220 likes | 354 Views
Decadal modulation of the ENSO impact on the Hydroclimate over the United States . Kingtse Mo Climate Prediction Center NCEP/NWS/NOAA. Composites of drought indeces. All drought indices indicate Cold ENSO=> Drought over the Great Plains, Southwest and Gulf states. Questions.
E N D
Decadal modulation of the ENSO impact on the Hydroclimate over the United States Kingtse Mo Climate Prediction Center NCEP/NWS/NOAA
Composites of drought indeces All drought indices indicate Cold ENSO=> Drought overthe Great Plains, Southwest and Gulf states
Questions The impact of ENSO does not always obey the ‘composites’. WHY? • Each ENSO is different • ENSO impact on regional climate is modulated by decadal modes
Decadal modes • Trends • Atlantic Multi decadal oscillation (AMO) • Pacific Decadal Oscillation (PDO) • PDO Barnett and Cayan (1998), Gershunov and Barnett (1998) Decadal: fluctuations with time scales longer than 6 yrs
Observational Data100 yr data set is not long enough P and Tsurf • Monthly Precipitation (P) and surface temperature (T) over the United States from 1915-2006 (U. of Washington) • SST and SLP • SST- Smith reconstructed SSTs 1915-2006 • SLP – Trenberth and Paolino (1987) • Anomaly- departure from monthly mean Climatology from 1915-2006
Atlantic Multi decadal Oscillation AMO composite warm-cold phase Decadal influence is small. You can filter to get clearer signal, but the percentage of variance is small, so the net influence is small. AMO warm 1930-1959,1992-2006 AMO cold 1915-1925,1965-1990
AMO influence through ENSO Composite of SPI6 with all seasons together for different phase of the AMO Mo et al. 2009
Trends • Composites were formed for each season for (a) 1915-1960 and (b) 1962-2006 • ENSO – Nino 3.4 index over 0.8 std (less -0.8 std) • Results are given as composites: warm –cold weighted by the number of events • Statistical significance is determined by the Monte Carlo method
Surface temperature ENSO composites warm-cold Tsurf: Strongest signal is in winter and Spring In the recent decades 1. Cooling over the Southeast, the Gulf States has been weakening . 2. Warming over California and cooling over the Plains occur in Spring
It used to be….. • Warm ENSO cold over the South and warming over the North. • Things are changing • Cooling over the South has been weakening in winter. • For California and the Southwest, spring is warmer in the recent decades. • The great Plains is cooler in spring ENSO-T
ENSO-P • Strongest impact is in winter • In the recent decades, influence on California increases • Over the Pacific Northwest, impact comes in Fall
Changing in SSTAs Warming centered at the eastern Pac Warming centered At the Central Pac Ref: Yeh et al. (2009) , Kug et al. (2009), Kao and Yu (2009), Ashok et al. (2007)
ENSO teleconnections are changing • The impact of ENSO on P and Tsurf over the United States is not stationary. • More warm ENSO events in the recent decades • Warming over the Northern U. S. and cooling over the Southeast and the Gulf states are weakening • Impact over the Southwest and California intensified in the recent decades (Warm=> wet, Cold=> drought) • Impact over the Pacific Northwest comes early in Fall. (Warm= > drought, cold=> wet) With such small sample, are these real???
Two kinds of ENSO Eastern PAC X Central Pac x Yeh et al.2009(nature)
Composites of EP & CP ENSO Events were taken from Yeh et al. (2009) P Tair C Pac 7 events • Different type of ENSO contributes to the variability of impact • If models are able to capture two types of ENSO and impact, then they can be used for diagnosing phycical mechanisms E Pac 21 events
IPCC runs (focus on winter) • 20th Century simulations from 1880 to 2000 runs (20C3m) • Models: • GFDL_CM2.1 (700yrs) • GFDL_CM2.0 (412 yrs) • MicroC3.2-medres (453 yrs) • UKMO_HadCM3 (280 yrs) • MPI_Echam5 (814 yrs) • These are simulations with realistic ENSO (Meehl et al. 2007, Oldenborgh et al 2005) Data from the Earth system grid WCRP CMIP3
IPCC (JFM) • For each run, compute anomalies as the departure from the model mean • Form SSTA Nino 34 and Nino 3 indices • Select events when both indices are greater than 0.6-0.8 standard deviations • EP events when Nino34> Nino3 + d with a clear Maximum over the eastern Pacific, d= 0.1 to 0.3 • 5. CP events when Nino3>Nino34 +d with a clear maximum over the central Pacific
Ensemble ipcc model Central Pac Eastern Pac • Tsurf : more north-south warming/ cooling for the E-PAC and East-west contrast for the C-PAC • Cooling over the South weakens for the C –PAC • More rainfall over California and the Southwest and less for the Southeast for the C-PAC Eastern Pac Central Pac
Physical mechanims were examined using the GFDL_CM2.1 model runs • Physical Mechanisms • Eastern Pac: • Positive P anomalies are located over the area extending from 160W to 90W. • Walker circulation puts less rainfall over northern Brazil • The weakening of Hadley circulation implies more rainfall over the Southeast. • Central Pac: • Positive P anomalies are located over the deadline Precip (shaded), Chi wind (vectors)
GFDL CM2-1 • E-PAC • the North American jet extends to the Atlantic • Strm 200 –more zonal structure • Central Pac • jet extends to the west coast • Pacific-North American wavetrain
Decadal changes of ENSO impact • For forecast and monitoring: • Impact on the U. S. depends on both ENSO pattern as well as magnitudes. • Pay attention to the precip anomalies in the Tropics. • Different tropical convection pattern different wind and circulation responses different impact • These changes are not linear. The OCN magic may not work