V  I  I  G   0  1  V  I  I  G  3  1  V  I  I  Island. (F) Glacier change for and-terminating glaciers (greater than 3 km 2 ) on James Ross Island. (G) Scatter plot demonstrating a weak correlation between change in glacier area and initial glacier size (r 2 = 0.11; see Table 5). (H) Variance in rates of change for land-terminating glaciers on James Ross Island.