1 / 21

Michael K. Tippett, IRI Matt Barlow, AER Brad Lyon, IRI

Statistical corrections of Central Southwest Asia winter precipitation simulations (and forecasts!). Michael K. Tippett, IRI Matt Barlow, AER Brad Lyon, IRI. Introduction. CSW Asia drought (1998-2002). Winter (DJFM) precipitation is key. “May Kabul be without gold but not without snow”

manasa
Télécharger la présentation

Michael K. Tippett, IRI Matt Barlow, AER Brad Lyon, IRI

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Statistical corrections of Central Southwest Asia winter precipitation simulations (and forecasts!) Michael K. Tippett, IRI Matt Barlow, AER Brad Lyon, IRI

  2. Introduction • CSW Asia drought (1998-2002). • Winter (DJFM) precipitation is key. • “May Kabul be without gold but not without snow” • Tropical forcing • Recent drought associated with La Nina. • “Perfect Ocean for drought” • GCMs have poor simulation skill. • Is statistical correction possible?

  3. Precipitation related to elevation. Elevation Focus on NE region. Variability Climatology NE

  4. Is there any predictability? Extended GCM simulations forced by observed SSTs show little skill. Observations = CRU (New et al 2002) Period = ~1965 through winter 1997-98

  5. Approach • Identify basis for predictability • Physical mechanisms. • Possible predictors. • Linear regression between model output and observations (MOS). • Use model variables to predict observed anomalies. CCA.

  6. East Asia Jet Stream • Negative correlation between CSW Asia precipitation and EAJS strength. • Positive correlation between EAJS strength and Maritime Continent precipitation. Yang et al. 2002

  7. First EOF of DJFM 200 hPa reanalysis winds Correlation: Wind EOF1/Precip EOF1 = .66 Wind EOF1/NE = .58 First EOF of DJFM precip Model winds (ECHAM4.5) do not have the same variability or relation with observed precipitation.

  8. Precipitation anomaly patterns associated with NE precipitation anomalies.

  9. Physical connections • Positive MC precipitation anomalies. • Upper level convergence • Southerly flow + Coriolis effect = westerly flow. • Strengthened EAJS. • Negative CSW Asia precipitation anomalies. Idea: Use Western Pacific model precip to predict CSW Asia precip.

  10. MOS Details • 4-month DJFM seasonal averages. • Training • GCM simulations (observed SST) ~1965-98. • Western Pacific precipitation. • UEA/CRU New precipitation data set interpolated to T42 grid 1950-1998. • Corrected simulation skill estimate • Leave one year out cross-validation. • Forecasts 1998-2002 • GCM forecasts made in Oct. • Forecast SST (NCEP Pacific + climatology + Stat. Atl.).

  11. Leading CCA modes Western Pacific model precip (simulation) predicting CSW Asia precip.

  12. Leave-one-out cross-validated anomaly correlation of MOS corrections Corrected simulation skill

  13. GCM simulation skill

  14. GCM Forecast Oct 1998 MOS Corrected Forecast DJFM 1998-99 Observations

  15. GCM Forecast Oct 1999 MOS Corrected Forecast DJFM 1999-2000 Observations

  16. GCM Forecast 2000 MOS Corrected Forecast DJFM 2000-2001 Observations

  17. GCM Forecasts Oct 2001 (More models now!) DJFM 2001-2002 Observations

  18. MOS Corrected Forecasts DJFM 2001-2002 Observations

  19. GCM Forecasts Oct 2002 Observations 2001-2002 Reports of snow.

  20. MOS Corrected Forecasts Observations 2001-2002 Reports of snow.

  21. Summary • GCMs show poor simulation skill for CSW Asia winter precipitation. • CSW Asia and western Pacific precipitation are associated via EAJS. • Observed winds are a good predictor. • Model winds are not. • Western Pacific model precip is a usable predictor. • MOS corrections improve simulation skill. • Some promising forecast results (1998-2003).

More Related