SUMMARY OF FORECASTS

For ENSO Condition:

Dynamical methods: The Scripps/MPI hybrid coupled model predicts warm conditions for winter 1996-97. The standard and the new Lamont Doherty Earth Observatory models predict continuing cold conditions through winter 1996-97. The NCEP coupled model calls for continuing somewhat cool SSTA in east-central Pacific from now to boreal winter 1996-97, but with some warming near dateline by then. The Australian BMRC low order coupled model forecasts current cool NiZo 3 SST to warm in second half of >96, with mild warm event winter 1996-97, lasting into spring. The Oxford coupled model calls for cool but moderating SST between now and winter 1996-97. The COLA coupled model predicts a warm episode for winter 1996-97, lasting into 1997.

Statistical methods: CDC=s inverse modeling predicts continuation or even strengthening of below normal east-central Pacific SST for summer '96, followed by slow return to near-normal conditions by winter 1996-97. The Australian BMRC non­linear analogue predicts somewhat above normal SOI to August 1996. UCLA=s SSA­MEM predicts NiZo 3 SST slowly increasing to normal by early '97, as SOI decreases from high-normal to normal. JPL/Utah State=s CSSA/MARS Aanalog@ system predicts a decrease to normal SOI conditions by 1997, negative values 1998. Univ. of British Columbia=s neural net model predicts a coolish-normal NiZo 3 for winter >96-=97. The Chinese Academy of Sciences/NCEP EOF-iteration model calls for slowly neutralizing SST through winter >96-97 with NiZo 4 above normal by then. NCEP/CPC's CCA predicts slightly below normal NiZo 3.4 (120^E­170EW) SST neutralizing by early winter and becoming warm by late winter '97. NOAA's constructed analog predicts dissipation of cool NiZo 3.4 SST by fall 1996, warming slightly for winter 1996-97 and especially spring. NCEP=s 4-model consolidated forecast projects slightly cool SST through 1996, switching to slightly warm by boreal spring 1997.

For United States/Canadian and England Temperature:

Dynamical methods: The NCEP coupled model forecasts warmth in lee of northern U.S. Rockies Jul-Aug-Sep '96. The GISS SI95 climate model forecasts positive anomalies in southwest U.S. summer >96 (low to moderate confidence).

Statistical methods: The UK Met. Office predicts somewhat cool for central England Jul-Aug >96. NCEP/CPC's optimal climate normals scheme predicts warm Southwest and much of eastern half of U.S. Jul-Aug-Sep. NCEP/CPC's CCA predicts positive anomalies much of eastern U.S., Texas, the Southwest, California, Hawaii Jul-Aug-Sep. NCEP/CPC=s screening multiple regression predicts above normal south Texas, Lake Superior; cool South Carolina and Southwest in Jul-Aug-Sep. The CCA of the Climate Research Branch in Downsview, Ontario predicts cool in arctic Canada and southeastern Ontario, warm most of western and south-central Canada Jul-Aug-Sep >96.

For United States/Canadian and Chilean Precipitation:

Dynamical methods: The GISS SI95 climate model predicts rainfall deficits in southwest U.S. summer >96 (low confidence). The NCEP coupled model does not forecast any coherent patterns in U.S. Jul-Aug-Sep. >96.

Statistical methods: NCEP/CPC=s optimal climate normals and NCEP/CPC's CCA predicts no coherent mainland U.S. patterns Jul-Aug-Sep, but CCA predicts wet southern Hawaii. NCEP/CPC=s screening multiple regression predicts dryness in Oregon, wet New Mexico Jul-Aug-Sep. The CCA of the Climate Research Branch in Downsview, Ontario calls for wet in much of arctic Canada for Jul-Aug-Sep. The Univ. of Chile Dept. of Geophysics predicts normal to dry conditions for central Chile for Jul-Aug >96.

For Tropical/Subtropical Rainfall:

Dynamical methods: The GISS SI95 climate model calls for wet in northeast Brazil summer >96.

Statistical methods: The Univ. of Wisconsin neural net calls for near-average rainfall for coastal East Africa for boreal autumn >96. The UK Met. Office multiple regression and discriminant analysis forecasts indicate dry conditions for tropical North Africa (except wet Guinea Coast) for Jul-Aug-Sep >96 (w.r.t. 1951-80 mean). The NCEP/African Desk=s CCA predicts normal to slightly wet in Sahel Jul-Aug-Sep >96 (w.r.t. 1955-94 mean). NCEP/CPC's CCA predicts weak anomalies at most tropical Pacific islands through boreal winter 1996-97, with dry tendency spring >97 in U.S.-affiliated islands sensitive to warm ENSO conditions.

For 1996 Tropical Storm Activity:

Statistical methods: The CSU LAD regression approach forecasts approx. near normal Atlantic >96 tropical storm activity. The FSU Poisson and tropical/baroclinic approach (see December issue) predicts approx. average >96 season for Atlantic hurricanes, somewhat below average for intense hurricanes. FSU=s Bayesian probabilistic model (see March issue) indicates 1.25 (~60% of normal) intense Atlantic hurricanes for 1996.