EO based glacier monitoring THEMES 1. WGMS & GLIMS within GTN G: strategic set up 2. GlobGlacier & Glaciers_cci: EO based products 3. LDCM & Sentinel 2: future monitoring perspectives Frank Paul* Department of Geography, University of Zurich * with contributions from M. Zemp (WGMS) and A. Kääb (Oslo)
History of the international glacier monitoring Internationally coordinated glacier observation initiated in 1894 by the International Glacier Commission GTN G Steering Committee Combination of former ICSI services (PSFG, TTS/WGI) into the WGMS in 1986 Advisory Board Executive Board Since then WGMS continues to collect and publish standardised worldwide glacier data Glacier changes with time (glacier fluctuations) Spatial distribution of perennial surface ice (glacier inventories) Global Land Ice Measurements from Space World Glacier Monitoring Service US National Snow and Ice Data Center Today, the international glacier monitoring is coordinated by WGMS, NSIDC and GLIMS within the Global Terrestrial Network for Glaciers (GTN G)
GTN G: The structure behind it
GTN G: set up and goals In cooperation with NSIDC and GLIMS, the WGMS runs the Global Terrestrial Network for Glaciers (GTN G) as part of GTOS/GCOS for the UNFCCC. This network follows a multi level monitoring strategy, integrating in situ measurements, remote sensing and numerical modelling.
Observing strategy main goals of long term observations: process understanding model validation change detection impact assessments change detection: Haeberli et al. (2000) rate of change acceleration trends pre industrial variability change patterns integrated / tiered observing strategy Tier 1: multi component obs. system across environmental gradients Tier 2: process understanding and model calibration => extensive energy/mass balance, flow Tier 3: regional indicators => mass change (index stakes, photogrammetry, LIDAR) Tier 4: regional representativeness => cumulative length change of selected glaciers Tier 5: global coverage => inventories (remote sensing/geoinformatics)
Gobal and regional glacier mass budgets WGMS (2008)
WGMS (2008) Global and regional front variations
Global glacier distribution (WGI)
WGMS data products
One stop portal on www.gtn g.org FoG WGMS WGI WGMS / NSIDC GLIMS NSIDC glacierphoto NSDIC
ECVs & satellite contributions GCOS107 Radic & Hock Accuracy 3%, Resolution: 30 m, Precision: 0.01 km 2, Sensors: optical (Landsat) Raper & Braithwaite IPCC WGI vs. DCW
EO based products for glacier monitoring Area Glacier outlines (vector) from ratio image & manual editing Drainage divides & topographic parameters from DEM fusion Archived Landsat TM/ETM+ data with USGS L1T correction Elevation change Repeat altimetry in mostly Arctic regions for trend detection DEM differencing (geodetic mass balance) for entire glaciers freely available (SRTM, GDEM, USGS) and national DEMs Snow covered area threshold on TOA reflectance (requires topogr. correction) proxy for AAR and calibration of mass balance models Velocity fields InSAR, offset tracking (microw.), image matching (optical) averaging over weeks/years, requires good contrast glaciers in key regions, huge archives (PALSAR, ETM+ Pan)
AREA: Take a satellite image (Landsat)... scene from the USGS archive (South Georgia)
... and convert it to outlines outlines (yellow) from Landsat ETM+ after editing
(this works also for tiny glaciers in the Alps) Mosaic of 10 Landsat TM scenes acquired in 2003
Change assessment: area and length
Change assessment: area and length 1997 2006 most important: accurate orthorectification
SNOW: take a Landsat image & a DEM Topographically corrected TOA reflectance band 4 Snow covered area (white) after applying a threshold
VOLUME: take DEMs from 2 points in time Zinal Randa Allalin Gorner Findelen Belvedere SRTM (2000) swisstopo (1985)
... and determine overall volume change Silvretta Rhone Pl. Morte Aletsch Gries Basodino Forno Trient Gorner key method to assess the representativeness of the mass balance glaciers for the entire region
ELEVATION CHANGES: trend analysis from repeat altimetry (ICESat, RA 2)
VELOCITY: offset tracking from RADAR sensors optical image (Landsat ETM+)
VELOCITY: offset tracking from optical sensors
Area: precision of manual digitizing
Comparing Landsat with aerial photography automatic with Landsat TM 30 m rock fall manual digitizations
System context for glacier area
Snow reflectance & spectral bands
Swath width G Sentinel 2 Landsat ASTER
Processing levels & products
100 km tiles
Processing & distribution
Most critical issues (key concerns) Orthorectification L1B > L1C: SRTM or similar, SRTM improved Status of discussion and possible solutions L1C 100 km tiles Data granule size, seamless mosaicing tool (easy to use, batch possibility, collaborative ground segments) L1C UTM projection possibel solutions LDCM might be operational before, glovis already used, interest in Sentinel 2?
Othoprojection errors due to DTM errors Image plane H h Vertical DTM error h DTM Terrain R r R / H = r / h Map plane Horizontal distortion r
Getting global: EO Archives & DEMs The longest time series available is from Landsat 5 TM (for free at USGS)
PVASR: RR results (Landsat debris free) 90 m automatic with Landsat TM
PVASR: RR results (reference data) 90 m automatic with Landsat TM manual on Landsat TM