Assessment of the NCEP CFS Hindcast Runs on Predicting the North American Monsoon Lindsey Long and Jae Schemm NOAA/NWS/NCEP/CPC Lindsey.N.Williams@noaa.gov Runs courtesy of Suranjana Saha NOAA/NWS/NCEP/EMC Introduction Focus on Precipitation and Atmospheric Circulation Look at both seasonal and interannual variability Five varying runs allow us to look at the affect of: Lead Time Updates to the CFS model Horizontal Resolution Description of Runs 5 different hindcast runs from 1981-2006 Two 15-member ensemble mean hindcasts from current operational CFS 1-month lead (May) 3-month lead (March) Three single runs with differing horizontal resolutions using an updated CFS T62 T126 T382 Summary of Runs Run CFS Version Resolution Initial # of Runs Condition 1 GFS 2007 / MOM3 T382 May 15th Single 2 GFS 2007 / MOM3 T126 May 15th Single 3 GFS 2007 / MOM3 T62 May 15th Single 4 Operational T62 May 15-Member Ensemble Mean 5 Operational T62 March 15-Member Ensemble Mean Complex Terrain Demonstrates the need for finer horizontal resolution Baja California Gulf of Gulf of Mexico California Sierra Madre Occidental NAME Sub Regions Arizona/New Mexico Region (AZNM) ; (Higgins et al. TIER 1.5 1998) CORE CORE Region AZNM (Gutzler 2004) Tier 1.5 Region (NAMAP2) CFS JJAS Precipitation [cm], 1981-2006 URD (1x1) T382 T126 T62 Oper T62 May Ensm Oper T62 March Ensm CFS JJAS Precipitation Model Minus Obs [cm], 1981-2006 URD (1x1) T382 T126 T62 Oper T62 May Ensm Oper T62 March Ensm Seasonal Cycle a) AZNM of Precipitation CORE Region shows good example of b) CORE increased accuracy with resolution and model upgrade TIER 1.5 shows less improvement with c) TIER 1.5 new model runs URD T382 T126 T62 T62 T62 May March Seasonal Cycle a) AZNM of Precipitation CORE Region shows good example of b) CORE increased accuracy with resolution and model upgrade TIER 1.5 shows less improvement with c) TIER 1.5 new model runs In all regions, T382 run does well capturing monsoon onset (June-July) URD T382 T126 T62 T62 T62 May March CFS Diurnal Cycle of Precipitation JJAS (mm/day) South East Great Plains Mountain AZNM l CORE T62 T126 T382 NARR Interannual a) AZNM Variability b) CORE Anomalous Precipitation for JJAS c) TIER 1.5 URD T382 T126 T62 T62 T62 May March Interannual a) AZNM Variability b) CORE Anomalous Precipitation for JJAS Decadal Trend c) TIER 1.5 present in observations and CFS URD T382 T126 T62 T62 T62 May March Precipitation Correlations to URD a) AZNM Monthly d) AZNM Seasonal b) CORE Monthly e) CORE Seasonal c) TIER 1.5 Monthly f) TIER 1.5 Seasonal T382 T126 T62 T62 T62 May March Atmospheric Circulation Prominent upper-air features in the NAME region: Anticyclonic flow over the region Low-level jet over the Gulf of Mexico Low-level jet over the Gulf of California Atmospheric Circulation (JJA) 300 mb Streamlines, 850 mb Wind Vectors NARR T382 T126 T62 Oper T62 May Ensm Oper T62 March Ensm Atmospheric Circulation (JAS) 300 mb Streamlines, 850 mb Wind Vectors NARR T382 T126 T62 Oper T62 May Ensm Oper T62 March Ensm Gulf of California LLJ, 850 mb Winds Strong Southerly Flow Weak Westerly Flow NARR T382 Conclusions As expected with the complex terrain of the NAMS, increased resolution does improve the precipitation forecast. Correlations show Interannual Variability is improving with resolution Improvement also seen at the Seasonal and Diurnal time scales. Comparisons of the T62 runs highlight the improvement provided by the updated atmospheric GCM model. Change in lead time for Ensemble Mean hindcasts show only a slight improvement in May versus March The general circulation is still a problem area. Missing anticyclonic flow over the monsoon area Lacking the southerly flow associated with the LLJ; therefore, the warm moisture brought into the region by this jet is missing. Thank you! Questions?
Pages to are hidden for
"Assessment of the NCEP CFS Hindcast Runs on Predicting"Please download to view full document