Space Weather Forecasting Joseph M' Kunches Chief, Space Weather Operations NOAA Space Environment C

1
Space Weather ForecastingJoseph M.
KunchesChief, Space Weather OperationsNOAA
Space Environment CenterBoulder,
ColoradoJoseph.Kunches_at_noaa.gov
Living With a Star and the Exploration Initiative
Workshop Washington, DC April 5, 2004
2
Proton Forecasts from SEC

• Produced daily for 1-3 days forward
• Utilize active region information and -- if
  eruptive already occurred -- LASCO Coronagraph,
  GOES SXI, and ACE EPAM data
• Have not improved appreciably as predicting
  eruptive event still elusive
• Produced as event occurs
• Uses GOES, SOHO, and ACE data
• Have improved over the last few years as more
  data are available, as evidenced by excellent
  predictions during Halloween Storms 2003

3
Daily Forecasts

• There is a strong need to know the value of the
  predictions
• Forecast verification is a necessary component in
  ascertaining value
• Daily forecasts are difficult due to the
  additional aspect of needing to predict the solar
  eruptive as well as the proton characteristics
  that may follow

4
Verification Statistics

• Results are for 1-3 day lead time forecasts
  issued daily
• For explanation of terms go to http//sec.noaa.gov
  /forecast_verification.html
• Forecasts show a modest level of skill

5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
Event-Driven Forecasts

• Include start, max, and predicted flux no
  spectrum as yet
• What predictions will Mars mission require?
• Commercial airlines serve as an example in
  defining requirements
• Airlines want 8 hour lead time, a predicted end
  time, and peak flux
• Mars and Lunar scenarios probably will want
  spectral information
• All Clear desirable for EVAs
• Need 360o monitoring to safeguard against missed
  events
• Improved models necessary to account for radial
  and azimuthal separation

10
Halloween Storms 2003

• EXCELLENT!
• 13 Radiation Storm (i.e., SEP) days, Oct. 26
  Nov. 7
• 4 Warnings issued
• 31 Alerts issued
• 195 minute average lead time
• 0 missed warnings
• 0 false alarm rate

11
Summary

• Two types of forecasts, daily and event-driven,
  are produced by SEC
• Event-driven forecasts are the most skillful
• Both types of forecasts rely heavily on GOES,
  SOHO, and ACE observations
• To forecast for a Mars mission, a full suite of
  similar instruments must be flown, spatially
  optimized for full observational capability
• Appropriate models need to be employed to fill in
  both spatial and temporal voids