Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | on 1 . The − 1 erence − ff Discussions 3–4 myr ± erence near their , or 5 % of its ice ff The Cryosphere 1 − in 12 yr, are near the 1 − erence of up to 30 myr ff region around South Pole encom- or 5 %). No change is detected on (4.1 %), respectively. Bindschadler 2 1 − 1 − erence of 50 myr ff 1,2 1716 1715 erence in glacier speed is below the 8 % level. (25.3 %) at their grounding lines. The slow down ff 1 − (5 %). Most of these changes in glacier speed extend 1 − (6.7 %) and 35 myr , and E. Rignot 1 1 − , J. Mouginot 1 (16.7 %) and 100 myr 1 − This discussion paper is/has been under review for the journal The Cryosphere (TC). Please refer to the corresponding final paper in TC if available. University of California Irvine, DepartmentJet of Propulsion Earth Laboratory, System Pasadena, Science, California, Irvine, USA California, USA dynamics that have almost no impact on the overall ice balance of the region. ner/Ronne Ice Shelves, theWe 12 detect the yr largest di Slessor slow down and with Recovery a Glaciers,tion 12 equivalent Ice yr to Stream velocity shows 6.7 a % di modestBailey, and Rutford, speed 3.3 and up %, (30 Institute myr respectively. Ice Founda- down Streams. On rather the uniformly Filchner Ice withfront Shelf speed, a proper, ice which 12 slowed we yrwe attribute conclude velocity to that di the an ice 12their km streams counterparts and advance in ice in shelves the its in Amundsen ice this and front broad Bellingshausen region, position. seas, in Overall, contrast exhibit with changes in ice i.e., far beyond what wasare previously MacAyeal known. Ice Also Stream slowinggrounding and in lines Byrd the of Glacier Ross 30 with Ice myrIce Shelf a Stream sector is 12 faster yr byon velocity 20 myr di the Ross Ice Shelf along the ice streams’ flow lines. At the mouth of the Filch- partial coverage acquired in 1997,map we of process ice the velocity data changeslargest to with observed assemble a a changes, nominal comprehensive an precisionfrontal of increase regions of detection in the of speed largelarge ice of tabular shelves 100 blocks and myr that are associated willreveal with eventually a the form slow slow icebergs. down detachment On of of Mercerof the and Ross 50 myr Whillans Ice Ice Shelf, Streams ourspreads with data 450 a km 12 upstream yr of velocity di the grounding line and more than 500 km onto the shelf, We report changes in ice velocitypassing of the a 6.5 Ronne/Filchner million and km streams Ross and Ice glaciers Shelves that and constitutemetric their a synthetic-aperture catchment significant radar areas. portion (InSAR) Using coverage of the of first the the full ice region interfero- completed in 2009 and Abstract The Cryosphere Discuss., 6, 1715–1738,www.the-cryosphere-discuss.net/6/1715/2012/ 2012 doi:10.5194/tcd-6-1715-2012 © Author(s) 2012. CC Attribution 3.0 License. Twelve years of ice velocityAntarctica change observed in by RADARSAT-1 and -2 satellite radar interferometry B. Scheuchl 1 2 Received: 13 April 2012 – Accepted: 29Correspondence April to: 2012 B. – Scheuchl Published: ([email protected]) 15 MayPublished 2012 by Copernicus Publications on behalf of the European Geosciences Union. 5 25 15 20 10 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , Rignot et al. ) but the area 2009 , ). Secondly, spaceborne remote cult, but a map of Siple Coast ice ), field measurements continue to ffi ). 2008 , 2003 2008 Floricioiu and Jezek , , erence map is created using the two data ). ff 1718 1717 2002 ´ , ottir et al. ). One goal of STG was to achieve pole-to-pole S. More recently, TerraSAR-X has added capabilities ◦ 2008 , algeirsd ð ). InSAR satellite data coverage in Central re- A South were acquired by the Japanese ALOS PALSAR and Bindschadler et al. ). During the International Polar Year (IPY 2007–2008), the ◦ Joughin et al. Jezek and Drinkwater 2011 , 2010 , S using RADARSAT-2. RADARSAT-2 operates nominally in right looking ◦ Jezek and Drinkwater Joughin et al. Crevier et al. ; The first, and so far only, complete interferometric data coverage of Central Antarc- In 1997, RADARSAT-1 underwent a special orbit maneuver to switch to left looking In this paper, we present a new ice velocity map and a grounding line map based on cause the SAR needs to switch back and forth between the two modes along its orbit. mode in fall 2009 with2011 a ( limited gap fillercontributions campaign of to four complete space the(STG) agencies mapping were ( in coordinated spring byInSAR the coverage IPY of Space the Tasking Group ice the sheets. coast For to Antarctica,the about European standard 80 Envisat right ASAR. The lookingsouth Canadian data Space of cover- Agency 80 committed tomode. fill Planning the and region executing left looking mode acquisitions limits data acquisition be- data strips made data processingstreams and mosaicking was di successfully generatedvelocity using measurements ( these data, complemented with othertica ice with interferometric SAR data was achieved using RADARSAT-2 in left looking hence pointing away frompendent South but Pole typically and south of leavingin 80 a left coverage looking gap mode thatcoverage that provided is are by sensor being this radar de- explored is ( not as comprehensive asmode. RADARSAT-1/2. Most of thewith campaign a was number of devoted interferometric tointerferometric pairs pair radar acquired along amplitude after each mapping that track. The in of limited a Antarctica, data test collect mode, in with combination only with one short We use InSAR data from thethe 1997 Antarctic RADARSAT-1 2009 Mapping mapping RADARSAT-2 Mission campaign. (AMM) Imaging and ofcapability, South which Pole RADARSAT-1 requires used left in looking 2 experimental is mode able in to 1997 use and operationally RADARSAT- since 2008. Most other SAR sensors are right looking, RADARSAT-2 InSAR data collected in falldata 2009. collected by We in RADARSAT-1 revisit 1997. and Asets re-calibrate di to the reveal for InSAR the firstin time the changes in 12 speed yr overchanges a interim. observed vast After extent along of exposing Siple Central the Coast Antarctica the and details ongoing the of evolution Filchner/Ronne the of sectors glaciers data and and conclude processing, ice on we shelves discuss in the these regions. 2 Data be vital in glaciology (e.g. tional Polar Year (IPY) ( velocity. Projects and area coverage have evolvedmost from single recently glaciers, covering over ice the fields vast2011b to ice sheets of Greenlandmains and sparse. Antarctica The ( firstPole took complete place interferometric in 2009 mapping as campaign part covering of a South coordinated acquisition campaign for the Interna- measurements of key areas atcific dense motion temporal patterns spacing. (i.e. For detailed analyses of spe- sensing satellites are a meanscampaigns. to They measure allow ice velocity datathat without collection would the over be necessity vast practically of impossible areas, ground pean to thereby ERS collect providing satellites in in the information the field.Radar early Since (InSAR) 1990’s, the spaceborne data launch Interferometric of have Synthetic the become Aperture Euro- the single most important means of measuring ice Ice velocity is crucialsheets information and for for studying estimating ice the dynamics.the Satellite way mass velocity information balance information has is of fundamentally collected changed has glaciers today. Firstly, Global and simplified Positioning ice System the (GPS) way ground measurements are made, allowing for high precision 1 Introduction 5 5 25 15 20 10 25 20 15 10 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Rignot ; Morena ). For the Egbert and 1999 , 5.4cm vertical 2007 , ± incidence angle of ) and the TPXO6.2 3.8cm ( ◦ ± 2008 , 2.2 and Bamber et al. ± Michel and Rignot ). A second tidal modulation is S; and (2) 43 tracks in Extended sets from speckle tracking. Since ◦ ff in speed (at 28 sets obtained from speckle tracking ff 1 2000 − Padman et al. , to 87 ◦ erent modes were combined: (1) 85 tracks ff erent acquisition times of and RADARSAT-1 ff 1720 1719 ) to estimate vertical position above the ellipsoid erence between epochs is converted into a range Rignot et al. ), but is large when examining temporal changes in ff 2002 1 , opt tide model ( − ), the root-mean-square residual magnitudes for the four 2008 , ). The precision of the tidal model is estimated at cult if coast-to-coast acquisitions are absent or control points of ffi 2012 , ). ). In areas where phase unwrapping is successful, the unwrapped inter- Egbert and Erofeeva ). To cover the entire area, two di 2002 Padman et al. , 2011b 2004 summarizes the and RADARSAT-1 -2 data collected and used in this study. , , 1 Here, we use the CATS2008a Mouginot et al. calibration method for the 1997 data.nunataks, Namely, we divides identify with areas of zero zero surface velocity (domes, slope) in the year 2009 ice velocity map to obtain 3.1.2 Velocity calibration A critical step intask data is processing rendered is di theknown absolute (or calibration zero) of velocitya the are few velocity not 1997 data. known InSAR This aa data map priori. could of be This Antarctica analyzed is thatcalibrated in employs one coverage coast prior of of to studies. the the coast entire The reasons tracks continent recent and why enables assembling provides the only of a application comprehensive of a new, improved Erofeeva of ice shelf velocitiesice (or velocity. 1 kmyr load model ( a potential variation of thebuttressing horizontal and rate driving of stress motion forcesvertical with on component, tides the the due ice error to shelves amountsthe small ( to radar changes 37 illumination myr in away fromterferometric the data vertical acquisitions. direction) per Such 1 an m error tide is change small between compared in- to the absolute value ( main tidal constituents for the load model are between oceanic tides, i.e. hours to days ( for each data point on ice2. shelves The at the corresponding di elevationdisplacement di which is subtracted from the– range or o the interferometric phase – before converting these displacements into ice velocity displacement based on a comparison withIce 16 Shelf independent ( tide records at or near Ross 3.1 Velocity measurements Two dimensional ice velocity isnation extracted of assuming speckle surface tracking parallel and interferometric flow analysis from ( a combi- is a vertical motionChanges of the in ice air shelf pressure caused are by maintenance also of important hydrostatic and equilibrium. operate on the same time scale as The RADARSAT-1 and -2 interferometriclocity data (1997 allows and 2009) the and extraction grounding of line ice position surface (2009 ve- only). et al. ferometric phase is used inthree range instead data of cycles range o areand available combine them in to 2009, reduce we data generate noise. two3.1.1 velocity products per Tide track correction The rise and falllation of of ice the shelves velocity with signal changes over in the floating oceanic extension tide of results ice in sheets. a One tidal modulation modu- 3 Methods two modes. Residual coveragecampaign gaps were caused closed by with acquisitionble a conflicts gap during filler the campaign 2009 that took place in Spring 2011. Ta- et al. in Standard 5 mode forHigh the 4 region mode between for 78 Southeach Pole and track vicinity. to Three consecutive enabletions orbits the were were acquired generation planned along of with two a 24 small day overlap interferograms region per between track. the Acquisi- ground coverage of the Careful planning was required to avoid conflicts and maintain sensor health ( 5 5 25 20 10 15 20 10 15 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | ). The erential ectively ff ff Mouginot ). 2011c , 1998 , erence to detect ff (higher because fewer Rignot 1 − Rignot et al. ected in the best manner. ff 2.3myr ± ). An ice shelf flaps up and down with 2011a 1722 1721 100m. Unlike methods relying on changes in erence maps but not necessarily in the original , ; and the precision of detection of changes in ± ff 1 . The corresponding 2009 velocity variation over − 1 − erence maps requires special care during processing of for the 12 yr period. These are nominal errors. Locally, 2–3 myr ff 2.8myr 1 ± Rignot et al. − ± erences on ice shelves. The 1997 and 2009 velocity maps were ; the 1997 velocity variation is South, which is now available online ( ff 1 − ◦ erence (24 days or one cycle), which we then di erence estimate 3–4 myr ff 10–20 cm) or even kinematic GPS (a few cm) ( ff ± ± erence of ff 1.8myr ). ± 2012 , ectively increasing the number of control points and propagating the calibration pro- The 2009 RADARSAT-2 AMM includes three consecutive cycles per track. This ac- Looking at stagnant areas throughout our study area (more than 77 000 data points ff ing line south ofproduct 80 has a positional accuracy of changes in the oceanicInSAR tide. – This DInSAR vertical motion –laser is at altimetry detected a ( directly precision using of di millimeters, i.e. ordersquisition of strategy magnitude makes better it than possiblethe to same generate time two interferograms di tidal per motion track and spanning map groundingresents line the position. first The comprehensive DInSAR RADARSAT-2 high-resolution, data InSAR-based set mapping rep- of the ground- 3.3 Grounding line detection The grounding line (GL) isbecome the afloat transition boundary in where thecrucial ice ocean. detaches for Knowledge from ice of the sheet bed theof mass to exact ice-ocean balance location interactions calculations, of ice ( the sheet grounding modeling, line and is the analysis and Pecora Escarpment nearthe Foundation velocity Ice di Stream), wethe find area a is 1-sigma variationoverlapping of tracks are available forvelocity the mapping 1997 is data). about Wespeed conclude is that about the precisionerrors of may exceed these values. progressively revised together, inThe an resulting iterative 1997 map fashion,tracks covers to were more included minimize area in track the compared boundaries. mapping. to prior mappingswere as used additional in Siple Dome, Raymond Ice Ridge, Central Antarctica North of Titan Dome, analysis of velocity di in the 2009 map thatchange could map. not be The seen implementation before of but needed a to tidal be correction fixed to was yield an a quality important step for the 1997, leaving out only athe few calibration pairs and with mosaicking poor processesucts. correlation to A levels. yield Great high high care quality is qualityRADARSAT-1 1997 applied 1997 change data in detection map alone. prod- for Similarly,plex change the ensemble calibration detection of and cannot ExtendedSouth mosaicking be High Pole of 4 assembled in the 2009 and using cannot com- Standardof the be overlapping, 5 done coast-to-coast tracks at mode from a RADARSAT-2 Envisat data comparable ASARet around and level al. ALOS of PALSAR precision ( without the help mosaicking errors become visible invelocity di maps if theHere, calibration we expect and identical measurements mosaicking between overlappingthe was adjacent measurements not tracks because are e collectedcalculating with an a average time velocityscale separation over of fluctuations, that 24 for time days, instanceextended period hence only associated that over e with 3 smoothes months, tides. which outdata. minimizes In In shorter-time the the addition, impact process of of the generating long-term data the changes 1997 on collection map, the we found small residual imperfections points. To continue the process,ing new by tracks using are ice addede velocity in from the overlapping calibration calibrated andcess tracks mosaick- across (hence the non-zero entire surveyed velocity), domainthis to approach, obtain a we consistent calibrate calibration and scheme. With mosaic together 102 pairs of RADARSAT-1 data from 3.2 Velocity di The generation of velocity di individual velocity maps because data noisedetect must be subtle minimized changes to in increase speed. the Track boundaries ability are to critical areas where calibration or control points for the 1997 mosaic. A few tracks are calibrated using these control 5 5 25 20 15 10 10 25 15 20 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , , b) ec- 2 ). ff 2002 erence , ff 2011b , ). The 1997 Rignot et al. 2005 , Joughin et al. a) and 2009 (Fig. cient to explain resid- Rignot et al. 2 ffi Haran et al. ), we detect a residual signal which is 4 ) – was studied in more detail at discrete c and 2 1724 1723 1976 ( ). In contrast to these earlier studies, our change ). We may ignore tide model results upstream of the 4 2005 , c) reveals many interesting features, which we will now 2 Thomas 2002 ). The MOA-based grounding line is about the same as the , ). In addition, recent observations from instrumented mammals 4 2011a , set between these two positions is equivalent to the width of the flexure erent times. This demonstrates that the tidal signal is removed e ff ff shows the surface velocity product for 1997 (Fig. Joughin et al. erence map (Fig. shows the impact of our tide correction on data quality and track to track 2 ff 1 cient quality to implement a tide correction that eliminates patch boundary be- ). The o Rignot et al. ffi ). Our map includes a partial coverage of the Filchner-Ronne sector. The 2009 Figure The di overlaid on the MODIS mosaic of Antarctica or MOA ( DInSAR groundling line forual Slessor errors Glacier (see but Fig. suggest this that is the sub-ice-shelf not cavitywhich su at is that another location is sourceexample deeper of than illustrates error previously that thought, in a theof precise tidal the knowledge predictions sea of (Padman, floor thetecting pers. are grounding long-term, comm.). critical line subtle This for position changes estimating and in tidal depth ice motion shelf beneath velocity with ice InSAR. shelves and for de- map is nearly complete,atically with poor residual levels gaps of in temporal West coherence Antarctica of the caused InSAR by signaldiscuss. the ( Most system- notable isWhillans a widespread Ice slow Streams down (formerlyearly on known reports Ross as Ice date Shelf Ice back and Streams to Mercer A and and B). This slow down – DInSAR grounding line forley Bailey Ice Ice Stream, Stream but arebetween some not MOA-based ice and rises accounted in DInSAR forThere, the based in the mouth grounding MOA-based the of lines grounding MOA-based Bai- occurs linedetected product. with is on DInSAR A 150 Slessor km (see large Glacier. too Fig. di far inland from its actual location coverage is the firstextent, interferometric with coverage improved of calibration. the PriorRoss region analysis Ice that of Shelf is the sector, presented with 19972005 a data to mix were its of focused full InSAR on and the non-InSAR data ( locations ( upstream of the grounding line of Whillans Ice Stream (and about 300 km for Mercer line ( detection map reveals the full extentcontinuous of coverage. the We area note of slow in down particular through that its extensive the and slow down extends about 450 km lem, we find thatdata the for Padman Ronne/Filchner model region, uses that a is grounding less line precise than position the based interferometric on grounding optical boundary. Without tide correction,shelf velocity it in would our not studyto area, be the because possible tidal the to actual modulation.of signal detect We su is changes find comparable in in thattween magnitude ice the tracks tide on estimates theacquired from ice at the shelf di Padman astively, model well consistently are as and inconsistencies accuratelythe between in tidal overlapping our tracks correction data. isGlacier A and questionable. Bailey few Ice For notable Stream instance (see exceptionsnot Figs. spatially near exist consistent where the with the groundinggrounded signal ice, line recorded which of in suggests the Slessor that surrounding the areas, tidal including correction on is in error. Looking at this prob- 4 Results Figure 2011a zone, which is notand a also constant. with The the tidal widththerefore cycle of because not the of possible the flexure to visco-elastic zone compareconfidence. nature varies the The of with the 1997 2009 deformation. ice and grounding It thickness line 2009future is changes. will mappings however at serve this as time a with reference high for measuring ment of theInSAR grounding data line from 1997 position. because Groundingphase these coherence data line do to mapping not map includethe cannot grounding outer repeat be lines limit pairs. of Methods do done tidal using not flexinggrounding using instead, line. detect which This the is was typically demonstrated grounding in several km line the seaward case per of of the se Evans actual Ice but Stream in ( surface slope, aspect, or visible features, the DInSAR approach is a direct measure- 5 5 10 15 20 25 15 25 20 10 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , ) 1 1 − − level 1 erence ). Bind- − ff 1 (25.3 %), − erence of 1 ff − erence of less and 30 myr ff 1 − Joughin and Padman at about 50 myr v or 5 %). A speed up can be 1 − between 10 myr (16.7 %) and 100 myr v 1 erence in % of the glacier speed at − ff erence between 1997 and 2009 also ff up to 20 myr 1726 1725 erences between stable shelf portions and fast v ff erence between the two epochs exceeds 100 myr ff erence overlaid on the MOA, the 2009 grounding line, as ff we noted the presence of residual errors in tidal error correc- . Wide cracks are observable in the respective radar amplitude 1 4 , we provide the velocity di erent from the pattern on the eastern half which is determined by − erence most likely results from the pre-calving event that hap- ) and plot changes in speed along these flow lines at the center v ff ects all the tributaries of these ice streams up to their presumed ff . Locally, velocity di ff 1 or 6.7 %). Here, the slow down (d − or 5 %. On nearby Slessor and Recovery Glaciers, the decrease in 1 − or 4.1 % at the grounding line. The slow down extends roughly 100 km 1 2011d − 1 , − shows a more detailed analysis for the Ross Ice Shelf region. The map shows detail for the Ronne-Filchner Ice Shelf. The 1997 coverage is less erence, d ff 3 4 ), i.e. at the percent level. This means that we observe little change in speed of ) mixes data from 1997 and 2000 and is not used here. The DInSAR grounding 2 erence near the grounding line is up to 50 myr erence pattern on the western portion of Ross Ice Shelf is determined by these two Figure Figure At the northwestern edge of the Ross Ice Shelf, we observe a zone of speed up ff ff 2003 Rignot et al. about 10 myr Whillans and change signatures. extensive compared to Ross Ice( Shelf. A more complete mapline by is shown in green. Forgrounding Slessor line Glacier, we (yellow/black also dashed indicate line)a the discussed yellow erroneous arrow MOA-based earlier. in In Fig. tion, the i.e., region the indicated real by changeIce in Stream, ice we velocity detect is a masked small by these slow residual down errors. in On speed Bailey with a 12 yr velocity di Stream slowed down slightlyof between up 1997 to and 30 myr 2009can with be a traced 12 to about yrdi 300 velocity km di upstream of theice grounding streams and line. is The di respective velocity di respectively. Most likely, as amust result have of migrated slow downstream between downcreases 1997 and from and the thickening, interior 2009. their toward The groundingsubsequently the decrease remains line grounding fairly in zone, constant speed to on in- peak theschadler ice at Ice shelf the Stream (with displays grounding d a lineis modest and speed only up partially near the coveredtraced grounding in line up region, our to which data about (d 200 km upstream of the grounding line. I contrast, MacAyeal Ice the grounding line insults the show text that Byrd below Glacier to slowedthan down provide 35 slightly myr some with context a 12 forupstream yr each of velocity glacier. di the Our grounding re- and line Whillans and Ice decreases Streams with experience a distance. more For significant comparison, slow down. Mercer Their 12 yr velocity velocity di pretation. The 2009 grounding line is indicated for reference. In addition to the 12 yr over 12 yr versus(Fig. absolute speeds inthe the glaciers range and ice of shelves several between hundred the two meters dates. pershows year the 12 yrwell velocity as di the 19972004 and ice 2009 front ice( position. fronts. We In generate addition,of flow MOA major lines provides outlets. using information Markers the on are 2009 the set velocity every vector 100 km map on the map and plots to ease inter- imagery at the transitioneas, which boundary justifies between our labeling speedevents of up these represent areas the and as largest no pre-detached.speed changes These speed up two in up pre-calving and speed the for in widespreadof both our slow the ar- down map. of other Outside the of observed Mercer these an changes Whillans areas are Ice of small Streams, most in magnitude, i.e. at the 10–20 myr that coincides with aular portion iceberg. of The thethough ice velocity the shelf di main thatpened is di past about 1997. to A detachof similar and pre-calving the event form Filchner is a Iceexceeds detected tab- 100 Shelf, along myr where the the northeasternblocks velocity front reach di 150 myr source, i.e. where theon ice limited motion discrete signal points meetsupstream up our to for noise 400 Mercer km floor. Ice upstreamreaching Prior Stream. for more studies Whillans Our than reported Ice result 500 Streamthe km also and ice onto shows stream 150 km Ross changes the determine Ice extent the Shelf, of regional an the velocity indication changes slow on that down the the shelf. signatures of Ice Stream), and a 5 5 25 15 20 10 25 15 20 10 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | . ; ). 1 − 2006 2008 , ). Since , ), which 4 for Filchner up to about erence over 1999 1 ff v , − Pritchard et al. erent times and for Rignot et al. , which is negligible ff 1 Gudmundsson − , equivalient to 6.7 % and 1 ), no change in ice surface − Lazzara et al. or 5 %. Coverage of Rutford Ice 1998 , 1 − erence reaches 50 myr ff and Henry Ice Rise. The 12 yr velocity Rignot ff or 5 %). We detect only a small slow down 1 1728 1727 − erence is up to 30 myr level versus ice shelf speeds of up to 1000 myr ff 1 − of less than 10 myr : 1997 and MOA ice front) ( ) and a mass budget close to zero ( 4 or 7.5 % of the Ice Shelf velocity in the area. We note v 1 ). The relationship is complex, with diurnal and fortnightly − 2009 , less than 30 myr 2008 v , erence) in the Filchner/Ronne sector. The impact of this slow down on erence at the 50 myr ff ff ´ ottir et al. Pritchard et al. , or 3.8 %). 1 − ) detect a slight thickening of Slessor Ice Stream, which is consistent with our erent tidal modulation, we are quite sure that the changes in speed detected on the erence is about 30 myr algeirsd Recovery and Slessor Glaciers are exhibiting the largest slow down (when expressed On the Filchner Ice Shelf, the magnitude of the slow down is real, but small with a 12 On Ronne, multiple iceberg calving events caused a retreat of the ice front between On the Filchner/Ronne Ice Shelves proper, we detect a widespread and spatially rel- ff ff ð 2009 compared to the mass budget of the entire region. In a recent study, elevation ( in velocity di grounding line fluxes is at the percent level, or less than 1 Gtyr yr velocity di Between 1997 and 2009, the ice front advanced by about 12 km (see Fig. report of a morealtimetry extensive data area of from slight Recoveryspeed slow being Ice down. confined Stream, closer No to which thickening the is is grounding apparent consistent line in region with in the slight our data. changes in ( increased side shear andfirmed would with explain model the simulations. slow down. This aspect needs to1997 and be 2004 con- (compare Fig. 2004, the icewhere front large has icebergs been calving after advancing. 1997 The resulted in situation a is net retreat similar of on the ice Ross front. Ice Shelf, residual signal is smallstream, however, which it does suggests not that1997 extend Rutford and into Ice 2009 the Stream surveysthat grounded did revealed at part stable not of a grounding change the detectable line speed ice positions level. between ( This the is consistent with earlier studies fluence on the change map.5 The % measured of changes the in absoluteslight speed velocity, over slow with 12 down yr slight downstream. are speedtidal less This up correction than pattern upstream rather of is thanto the suggestive a inaccurate grounding of real estimates line residual change and of uncertainties in the ice in depth dynamics. of Tidal the errors sea could floor be in due the sub-ice-shelf cavity. The period of 24 days is larger than one periodicity, there is potential for residual tidal in- periodicity and variations oscillating between 20 % and 12 %. While our observation di ice shelves are not tidaltogether related. correctly, If especially they at were, patch it boundaries.therefore would The represents be signal a impossible observed true to on mosaic change the the in ice data overall shelf speed between the two epochs. 5 Discussion The velocity of Rutford Ice StreamA is influenced by oceanic tides ( small and mainly covers thedi area around Kor here that because oursensitive InSAR to transitory data signals combinemore, occurring data because at acquired our much 24-day mosaic lower apart, temporal combines we frequencies. many Further- are tracks acquired not at di tide correction. Even with thisthe uncertainty grounding present, line the (d dataon reveal Institute velocity Ice increase Stream, at Stream i.e., d is spotty, particularly12 around yr upstream the of grounding the15 line. myr grounding line The indicates velocity a di higher speed in 2009atively (d uniform slow down. TheIce 12 Shelf, yr or velocity di 5 % of its ice front speed. The observed region on Ronne Ice Shelf is speed is larger. The 12 yr3.3 % velocity di for Slessor andwithin 100 Recovery km Glaciers, upstream respectively. of Thispervasive the over slow grounding a down line vast region is areaa in confined for spike the to Slessor. in case The speed of flow change Recovery line near but for the more Foundation grounding Ice line, Stream which shows may indicate residual errors in 5 5 10 20 25 15 25 15 20 10 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | . 1 − erent ff Helsen ( 1 − , Byrd flow 3 . Our change map ). We conclude that 1 on average and de- − 1 − 2008 , 3–4 myr ± erence for all glaciers and ice 2–3 myr ff ± ) report thickening of stagnant Kamb Stearns et al. ) the ice streams and ice shelves in the 2009 ( 2008 1730 1729 , ). Our analysis shows a general agreement of the ected by tides and the drainage of sub-glacial lakes. ff ). The authors show a 48 yr record of ice velocity that 2005 Pritchard et al. , ). Rignot et al. 2008 , , but thinning of Whillans and Mercer Ice Streams at 20 cmyr 1 2002 2012 − , , The authors wish to thank Mr. Yves Crevier (Canadian Space Agency) for ). A similar analysis should be pursued for the 1997–2009 time period. Stearns et al. Thomas et al. Joughin et al. 2008 , Other than those, the Filchner/Ronne sector shows few changes, with most ice On Ross Ice Shelf sector, The slow down of Whillans and Mercer Ice Streams was analyzed using an earlier, velocities are a Aeronautics and Space Administrations MEaSUREs and Cryospheric Science Programs. the past 12 yr, whicha suggests that strong the impact ice on dynamics the of mass the budget entire ofAcknowledgements. region the does Antarctic not continent. have championing the 2009 AMM RADARSAT-2 and for hisin support 2011. in Mr. filling remaining Gord coverage Rigbyexecuting gaps RADARSAT-2 and data Mr. acquisitions. Thomas RADARSAT-1 data Loganrie were from Padman provided MDA provided by were advice ASF. Dr. instrumental onCalifornia Lau- the in Irvine tidal planning and models. and at This Caltech’s work Jet was Propulsion performed Laboratory at under the a University contract of with the National proper are defined byOverall, the however, ice the streams observed and changesregion. glaciers have We little that therefore impact constitute conclude on its that theand in catchment mass contrast Bellingshausen area. with balance Seas their of ( counterparts the broad in region the under Amundsen investigation herein have not been changed in a significant way in streams slowing downstreams slightly. in The the 12 region yr isbudget below of velocity 8 the %, di which region.Filchner is The of Ice most little Shelf, distinct consequenceSlessor consistent for signal, Glacier the with however, over overall is a an mass a largewe ice sector slight confirm shelf that the slow warrants slow advance, down further downference and of of study. of On Mercer the a 16.7 and the % small Whillans Ross and Icethe Ice slow 25.3 Stream, %, Shelf, entire down with respectively. spatial a Our of 12 extent changestream yr of map and velocity the shows, dif- onto for change. the the The shelf. first slow Regional time, down, variations extends in hundreds velocity changes of on km Ross up- Ice Shelf and provides ice motiontection results of with changes a inprovides precision speed new, important with of information a aboutlack precision the thereof. of evolution Filchner of and about this Ross sector Ice Shelves of exhibit Antarctica, signs or of pre-calving events. despite the existence of ato mini maintain surge a between rather 2005 steady and flow 2007, regime Byrd over the Glacier last continues partial few version decades. of theyears 1997 ( map, combined with ground measurements from di et al. line) are consistent with the pre-surge values ( mean annual deceleration for theslow trunk of down Whillans at Ice the Streampared grounding but to an line increased pre-1990 near rate ground Crary of the measurements. Ice 2009 A Rise velocity more for map ourwhere detailed ( observation with shelf-wide period comparison pre-1990 com- of ground measurements will be discussed else- Ice Stream at 60 cmyr The latter is opposite toning must what therefore we be expect related to forfor changes ice the in stream surface time mass slow period balance. down. Indeed, 1995–2003that The an revealed analysis may observed a thin- have decrease caused in an snowfall during average that decrease time in period surface elevation of 6 cmyr We present a 12 yrlargest change ice record shelves in and ice many motion in major Central ice Antarctica, streams covering and the glaciers. two Data quality is excellent A 10 % speeddrainage up ( was observedreveals between no 2005 changes and inThe 2007 speed velocity as along record the ahas main prior result trunk not of to of been sub-glacier 2000 Byrd observed. is between The 1960 sparse, 2009 and so velocities 2005. a of surge Byrd cycle Glacier may (See be Fig. possible but 6 Conclusions 5 5 25 15 20 10 15 20 10 25 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , , , i 1719 , available , last access: 0183:EIMOBO h 1727 1728 1726 , 1727 , 2006. , doi:10.1126/science.1153894 1722 , 2003. doi:10.1007/s00254-008-1393-y 1717 doi:10.3189/002214311796406158 doi:10.1016/j.epsl.2007.03.005 , 2008. 1729 , 1717 1729 1732 1731 1724 1728 , , 1718 1729 , 1724 1721 doi:10.1038/nature05430 , 1720 1724 1718 , 2003. cient inverse modeling of barotropic ocean tides, J. Atmos. , 1999. 1728 1718 ffi , 2009. 1720 doi:10.1029/2003GL016941 http://dx.doi.org/10.1175/1520-0426(2002)019 , 2008. , 2005. 1719 , 2002. , 2009. 1720 1720 doi:10.3189/002214308786570872 1718 , , 2002. ´ ottir, G., Smith, A. M., Murray, T., King, M. A., Makinson, K., Nicholls, K. W., and http://www.sciencedirect.com/science/article/pii/S0012821X07001628 1729 1717 1717 algeirsd similation of ICESatdoi:10.1029/2008GL035592 laser altimetry over ice shelves,ning Geophys. on Res. thedoi:10.1038/nature08471 Lett., margins 35, of L22504, the Greenland and Antarctic ice sheets, Nature, 461, 971–975, mined by accumulation variability, Science, 320, 1626–1629, gaard, E., Davis, C. H., Li, Y., and Goodwin, I.:2008. Elevation changes in Antarctica mainly deter- satellite-radar interferometry, J. Glaciol., 44, 469–476, 1998. methods, in preparation, 2012. Antarctica (MOA) image map,Colorado, National USA, Snow updated 2006, and 2005. Ice Data Center, Digital media, Boulder, ing radar images: comparisonGlaciol., of 45, the 93–100, phase 1999. correlation method with radarand interferometry, J. mission planning, Can. J. Remote Sens., technical note, 30, 304–320, 2004. Antarctica, Nature, 444, 1063–1064, doi:10.1029/2001JB001029 doi:10.1029/2005GL024319 2.0.CO;2 IPY: TerraSAR-X imagesdoi:10.1007/s00254-009-1743-4 the System, Environ. Geol., 58, 457–458, Geophys. Res. Lett., 30, 1477, stream velocities: observation and analysis, J. Geophys. Res.Horgan, (Solid Earth), H., 107,eration Peters, 2289, L., of Winberry, Whillans P., Ice Das, Stream, S. B., Westric synthetic and Antarctica, aperture radar, Catania, Geophys. J. Glaciol., G.: 56, Res. 1026–1042, Continued Lett., decel- 32, L22501, doi:10.1080/10889379909377676 Ocean. Tech., 19, 183–204, 2011. On the recent calving of icebergs from the Ross Ice Shelf, Polar Geogr., 23, 201–212, lessons learned, Polar Science –ference, Global 8–12 Impact, June International 2010, Polarretariat, Year Oslo, Oslo Abstract Norway. Science Oslo, 382880, Con- Research 2010. Council of Norway, IPY-OSC Sec- snapshot year: an update, Environ.2008. Geol., 55, 1379–1380, ternal forcing, Earth Planet.online: Sc. Lett., 257,May 1–13, 2012, 2007. slip motion indoi:10.3189/172756403781816284 the mouth of Whillans Ice Stream, Antarctica, Ann. Glaciol., 36, 263–272, Behar, A. E.: Tidal influenceflow on and Rutford basal Ice processes Stream, from West54, closely Antarctica: spaced 715–724, observations GPS of and surface passive seismic stations, J. Glaciol., ð Padman, L., Erofeeva, S. Y., and Fricker, H.Pritchard, H. A.: D., Improving Arthern, R. Antarctic J., tide Vaughan, D. models G., and by Edwards, L. as- A.: Extensive dynamic thin- Helsen, M. M., van den Broeke, M. R., van de Wal, R. S. W., van de Berg, W. J., van Meij- Rignot, E.: Hinge-line migration of Petermann Gletscher, North Greenland, detected using Mouginot, J., Rignot, E., and Scheuchl, B.: Mosaicking of ice motion in Antarctica: data and Haran, T., Bohlander, J., Scambos, T., Painter, T., and Fahnestock, M.: MODIS mosaic of Michel, R. and Rignot, E.: Flow of Glaciar Moreno, Argentina, from repeat-passMorena, shuttle L. imag- C., Pearce, C., Olfert, L., Elder, C., and Rolland, P.: RADARSAT-2 order handling Gudmundsson, G. H.: Fortnightly variations in the flow velocity of , West Joughin, I. and Padman, L.: Melting and freezing beneath Filchner-RonneJoughin, Ice Shelf, I., Antarctica, Tulaczyk, S., Bindschadler, R., and Price, S.Joughin, F.: Changes I., in Bindschadler, West R. Antarctic ice A., King, M. A., Voigt, D.,Joughin, Alley, I., R. Smith, B., B. Anandakrishnan, E., S., and Abdalati, W.: Glaciological advances made with interferomet- Floricioiu, D. and Jezek, K.: International viewpoint and news – Antarctica during the Egbert, G. D. and Erofeeva, S. Y.: E Lazzara, M. A., Jezek, K. C., Scambos, T. A., MacAyeal, D. R., and van der Veen, C. J.: Jezek, K. and Drinkwater, M.: International viewpoint and news – global interagency IPY polar Crevier, Y., Drinkwater, M., and Jezek, K.: IPY-STG SAR coordination group synopsis and Bamber, J. L., Alley, R. B., and Joughin, I.: Rapid response of modern dayBindschadler, ice sheets R. to A., ex- Vornberger, P. L., King, M. A., and Padman, L.: Tidally-driven stick- A References 5 5 30 25 30 25 20 10 20 15 15 10 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , , , last ◦ erential ff 47.2 ◦ 1730 , 44.3 1736 doi:10.1038/ngeo356 1722 , 1728 a 1724 doi:10.1029/2011GL047109 ◦ , ◦ 14) to Pole + 57 ◦ , 2008. 39.5 1719 , 86.5 1717 ◦ http://nsidc.org/data/nsidc-0484.html ◦ ◦ to 87 S5 EH4 ◦ 36.7 , 2011b. 1734 1733 doi:10.1038/ngeo102 , last access: May 2012, 2011c. 1724 ◦ 1720 1729 34.1 , 2000. ◦ 1725 ST2 ST3 ST4 ST5 ST6 ST7 1724 , 1723 , doi:10.1126/science.1208336 limited gap filler in 2011) modes used for interferometric data collects in Antarctica. + 1722 1729 Left looking RADARSAT-1 (September–October 1997) and RADARSAT-2 (February– 14 tracks were aquired in 2011 to close data gaps from the 2009 AMM. Coverage (lat) 77.5 RADARSAT-1 1997 Range SpacingAzimuth Spacing 8.1 mIncidence 5.3 Angle m 11.6 m 28.0 5.0 m 11.6 mRADARSAT-2 5.1 m 2009 Range 11.6 Spacing mAzimuth Spacing 5.2 mIncidence Angle 11.6 m 11.6 m 5.0 m 5.2 m 11.8 m 5.1 m 41.3 11.8 m 5.1 m No. of Tracks 43 85 ( No. of Tracks 46 10 12 12 9 13 a doi:10.1029/1999JC000011 low the Filchner andferometry: comparison Ronne with Ice tide Shelves, Antarctica, model predictions, using J. synthetic Geophys. aperture Res., radar 105,Meijgaard, inter- 19615–19630, E.: Recent Antarctic icemodelling, mass Nat. loss Geosci., from 1, radar 106–110, interferometry and regional climate satellite radar interferometry, Geophys. Res.2011a. Lett., 38, L10504, Antarctica, in preparation, 2012. 1427–1430, NASA EOSDIS DAAC at NSIDC, available online: outlet glacier caused by2008. subglacial floods, Nat. Geosci., 1, 827–831, sheet, Nature, 259, 180–183, 1976. http://nsidc.org/data/nsidc-0498.html access: May 2012, 2011d. ferential Satellite Radar Interferometry, NASA EOSDIS DAAC at NSIDC, available online: Table 1. April 2009 Rignot, E., Padman, L., MacAyeal, D. R., and Schmeltz, M.: Observation of ocean tides be- Rignot, E., Bamber, J. L., van den Broeke, M. R., Davis, C., Li, Y., van de Berg, W. J., and van Rignot, E., Mouginot, J., and Scheuchl, B.: Antarctic grounding line mapping from di Thomas, R., Rignot, E., and Scheuchl, B.: Substantial velocity changes on the Ross Ice Shelf, Rignot, E., Mouginot, J., and Scheuchl, B.: Ice Flow of the Antarctic Ice Sheet, Science, 333, Rignot, E., Mouginot, J., and Scheuchl, B.: MEaSUREs InSAR-Based Antarctica Velocity Map, Stearns, L. A., Smith, B. E., and Hamilton, G. S.: Increased flow speed on a largeThomas, East Antarctic R.: Thickening of the Ross Ice Shelf and equilibrium state of the West Antarctic ice Rignot, E., Mouginot, J., and Scheuchl, B.: MEaSUREs Antarctic Grounding Line from Dif- 5 15 10 20 25 Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

). Blue +100

overlaid on 0 2011b

,

(b) dv [m/yr] -100 Berkner Island

700 500

900

Henry

300

Ice Rise 100 and 2009 Rignot et al.

100 (a)

500 300

700 Ronne Ice Shelf Ronne Ice 900 velocity contour [m/yr] Korff 100 c dv 12 years (2009-1997)

Ice Rise

-100 m/yr -100 m/yr 0 +100 m/yr >3000

1000

100 10

IS: Ice Stream Gl: Glacier v_1997 [m/yr] <1.5 Byrd Gl Recovery Gl Ross Ice Shelf Slessor Gl applied 1736 1735 Mercer IS (2009–1997) for the entire region. Superimposed on Foundation IS Tide correction Tide Baily IS v Tide not corrected Tide Roosevelt Island Institute IS Whillans IS MacAyeal IS MacAyeal Bindschadler IS b Ronne- Filchner Ice Shelf Rutford IS Van der Veen IS der Veen Van

Crary

Ice Rise >3000

1000

erence d

100 ff 4000 10 3500

3000 v_1997 [m/yr] 2500 <1.5

2000 erence calculated using two consecutive RADARSAT-2 pairs acquired 0 500 1000 km ff

Siple

Dome Siple Coast Siple

dv 24 days (2009)

Ross Ice Shelf

500

1000

500

1500

1000 1500 2000 -40 m/yr 0 +40 m/yr elevation contour [m] shows the velocity di 0 250 km a 1000 Impact of tide correction to InSAR based ice velocity estimates. The example on the Ice surface velocity maps for Central Antarctica for 1997 (c) are velocity contour lines taken from the IPY ice velocity map ( tones indicate a slownear down. Roosevelt The Island two and Filchner dark Ice red Shelf) regions indicate on pre-calving the events. ice shelf edges (Ross Ice Shelf Fig. 2. in 2009 without (top) andference map with spanning (bottom) 12 tide yr correctionimpact applied. (tide of The correction tide right was correction applied examplethe for shows importance to a of the the this dif- 2009 1997 step RADARSAT-2 data). data for The absolute (top: velocity calibration not and corrected; to bottom: generate a tide seamless corrected) mosaic. shows Fig. 1. left shows a velocity di MOA. (c)

Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

dv [m/yr] dv dv [m/yr] dv dv [m/yr] dv

dv [m/yr] dv dv [m/yr] dv dv [m/yr] dv dv [m/yr] dv 100 50 0 -50 -100 100 50 0 -50 -100 100 50 0 -50 -100 250 800 -50 -100 100 50 0 100 50 0 -50 -100 100 50 0 -50 -100 100 50 0 -50 -100 1000 250 500 800 1000 dv dv 800 200 dv dv dv 600 200 800 400 600 600 150 150 600 300 400 400 Byrd GL 400 100 Baily IS Whillans IS 100 400 200 Slessor GL MacAyeal IS Recovery Gl Foundation IS 200 dv R-1 1997 R-2 2009 R-1 1997 R-2 2009 R-1 1997 R-2 2009 200 50 200 dv R-1 1997 R-2 2009 R-1 1997 R-2 2009 R-1 1997 R-2 2009 R-1 1997 R-2 2009 50 200 100 B Distance B’ 0 0 Y Distance Y’ 0 F Distance F’ E Distance E’ S Distance S’ R Distance R’

B Distance B’ 0 0 0 0 0

0 0

0 0 0 0

Ice velocity [m/yr] velocity Ice Ice velocity [m/yr] velocity Ice

800 600 400 200 [m/yr] velocity Ice Ice velocity [m/yr] velocity Ice 800 600 400 200 [m/yr] velocity Ice 800 600 400 200 Ice velocity [m/yr] velocity Ice 800 600 400 200 [m/yr] velocity Ice 800 600 400 200 800 600 400 200 800 600 400 200 1000 1000 1000 1000 1000 1000 1000 Y

R

dv [m/yr] dv dv [m/yr] dv dv [m/yr] dv Y’ [m/yr] dv 1738 1737 100 50 0 -50 -100 0 -50 -100 100 50 100 50 0 -50 -100 0 -50 -100 100 50 S 500 300 1200 1000 2009 Ice front 2009 Grounding line 1997 Ice front Flow lines 2009 Ice front 2009 Grounding line 2009 Grounding 1997 Ice front Flow lines 250 1000 dv dv dv 400 800 A’ F B 200 800 B’ 300 600 B’ 0 250 km 500 S’ 150 600 D’ 200 Rutford IS Mercer IS 400 Institute Gl A I 100 400 Bindschadler IS R’ erence. The vertical dashed line is the 2009 grounding line. E’ R-1 1997 R-2 2009 R-1 1997 R-2 2009 ff F’ 100 dv R-1 1997 R-2 2009 R-1 1997 R-2 2009 50 200 200 I Distance I’ 0 0 U Distance U’ Distance U 0 0 0 A Distance A’ 0

D Distance D’

0 0

Ice velocity [m/yr] velocity Ice

800 600 400 200 [m/yr] velocity Ice 800 600 400 200

B

1000 1000 Ice velocity [m/yr] velocity Ice Ice velocity [m/yr] velocity Ice U’ I’ 800 600 400 200 200 800 600 400 1000 1000 Velocity Difference Velocity erence map (2009–1997) detail, 1997 and 2009 ice front, and flow line plots erence map (2009–1997) detail, 1997 and 2009 ice front, and flow line plots Velocity Difference Velocity ff ff 0 250 km 500 D U -100 m/yr 0 -100 m/yr 0 +100 m/yr E -100 m/yr 0 -100 m/yr 0 m/yr +100 Velocity Di Velocity Di erence. The vertical dashed line is the 2009 grounding line. ff for Ronne Filchner IceThe Shelf overlaid yellow-black dashed on MOA. lineerrors Each is (marked flow an line by incorrect a contains groundingand markers yellow line 2009 arrow). every leading and 100 The km. the to flow velocity residual line di tide graphs show correction the absolute velocities for 1997 Fig. 4. for Ross Ice Shelforientation. overlaid The on flow line MOA. graphs show Eachdi the flow absolute velocities line for includes 1997 and markers 2009 every and the 100 velocity km for easier Fig. 3.