PROGRESS IN ADDRESSING SCIENCE GOALS FOR GLACIER OBSERVATIONS BY MEANS OF SAR

Transcription

1 PROGRESS IN ADDRESSING SCIENCE GOALS FOR GLACIER OBSERVATIONS BY MEANS OF SAR Frank Paul, Thomas Nagler, Dana Floricioiu, et al. PSTG, 13/9/2016

2 Panmah, Karakoram

3 Observed glacier products and sensors Product Area Facies Elev. change (ALT) Elev. change (DEM) Velocity Sensor Optical (Radar) Optical Altimeter (opt./radar) Optical / Radar Optical / Radar Format Shape file (vector) Shape (line) Shapefile (csv) Geotiff (raster) Shapefile (csv) Sources GLIMS/RGI raw data ask scientist ask scientist (WGMS) ask scientist (GLIMS) Correction/ Validation Validation datasets Manual editing (visual) High-resolution data (Google Earth), coherence images Manual editing (visual) WGMS observations Filtering (slope, outlier) IceBridge / Cryovex co-registration, stable ground differences ICESat, LIDAR & national DEMs Filtering, stable ground, in-situ Automatic GNSS Challenges Global consistency, debris, snow, clouds, shadow, water contrast, Aufeis, precise DEMs for topogr. correction clouds / footprint size, interpolation, short time series Co-registration, data voids, penetration, cell size, projection, sensor biases (jitter) orthorectification of input data (DEM accuracy), lack of contrast (optical) Archived datasets Corona, Hexagon Landsat MSS / TM / ETM+ (SLC on) Landsat MSS / TM / ETM+ (SLC on) ICESat GLAS EnviSat RA-2 SRTM, GDEM2, RAMP, NED / CDED, GIMP, SPOT-SPIRIT ERS-1/2, PALSAR Envisat ASAR, Landsat TM / ETM+ (SLC on) Ongoing missions Landsat ETM+/OLI Terra ASTER, SPOT, Sentinel 2A Landsat ETM+ /OLI, Sentinel 2, SPOT, ASTER Cryosat 2 ASTER14 DMO, TanDEM-X PALSAR 2, TerraSAR-X, Cosmo-Skymed, ASTER, Landsat OLI, Sentinel 2A Forthcoming Sentinel 2B Sentinel 2B Sentinel 3, ICESat 2 GDEM3, ALOS-DEM, World-DEM Sentinel 2B

4 Requirements for glaciers (technical) Satellite-derived products are complementary to field observ. in Glaciers_cci: Area (outlines/inventory), elevation change (alt/ddem), velocity (opt. / MW) not yet (but soon in Option 3): Facies (snow mapping on glaciers) Glaciers are globally distributed (200000) & require high res. Area/inventory creation is organized within GLIMS, but region selection is science based Elev. change / velocity: driven by data availability, science agenda, special events Techniniques are mature but some require manual editing velocity: ~ automatical, elev. change: pre-processing manual, area: post-processing manual accuracy assement difficult (lack of reference data) but not impossible (precision more easy) Processing lines require a precise DEM as a key input either for pre-processing (orthorectification) or deriving products (divides, topo parameters) current spatial resolution and accuracy of DEMs is not sufficient in steep mountain terrain glacier surface elevations change quickly (100 m / decade), causing orthorectification errors the DEMs used for orthorectification (PlanetDEM) need to be freely available as well Systematic acquisitions of time series of DEMs for estimation of ice volume changes highly needed (exploit TDX archive)

5 Requirements for glaciers (observation) Observation requirements are different for each product Optical sensors: must be cloud free Area: only end of ablation period in specific years, facies (EOSS): dto. but in all years, velocity: during ablation season (each year), elevation change: as area but less critical Microwave: winter preferred for velocity/dems, summer for coherence images (area), short repeat intervals, C or X-band work but for elevation change snow penetration is important Timing for outlines dependes on latitude and climate Different in tropical and monsoon regions (end of dry period) Often seasonal snow in maritime climates and clouds in mid-latitudes (=> try each year) Global coverage ok (optical) but limited impact on acquisition Archive consolidation at USGS is highly appreciated, reprocessing of GLS1990 is encoraged reprocessing of ASTER archive to L1T is also appreciated, VIS is still fine for DEM & velocity Access data from other space agencies (Brasil/China/India)? not easy to find and even more difficult to get...

6 Sensors: timelines and applications Paul (Glaciers_cci) Paul

7 Glacier mapping and corrections TM3 TM5 snow crevasses debris bare ice rock Ratio manual work! Mask Outlines Paul

8 Accuracy: manual glacier delineation (Landsat) Landsat data: USGS Paul et al. (2015)

9 Glacier shrinkage in Patagonia: km N Landsat data: USGS

10 Time series of area changes Landsat data: USGS Paul (Glaciers_cci)

11 Glacier products from Sentinel-2 data Glacier outlines Thresholds applied on NDSI, Band 2, Band 8 & Ratio B2/B4 Snow/ice areas on glaciers Threshold applied on Band 8 within glacier outlines (AGI99) Copernicus Sentinel Data

12 Area altitude distribution of snow/ice areas 12

13 Glacier area changes Glacier Name Area 1969 [km²] Area 1998 [km²] Area 2015 [km²] Untersulzbach Kees 3,961 3,721 3,125 Viltragen Kees 2,467 2,127 1,698 Schlaten Kees 9,678 9,316 7,713 Rainer Kees 3,628 3,511 3,179 Dorfer Kees 4,552 3,798 2,848 Obersulzbach Kees 12,165 11,007 8,768 Habach Kees 3,567 3,289 2, ENVEO Sentinel-2 Glacier outlines 2015 Austrian Glacier Inventory 1998

14 Glacier mapping with Landsat 8 OLI & Sentinel 2 MSI OLI red/swir (30 m) OLI pan/swir (15 m) MSI red/swir (10 m) Sentinel 2: Copernicus 2015 Paul et al. (2016) Paul

15 Ice Velocity maps Karakoram,TSX/TDX

16

17 Covered Glacier area, and fraction of valid IV estimates 1 Mountain Glaciers Ice velocity retrieval works best with SAR Data with < ~ 5 m resolution 2 Glacier Dates Coverage (GLIMS) 1. Yinsugaiti 9/9/12-25/11/ % 2. Musita 9/9/12-25/11/ % 3. Qiaogeli 9/9/12-25/11/ % 4. Baltoro 9/9/12-25/11/ % 3 4

18 Svalbard Ice Cap Monitoring by Sentinel-1 Time Series of Ice velocity Of major outlet glaciers

19 Time series of Ice velocity - KRONEBREEN GLACIER

20 Ice Velocity Maps for Novaya Zemlya PALSAR Semzinel

21 Elevation change: DEM Differencing

22 Geodetic mass balance of Patagonian Icefields Wael Abdel Jaber & Dana Floricioiu (DLR-IMF) Investigation area: Northern & Southern Patagonian icefields Data: SRTM C-band DEM TanDEM-X Raw DEMs Elevation change rate maps SPI: 2011/ NPI: Mass balance of the icefields Region Area [km 2 ] NPI SPI NPI+SPI * Ice equivalent values

23 Data Needs

24 Data Needs