Richard L. Armstrong
University of Colorado
Boulder, Colorado
USA
Global Snow Cover ECV
Status and Progress in Observations
16th Session of the GCOS/WCRP Terrestrial
Observation Panel for Climate (TOPC-16)
10-11 March 2014, Ispra, Italy
Outline
•
•
•
•
•
•
•
•
•
Why a snow cover ECV?
Users and requirements
Previous ECV documents
Review of currently available products
Some user statistics
Enhanced data set activities
Innovative data sources
Snow cover trends
Community activities/efforts
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
Snow Cover ECV – Why?
Relevance and Benefits:
Recognized as a fundamental indicator of climate
variability and change. Highly sensitive to changes in
temperature and precipitation regimes.
Affects albedo, soil moisture, permafrost state, growth
conditions for vegetation.
Required for snow melt runoff forecasts and flood
prediction.
Approximately 1/6 world’s population relies on snow melt
runoff for water supply. (As much as 75-80 percent of
water in the western US results from snowmelt.)
Assessment and improvement of climate model
performance.
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
Snow Cover ECV
The ECV “snow cover” may include four
categories; snow covered area (SCA), snow
water equivalent (SWE), snow depth, and snow
wetness (presence of liquid water in the snow
cover).
There is a remote sensing capability to
determine, with varying degrees of accuracy, all
four of these parameters. The primary
monitoring product with the longest time series
is a continuous data record of
hemispheric/global snow covered area.
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
Users and Requirements –
Snow covered area (SCA) – (binary, percent)
- presence or absence of snow
- surface reflectance
- trend analysis/climate variability
- cal/val NWP and GCM products
Snow water equivalent (SWE) - (mm)
- hydrology
- frozen soil/permafrost
- trend analysis/climate variability
Time period and spatial resolution ?
- Typically first question asked by potential user
- both being continually enhanced
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
2009 GTOS ECV T5 Snow Cover
“Assessment
of the status of the development of the
standards for the Terrestrial Essential Climate
Variables”
• Definition and Units of Measure
• Existing Measurement Methods, Protocols and
Standards
 In situ and satellite
• Contributing Networks and Agencies
• Available Data and Products
• Recommendations
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
2011 GCOS -154 Systematic Observation Requirements for
Satellite Data Products for Climate – 2011 Update
The primary monitoring product is a continuous data
record of global snow areal extent (SCA).
Also desirable to have supplemental global information
of three additional properties: snow depth, snow water
equivalent, and the presence of water in the liquid phase.
Passive microwave sensors provide reasonable
estimates of snow depth, water equivalent and wetness
in many areas.
Combined or integrated products linking two or more
snow products have been generated by research groups
e.g. ESA Globsnow.
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
Satellite Remote Sensing of Snow
Hemispheric to Global Scale
SCA: Visible = GOES, AVHRR, MODIS, ASTER
Higher resolution (30 to 500 m)
Clouds and darkness obscure surface
Limited to surface characteristics
SWE: Passive Microwave = SMMR, SSM/I,
AMSR-E
Lower resolution (~ 10-25 km)
All weather & day/night
Sub-surface characteristics (mass estimate, soil
state ) -- for dry snow only
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado
“Primary Monitoring Product” Continuous Record of SCA
NOAA visible climate data record
1973
Nov
1966
Oct
1972
May
1975
ESSA, NOAA, GOES
Series
2007
1998
1990s
May 1999
1980-81
Weekly
190 km
digitized
METEOSAT
&
GMS added
Reanalysis of 1966-71
Feb 1997
Moving back in time decreases spatial resolution, spatial
IMS 24 km
coverage, temporal resolution, and retrieval/algorithm sophistication.
Feb 2004
IMS 4 km
Slide courtesy of Dave Robinson/Rutgers Univ.
Satellite-derived Data Sets at NSIDC - pre-NASA EOS 2000
NSIDC Northern Hemisphere EASE-Grid Weekly Snow Cover and Sea Ice
Extent Product, beginning October 1966 - derived from weekly NOAA
(AVHRR optical sensor) snow maps, 25 km. Evolved to NOAA IMS
(Interactive Multisensor Snow and Ice Mapping System) in 1997, daily data
at 24 km, and 4 km data starting in 2004. http://nsidc.org/data/nsidc0046.html
NSIDC Monthly Snow Water Equivalent Climatology Product. (25 km
EASE-Grid) beginning November 1978 . Snow water equivalent derived
from SMMR and SSM/I passive microwave sensors.
http://nsidc.org/data/nsidc-0271.html (no longer being updated)
NASA EOS Global Snow Cover Products at NSIDC
Visible/Infrared - Snow Covered Area (SCA)
MODIS (Moderate Resolution Imaging Spectroradiometer) 500 m and
0.05 deg (CMG) resolution, daily and 8-day, derived from two EOS
visible/infrared instruments: Terra satellite beginning December
1999, and Aqua, May 2002. http://nsidc.org/data/modis/index.html
Passive Microwave – Snow Water Equivalent
AMSR-E (Advanced Microwave Scanning Radiometer)
25 km resolution, daily, 5-day, monthly, Aqua, May 2002 to Oct. 2011.
http://nsidc.org/data/amsre/index.html
Near-Real-Time SSM/I-SSMIS EASE-Grid Daily Global Ice
Concentration and Snow Extent (NISE) 25 km resolution, updated
daily. http://nsidc.org/data/nise1
MODIS Snow Cover Products
• Number of users between 2010 and 2013:
10,174 (12 products)
• Number of countries acquiring data
between 2010 and 2012: 70
• Product summary: 12 snow cover
products at daily, 8-day, and monthly
temporal resolution. Temporal coverage is
February 2000 to present.
Ongoing and future activities
Various efforts are underway to develop new satellite
data products to extend length of records, increase
spatial resolution, and frequency of observation.
For example: as recommended by 2011 GCOS Sat. Supp.
* Exploit capability of the full resolution (1 km) AVHRR
data to extend data back to 1985.
* Take advantage of extensive oversampling of multisensor footprints to enhance gridding resolution for
passive microwave data.
* Provide percent snow cover rather than only binary.
* Integrate data types.
A snow cover climatology for the European Alps
derived from AVHRR data (1985-2011) Hüsler, F. et
al. 2014. The Cryosphere, 8, 73-90
➔ Data archive: AVHRR 1-km, daily,
back to 1985, pan-European
coverage
➔ Snow detection algorithm
applicable to any kind of AVHRR
generation
➔ Extensively validated and tested
for long-term consistency (intersensor)
➔ Overall accuracy around 90%
POD
➔ Application-dependent
parameters and products can be
delivered (for climatology,
phenology, hydrology, etc.)
14
Passive Microwave Earth System Data Record (ESDR)
(a NASA MEaSUREs project)
Objective:
Produce an improved, enhancedresolution, gridded passive
microwave ESDR for monitoring
cryospheric and hydrologic time
series
Before
•
Full record will include SMMR, all
SSM/I-SSMIS and AMSR-E
• Use newly recalibrated L1B SSM/ISSMIS FCDRs
• Enhance gridded resolution to as much
as ~1 km (channel-specific)
• EASE-Grid
2.0 takes advantage of extensive oversampling in
Image
reconstruction
sensor footprints to enhance gridding resolution. Resolution
enhancement using these techniques depends on frequency. Highestfrequency (~85 GHz) data resolution may be enhanced to ~ 1-3 km."
Investigators:
M. J. Brodzik, NSIDC
D. G. Long, BYU
R. L. Armstrong, NSIDC
http://nsidc.org/pmesdr
After
MODSCAG – MODIS snow covered-area and grain size,
percent w/in 500 m grid cell --- T. Painter, JPL
Provides Fractional Snow Cover
MODSCAG is a multiple endmember spectral mixture analysis model combined with
radiative transfer directional reflectance: spectrally unmixes allowing number of
endmembers and the endmembers themselves to vary on pixel by pixel basis.
Combined or integrated products linking two or more
data sources.
Matias Takala,et al. 2011, Estimating northern hemisphere snow
water equivalent for climate research through assimilation of
space-borne radiometer data and ground-based measurements,
Remote Sensing of Environment, Volume 115, Issue 12, 15
December 2011, Pages 3517–3529
ESA Globsnow - www.globsnow.info/ .
Combined or integrated products linking two or more
data sources, contd.
CMC Global Snow Depth Analysis:
Daily from March 12, 1998; Resolution: 1/3° Uses all available
real time snow depth observations from synops, metars
(meteorological aviation reports) and sa’s (special aviation
reports) on the WMO information system. Updated every 6 hours
using the method of optimum interpolation with an initial guess
field provided by a simple snow accumulation and melt model.
The dataset includes monthly snow depth and SWE, the latter
based on snow density estimates. Data are provided on a 24-km
polar stereographic grid covering the NH.
(http://nsidc.org/data/nsidc-0447)
Snow Cover Extent Anomalies
produced by D. Robinson
Seasonal Snow Anomalies
iSWGR
NASA (international) Snow Working Group for
Remote Sensing
Dr. Matthew Sturm et al. 2013
Snow from Air & Space:
Building a Vision
Mission: To engage U.S. and international
science communities in building a vision for
future snow remote sensing efforts, including
but not limited to, future NASA missions, to
promote international collaboration, and
engage in capacity-building within the U.S. and
international
snow
remote
sensing
communities. Encourage snow & snow remote
sensing research.
MODIS Albedo Product
Global Albedo is a MODIS Standard Data Product that began
routine production in 2000.
Products are operationally produced operationally every 16 days
at a 1km spatial resolution and archived as equal area tiles in a
sinusoidal projection (HDF-EOS format) at the EROS Data Center
(EDC) DAAC. (currently available daily by special request)
Products include: Global Albedo (MOD43B3),
BRDF Model Parameters (MOD43B1),
Nadir-BRDF Ajusted Reflectance (NBAR) (MOD43B4)
http://www-modis.bu.edu/brdf/product.html
Snow Cover Product Validation
- SCA -- Common approach is to validate a
given spatial resolution against a higher spatial
resolution satellite product, typically assuming
the higher resolution to be “truth” often without
objective evidence.
-SWE – validation data are in situ point and
transect manual measurements.
-Problems associated with validating areaintegrated satellite data against sparse point
data will always exist (sub-grid heterogeneity,
especially in mountain regions). Recent
applications of downscaled reanalysis data.
TOPC-16, 10-11 March 2014, Ispra, Italy – Richard Armstrong, University of Colorado