Transcript UWG Science Presentation Derksen

Access to NSIDC Data Across the Climate
Research Division, Environment and
Climate Change Canada
NSIDC DAAC UWG Meeting
August 9-10
Boulder, CO
Chris Derksen
SMAP freeze/thaw product
development and validation
Research Scientist
Climate Research Division
Environment and Climate Change Canada
Snow cover/climate analysis:
-annual assessments (i.e. Arctic Report Card)
-Co-I of CMIP6 land MIP
Radar remote sensing of snow:
-on land (new mission concept studies)
-on sea ice (Operation IceBridge)
Environment and Climate Change
Canada: Climate Processes Section
Science and Technology Branch
Atmospheric S & T
Water S & T
Wildlife/Landscape
S&T
Science
Integration
Air Quality Research Division
Meteorological Research Division
Climate Research Division
Climate Processes
Climate Monitoring & Analysis
Climate Chemistry
Climate Modelling & Analysis
[email protected]; [email protected]; Chris [email protected]
Science Question:
• How much snow is there in the NH, how does it vary from year-to-year, and how is it changing?
Main data sets:
• IMS-24km, IMS-4km
• MEASURES snow product
• GlobSnow and other SWE products
• Daily brightness temperatures
(melt onset; snow algorithm development)
• MODIS snow cover products (daily Climate Modeling Grid)
Data latency:
• daily for NH snow monitoring
• monthly for assessments such as BAMS State of the Climate
Data delivery:
• ftp pull initiated by user (works fine)
Issues:
• undetected missing daily IMS files (i.e. not documented). Can NSIDC automatically flag when an ftp failed
from NIC and generate a request for a replacement file?
• caveats/issues with data products are not clearly indicated and tend to be buried in the documentation
or not mentioned by the dataset PI.
• provide NetCDF version of MODIS snow cover products. The current format is EOS-HDF, often need to be
converted to NetCDF formats first, especially for climate model evaluation studies.
• For long time series data, data push tools (put data order on users’ ftp site) would be helpful.
Science Question:
• Validation of Operation IceBridge snow depth on sea ice products
Main data sets:
• DMS L1B Geolocated and Orthorectified images
• ATM L1B Elevation and Return Strength
• POS/AV L1B Corrected Position and Attitude Data
• Sea Ice Freeboard, Snow Depth, and Thickness
Data latency:
• Pull whenever updated
Data delivery:
• ftp pull initiated by user (works fine)
Issues:
• At the time of data access, there were no
geo-tools to subset the data into manageable
volumes. Had to download over 50 gb of data
and sort through it locally to get at what we wanted.
• Variety of file types and formats.
King et al. (2015)
[email protected]
[email protected]
Science Question:
• Evaluation of CMIP5 predictions of future
Arctic sea ice extent
Main data sets:
• Sea ice concentration (NT2/Boostrap)
Data latency:
• Daily and Monthly
Data delivery:
• ftp pull initiated by user (works fine)
Issues:
• Map projections require re-gridding for
CMIP-5 analysis
Laliberté et al. (2016)
[email protected]
Science Question:
• Evaluation of seasonal prediction skill for different sea-ice attributes in
ECCC’s seasonal forecasting system CanSIPS and evaluation of
operational forecasts
Main data sets:
• NSIDC MEASURES & Markus et al (2009) for melt onset
• Sea ice concentration (NT2; Bootstrap)
Data latency:
• daily for operational forecasts
• not an issue for historical forecast analysis
Data delivery:
• ftp pull initiated by user (works fine)
Issues:
• Climate modeling centre user group: limited background with respect
to satellite products; search, download, and use with little
consideration of product uncertainties
Sigmond et al. (2013)
[email protected]
Science Question:
• Determination of daily regional high resolution sea ice motion for operational forecast verification
Main data sets:
• Sea ice age (Tschudi et al.)
• AVHRR APP-x, (NSIDC via U of Wisconsin-Madison)
• Concentration (NRT NASA Team2)
• Other datasets (SLP, SST) are sourced from NOAA (i.e. NCEP/NCAR reanalysis)
Data latency:
• daily for operational forecast verification
• N/A for historical analysis
Data delivery:
• ftp pull initiated by user (works fine)
Issues:
• Subsetting is kind of a pain, since we have to
download and subset locally. Not that big
of a deal, but would be nice to have
DAAC-side subsetting.
• Map projections: move towards a single common
standard across all NSIDC-hosted datasets,
EASE-Grid 2.0 or Polar Stereographic?
Science Question:
• Validation of SMAP landscape freeze/thaw products
Main data sets:
• L3_FT_A
• Cal/val data (protected access)
• MEASURES freeze/thaw (Kimball)
Data latency:
• Weekly-monthly
Data delivery:
• ftp pull from JPL
Issues:
[email protected]
General Comments
• Very positive feedback on services provided by NSIDC DAAC, from the
perspective of both data providers and data users: very professional
interactions in all regards
• Most common issues pertain to user specific projection, subset needs, and
communication of data quality/uncertainty information
• Most issues/recommendations fall under the category of ‘nice to have’.
• The ECCC CRD user community wants access to data, DAAC-side data
preparation tools, efficient data transfer, to facilitate data analysis on our
end