SMAP L4_C Status

1
An Earth System Data Record for Land Surface
Freeze/Thaw State
John Kimball1, Kyle McDonald2, Youngwook Kim1
John Lucotch1 Joseph Glassy1 1The University
of Montana 2Jet Propulsion Laboratory, CalTech
NASA MEaSUREs/PoDAG meeting, Oct 15, 2010
2
Project Goals and Objectives

• Provide a consistent, long-term global record of
  landscape freeze/thaw (F/T) state dynamics for
  vegetated regions where frozen temperatures are a
  major constraint to ecosystem processes
• Distinguish F/T heterogeneity in accordance with
  mesoscale climate, terrain land cover features
• Quantify F/T linkages to vegetation productivity
  and carbon fluxes
• Distinguish natural variability from climate
  change signal
• Establish baseline conditions for future missions
  (SMAP).

Over half of the global land area is affected by
F/T cycles that constrain land-atmosphere water,
energy carbon fluxes
Freeze-Thaw Affected Regions
Frozen Period (Days/yr)
Source Kim et al. 2010. TGARS
3
Current Results and Output Products
Daily Freeze-Thaw Status SSM/I (37GHz, 25km Res.)
2004

• Daily F/T state maps
• - 4 discrimination levels Frozen (AM PM),
  Thawed (AM PM), Transitional (AM frozen, PM
  thaw), Inverse-Transitional (AM thaw, PM frozen)
• Global domain encompassing F/T affected areas
• - 66 million km2 or 52 of global vegetated
  area)
• Initial 20 year record (Final gt30 years)
• Detailed metadata on product QA/QC
• - Online documentation
• - Quantitative accuracy metrics (daily)
• - Qualitative QA/QC maps (annual)
• Quick-Visuals (thumbnails, animations)
• Software support (HDFView, Panoply)
• 2 data formats (binary, HDF-5)

Apr 10
Jul 19
Dec 26
Source http//freezethaw.ntsg.umt.edu
4
Example FT-ESDR Research Applications
Annual Non-Frozen Period Trend (1988-2007)
Source Kim et al. 2010. TGARS
5
FT-ESDR Development Status

• First ATBD release (Jun-09)
• Available online http//freezethaw.ntsg.umt.edu/
  publications.htm
• Baseline seasonal threshold algorithm single
  Tb data series
• Planned updates for error budget, data
  integration and validation
• Recent FT-ESDR releases (Mar/Oct-10 at NSIDC)
• AMSR-E global land parameter bundle (2002-2008)
• SSM/I based F/T record (1988-2007)
• Two more releases planned (FY11/12)
• 2 online data archives (NSIDC DAAC, NTSG-ESIP)
• Baseline (Tier-1) documentation of FT-ESDR
  accuracy
• Daily comparisons against global WMO air
  temperature records (3700 stations) over
  homogeneous land areas
• Spatial QA/QC using ancillary data (e.g. DEM,
  Land cover heterogeneity)
• Current product characteristics
• - Global EASE-grid projection 25-km Res Daily
  repeat (AM, PM, CO) 20-yr record (1988-07)
  derived from SSM/I 37V GHz Tb series HDF-5
  binary formats.
• Product quality Good (i.e. Validated Stage II
  based on EOS product maturity guidelines)
• - Mean annual F/T classification accuracy gt80
  relative to WMO station observations (Tair).

6
FT-ESDR Algorithm Baseline
Pixel-wise Calibration using Tmx/Tmn from Global
Reanalysis
Seasonal Threshold Approach
Annual Definition of SSM/I (37V GHz) Tb F/T
Reference States
Frozen
Non-Frozen
Source Kim et al. 2010. Developing a global
record of daily landscape freeze/thaw status
using satellite passive microwave remote sensing.
IEEE TGARS, DOI 10.1109/TGRS.2010.2070515.
7
Baseline (Tier-1) FT-ESDR Validation QA/QC
Spatially Explicit Quality Assessment
Global Comparisons with WMO Daily Air
Temperature Observations
Cross-channel F/T classification assessment
Source Kim et al. 2010. TGARS
8
Secondary (Tier-2) FT-ESDR Validation

• Focused studies over intensive sub-regions
• Sub-grid scale terrain land cover
  heterogeneity effects
• F/T sensitivity to individual landscape elements
  (snow, soil, vegetation)
• Leverage planned NASA field campaigns involving
  synergistic measurements (CARVE, SMAP).
• Comparisons with other synergistic datasets
• Atm. CO2 anomalies, Satellite based snow cover
  extent, GPP/NPP, NDVI LST.

FT-ESDR Spring Thaw 1Snow Cover Extent
Anomalies over Northern (gt50ºN) Domain
r 0.64, plt0.001
1Source D.A. Robinson (http//climate.rutgers.edu
/snowcover)
9
Enhanced FT-ESDR Development Validation
Mean annual F/T Classification Accuracy

• Cross-sensor F/T comparisons
• Document F/T differences between overlapping
  sensor records SSM/I, SMMR, AMSR-E, ASCAT,
  SeaWinds, SMOS
• Utilize multiple sensor records for FT-ESDR
  production (e.g. empirical forward process
  models, data assimilation)
• Goal Document exploit synergies among
  multiple satellite records/frequencies/polarizatio
  ns to enhance FT-ESDR information, accuracy and
  QA/QC.

SMMR/SSM/I 37V GHz AMSR-E 36V GHz SeaWinds 13
GHz
2007 frozen period difference between SSM/I and
AMSR-E
10
Climate Change Monitoring of extent,
seasonal-annual variation trends in
frozen/non-frozen period
Human Health climate sensitive infectious disease
distribution and change, vector habitat change
Agriculture Frost status/potential, growing
season and potential yield
Potential FT-ESDR Research Applications
Forests phenology, productivity, carbon
source/sink activity, cold hardiness, vegetation
stress forest ecoregions
Water Resources snow cover status and frozen
soils monitoring, timing of ice breakup and
formation in rivers and lakes
Weather, Natural Hazards Weather forecasting,
runoff and flood risk potential
Species and Habitats Shift in ecological
zones, habitat change
11
Potential FT-ESDR User Community
Primary Thaw Date vs Spring Flood Pulse, Yukon
Basin, AK
Example FT-ESDR Cryosphere Hydrology
Applications
Primary Spring Thaw Date vs Spring River Ice
Breakup on Tanana River AK
12
Potential FT-ESDR User Community
Example Agriculture Application
2Soybean Harvested Acres 2007
2007 Spring Frost Index
Source 1http//www.nass.usda.gov/Statistics_by_Su
bject/index.php 2http//www.agcensus.usda.gov/Publ
ications/2007/Online_Highlights/Ag_Atlas_Maps/Crop
s_and_Plants/index.asp
13
Potential FT-ESDR User Community
Example FT-ESDR Carbon Cycle Ecosystems
Application
14
Potential FT-ESDR User Community

• Diverse User Community
• - Decadal Survey mission teams (SMAP)
• - Cryosphere
• - Ecosystems Carbon Cycle
• - Hydrology
• - Global Change
• - Atmosphere and Climate
• Community ID Interaction
• Literature search
• Publications
• Dear Colleague letters
• - Workshop presentations
• FT-ESDR user metrics
• User feedback
• SMAP SDT/WG activities

15
User Identification Through FT-ESDR 1Metrics
Reporting
1Preliminary results based on NTSG-ESIP
statistics from Jun-Aug, 2010.
16
FT-ESDR User Community Support

• Project website with online data archives,
  documentation metadata
• - Embedded links to NSIDC archives similar or
  synergistic data products
• Publication of methods, data software
• - Peer-review journal publications
• Public data archives (NSIDC DAAC, NTSG-ESIP)
• Data links through existing community data
  portals (NACP)
• Dear Colleague invitations for data download
  feedback
• SMAP SDT involvement
• - Baseline Info. for planned L3_F/T product
  development
• - F/T inputs for other product retrievals
  (L3/4_SM L4_C)
• - Cal/Val Applications WGs
• Development, publication, distribution of
  synergistic data bundles (AMSR-E)
• - FT-ESDR used as a frozen T constraint for
  production of higher-order AMSR-E product
  retrievals Inundated area, Soil Moisture, Air
  Temperature, Water Vapor VOD
• - AMSR-E science team support

17
FT-ESDR Project Summary

• Three planned FT-ESDR data releases
• - Initial data release (FY10 20 yr record
  1988-07) transfer to NSIDC completed
• - Additional F/T data release with AMSR-E land
  product bundle (2002-08)
• - Two more releases planned for FY11/12 (up to 33
  yr record 1979-011)
• Metrics reporting initiated for both NSIDC
  NTSG-ESIP archives
• Product Maturity Good (i.e. validated, Stage 2
  based on EOS product maturity guidelines)
• - Mean annual F/T classification accuracy gt80
  relative to 3700 WMO stations
• Accuracy adequate for broad set of science
  applications climate change, cryosphere,
  hydrology, ecosystems C cycle
• Methods publication (Kim et al. 2010. IEEE
  TGARS)
• Product readiness to support next wave of IPCC
  activity Good
• Work in progress
• Comparison use of multiple sensor records for
  improved information development of longer F/T
  record
• Detailed algorithm error budget development
• Validation (Stage 3) QA/QC

18
BACKUP SLIDES
19
Product Maturity Definitions
Source http//nsidc.org/data/docs/daac/ae_land3_l
3_soil_moisture.gd.html
20
Data for FT-ESDR Production and Validation
21
FT-ESDR User Community Statistics
Example user metrics from NTSG-ESIP archive
Daily Users
Distinct Users by Class
22
Cross Sensor F/T Comparisons
2004 Non-Frozen Season from SSM/I, AMSR-E
SeaWinds
23
Potential FT-ESDR User Community
Example FT-ESDR Ecosystems Application
Annual NPP (MOD17) vs FT-ESDR Spring Thaw Timing
Anomalies
24
Potential FT-ESDR User Community
Example Agriculture Application
2Corn Harvested Acres 2007
2007 Spring Frost Index
Source 1http//www.nass.usda.gov/Statistics_by_Su
bject/index.php 2http//www.agcensus.usda.gov/Publ
ications/2007/Online_Highlights/Ag_Atlas_Maps/Crop
s_and_Plants/index.asp