NASA / SPoRT Update

1
NASA / SPoRT Update
Coordination Call Friday 4/16/10

• Agenda items
• MODIS/GOES Hybrid
• UAHs CI and CTC data files to NOAA/HWT/EFP
• Lightning Data transfer, Psuedo-GLM code,
  Training
• Lightning Warning Algorithm
• MODIS/AMSR-E SST Composite

2
MODIS-GOES Hybrid
Work done by Gary Jedlovec, Matt Smith, Kevin
Fuell

• Running in real time, but currently working out
  the data ingest and bowtie correction logistics
  for MODIS.
• Need to remap MODIS to 2km IR to more accurately
  portray ABI channels
• Need to examine timing discontinuities at edges
  and possible limb-correction
• Product introduced to WFO Partners in March
  They expressed high interest.

MODIS 1km / GOES 4km IR over SW U.S. at 1845Z (2
MODIS swathes)
3
UAHs CI and CTC data files to NOAA/HWT/EFP
Work by Kevin Fuell, Jon Case, John Walker (UAH),
Chris Siewert, Greg Grosshans

• NASA/SPoRT asked to assist with the transfer of
  AWG Convective Initiation and a Cloud Top Cooling
  product for use during Spring Experiment 2010
• AWG CI product is by University of Alabama
  Huntsville
• SPoRT already sending LMA data in similar fashion
• SPoRT worked w/ UAH to make N-AWIPS ready file
• HWT has tested a file and will receive automated
  script from SPoRT and UAH

GEMPAK via SPoRT
NAWIPS via HWT
4
Real-time KSC Data
Work done by Geoffrey Stano, NASA Lightning Group

• Lightning group has obtained real-time LDAR
  data
• First successful ingest into AWIPS
• Currently at 1 km and 1 min resolution
• Finalize ingest to WFO Melbourne, FL during SPoRT
  visit April 23
• Efforts support Spring
• Experiment

Pseudo-GLM from KSC LDAR II
5
Data Transition
Work done by Geoffrey Stano, Kristin Kuhlman

• Finalizing data feeds to Spring Experiment
• 3 total lightning networks
• KSC, North Alabama, Washington DC
• SPoRT personnel to participate in Experiment

• Methodology for Pseudo GLM has been transferred
• Providing WES case studies for in-active
  lightning days
• May 3 and 6, 2009

6
Lightning Training
Work done by Geoffrey Stano, Eric Bruning

• Train forecasters on the use of the Pseudo GLM
  product
• Intended to be viewed before arriving in Norman

• Followed by in-person training on first day
• Educate forecasters on potential of GLM era
• Provide 2 case examples
• Working to incorporate into the NWS Learning
  Management System

7
WRF-based Lightning Algorithm
Work done by Bill McCaul, Jon Case, Scott Dembeck

• Ongoing work to use model microphysics to
  forecast lightning threat (graupel flux,
  vertically integrated ice)
• NSSL WRF has been configured to output Flash
  Rate Density using hourly max fields. See
• http//www.nssl.noaa.gov/wrf/
• Lightning Threat 1 based solely on graupel
  flux
• Lightning Threat 2 based solely on Vert.
  Integ. Ice
• Lightning Threat 3 combined use of Threats 1
  2
• Lightning Threat fields will be evaluated at
  Experimental Forecast Program (EFP) 2010

• Tie in with the Proving Ground
• Data assimilation from GLM to improve
  microphysics and storm environment
• Better initial characterization of microphysical
  fields and therefore improved convective
  forecasts in the very short term (1-6 hours)
• Request to use Lightning Threat Algorithm in the
  simulated WRF-ABI case event

8
Pseudo ABI Composite SST Product
Work done by Gary Jedlovec, Frank LaFontaine,
Jaclyn Shafer

• Application
• weather forecast community requires continuous
  spatial coverage of SST for model initialization
• marine weather users benefit from high resolution
  SST data
• Cloud cover cause a significant problem!
• MODIS /AMSR-E product relevant to ABI
• Composite fields of high resolution satellite
  derived SSTs provide a good way to provide this
  data
• Compositing fills in cloudy or data void regions
  with observations from the previous day(s)
  satellite passes
• Fixed time replacement preserves diurnal cycle
  captured with foursnapshots from polar orbiting
  data

9
Pseudo ABI Composite SST Product
Problem The regional weather forecast
community requires continuous spatial fields of
surface parameters for model initialization.
Composite fields of high resolution satellite
derived SSTs may provide a good way to provide
this data. However, persistent cloud patterns
can cause product latency and reduced accuracy.

• Previous work developed a SST composite product
  with MODIS data to provide high-resolution SST
  data over limited regions (Haines et al. 2007)
• Impact of MODIS SSTs (versus RTG) on fluxes of
  heat / moisture and subsequent weather forecasts
  was significant in coastal regions (Lacasse et
  al. 2008 Case et al. 2009)
• Regions of high latency reduced accuracy and
  impact

• Just finished a collaborative project with
  PO.DAAC at JPL
• Developed an enhanced SST composite product
  reduced latency
• MODIS and AMSR-E, PO.DAAC L2P data stream,
  latency weighted compositing algorithm
• Demonstrate improvement and impact on forecasts

10
Pseudo ABI Composite SST Product

• Approach overview
• Increased SST data availability transition from
  direct broadcast to more complete data source
• Bring in AMSR-E SST data for coverage in
  persistent cloud regions
• SST compositing algorithm changes latency,
  error, and resolution weighted product
• Use near real-time L2P data stream (JPL) for
  MODIS and AMSR-E passes more passes
• expand product coverage
• pixel by pixel quality estimates and bias
• slight additional delay in
• data access tolerable
• better cloud / rain detection
• AMSR-E data coarse resolution with
• no data near coast

11
Pseudo ABI Composite SST Product

• Approach details
• A collection of MODIS and AMSR-E SST values
  corresponding to the last 7 days is obtained for
  the JPL PO.DAAC for each pixel in a product
  region (at 1 km resolution), for the four Terra /
  Aqua overpass times
• MODIS proximity flags 4 and 5, bias adjustment
• AMSR-E proximity flag 4
• Apply latency weighted compositing scheme to the
  collection at each point in the 1 km resolution
  output file
• where SSTcp (I,j) is a composite SST value
  at a point (I,j),
• SST(I,j,k) is a SST collection (k is an
  index corresponding to date of data in
  collection),
• L(k) is the latency (in days) of a
  particular SST value in the collection, and
• Wt(d,k) is a data weight factor where d
  corresponds to either MODIS (Wt1.0), AMSR-E
  (Wt0.20), or some other value for another
  instrument source.
• The inverse latency formula uses all data in the
  collection and allows more recent data to have a
  greater influence on the composite
• The reduced AMSR-E weight factor (Wt) accounts
  for the large footprint compared to MODIS

12
Pseudo ABI Composite SST Product

• MODIS alone produces a high-resolution (1km) SST
  composite but with some latency issues and gaps
• AMSR-E alone reduces latency in the SST composite
  with coarser resolution data, but not near land
• Enhanced MODIS / AMSR-E SST composite a blend of
  both
• Product available 4x a day corresponding to Terra
  (day and night) and Aqua (day and night)

OSTIA or NESDIS POES/GOES SST product (with a
Wt0.20) used to fill in where neither MODIS nor
AMSR-E coverage is complete (a few coastal areas)
13
Pseudo ABI Composite SST Product
Validation 4 regions, 4 seasons, 4 times a
day 60-70 fixed buoys (coastal and gradient
regions), 90-100 drifting buoys (more open ocean)
per time - bias and rms

• Enhanced (black) much improved over MODIS only
  (red) at all times due to reduced latency
• Drifting buoy (solid lines) biases smaller than
  fixed (dashed)
• Biases generally lt0.20 C, RMS lt0.50 C
• Trend in MODIS only due to latency no trend in
  enhanced SST composite bias

14
Pseudo ABI Composite SST Product
Forecast impact High resolution SSTs have had a
positive impact on a forecast applications
15
Pseudo ABI Composite SST Product - GOES-R PG
Application

• Current product available in WRF EMS and selected
  WFOs
• 1km resolution (2km in WRF EMS), 4 times a day
• Planning to adapt product to GOES ABI resolution
• 1-3 hourly, 2 km resolution
• compare to current POES / GOES once daily
  product (Maturi)
• monitor accuracy
• Eventual dissemination to WFOs, etc.
• Will collaborate with AWG on application and use
• Positive impact on regional weather forecasts

• Real-time data available from NASA / SPoRT
  (GRIB2) ftp//ftp.nsstc.org/outgoing/lafonta/sst
  /grib2/conus/
• Also available from JPL PO.DAAC in GRIB2 and
  netCDF (May 2010)
• SPoRT http//weather.msfc.nasa.gov/sport/modis/
• sst_comparison.html