Choosing ensemble graphics and dataaccess functionality for the FIM realtime ensemble demo this hurr

1
Choosing ensemble graphics and data-access
functionalityfor the FIM real-time ensemble demo
this hurricane season.

• Tom Hamill
• [phone]
• tom.hamill_at_noaa.gov

2
Inputs

• FIM G8 (30 km) quasi-realtime global ensemble
  output, initialized from EnKF.
• produced 2x daily? possibly 4x.
• 20 members?
• output interpolated to pressure levels, lat/lon
  grid.
• GFS real-time ensemble forecasts
• Other centres data?

3
Some expected data processingof FIM ensemble

• Apply automated tracking algorithm of vortex
  (which specifically?).
• Attach metadata to tracked vortex (time, ensemble
  member, central pressure, etc.)
• Convert most important FIM ensemble fields (e.g.,
  that stored for TIGGE, hurricane-relevant data)
  to GRIB-2 (and XML) format
• Make current and past forecast data and
  observed/analyzed data available on public server
  
• for others who want to add FIM output to their
  web pages.
• for verification.
• Possibly some simple post-processing, e.g., to
  make a cone of uncertainty.

4
FIM ensemble web pagefunctionalities

• Perceived requirements
• Base capability functional by 1 August 2008
• Ability to display current and recent (last x
  days?) FIM ensemble forecasts.
• Password protection for web page access, but
  password given out readily to interested
  participants.
• Displays of a fairly rudimentary base set of
  hurricane-related products (what compromises base
  set TBD a few slides later).
• Convenient user interface that makes basic
  products easy to see, allows sophisticated users
  ability to dig in and get more information.
• Readily extensible to other areas, e.g,
  extratropics if later west-coast landfalling
  cyclones of interest.

5
FIM ensemble web pagefunctionalities

• Second-tier functionalities (do if there is
  time)
• Separate displays for different areas (wide view
  of Atlantic, Gulf, Caribbean, east Pac, west
  Pac).
• Click on a member of a stamp map or spaghetti
  plot to bring up that members data, plot time
  series.

6
Spaghetti of ensemble track forecastshttp//euler
.atmos.colostate.edu/vigh/guidance/northatlantic/
track_early1.png
I like this sort of graphic, but can we tweak
this to provide other information, e.g. size or
shape of dot intensity? Different colors for
different models (GFS, FIM, etc.).
7
STAMP MAPS -- in this case, for Rita. Add
looping over time, simulated reflectivity, other
information?
observed position
8
Stamp maps might include past track, area of high
winds, like this.http//www.nhc.noaa.gov/gifs/sur
facewindfield_example1.gif
9
Derived products?
I strongly suspect that in year 1, we just wont
be able to make these sort of products, or at
least to calibrate them.
10
Another derived product
11
Bad and good graphicshttp//www.wmo.ch/pages/prog
/amp/pwsp/documents/TD-1422.pdf
Choice of colors on left highlights the data
where there are insignificant differences from
climatology