Slide 1 - Environmental Modeling Center - NOAA

Document Sample
Slide 1 - Environmental Modeling Center -  NOAA Powered By Docstoc
					Precipitation, clouds and fluxes in the CFSRR
               Apr. 1998-Jul. 2000

--compared to ERA40, CMAP, SRB, ISCCP, AVHRR OLR
--ERA40 known to have too much tropical precipitation
    due to problems in correcting satellite biases
        --independent estimates are estimates
     --CFSRR contains stratus parameterization
       --stratus reduced after first few months
        --CFSRR to be tested in GFS parallel
           --clouds, fluxes 0-6hr forecasts
              --cloud liquid water 0-hr
 Global Mean Precipitation
    Aug. 1998-July 2000

         CFSRR ERA40   CMAP

Global   3.18   3.30   2.62
Land     2.41   2.1    1.96

ocean    3.51   3.82   2.9
Upward trend in reanalyses’
Precip over oceans

ERA40 much stronger trend
     (Thanks to HM van den Dool)

ERA40 most rain over tropical oceans   Ocean
  CFSRR most rain in extra-tropics

ERA40—excessive tropical rain could
   decrease rain in extra-tropics

CMAP not so good over mid-latitude
  oceans and for higher latitude
      frozen precipitation
Both reanalyses have more
of double ITCZ than CMAP
  ERA40 too much rain over tropical
    oceans, too little subtropics
CFSRR too little over Indonesia, Africa,
too much mid-latitudes, too much ITCZ
CMAP—trouble with ITCZ, mid-latitude
   oceanic precip, frozen precip?
      CFSRR east Pacific stratus
    Further from coast than ISCCP,
In South Pacific displaced to northwest

ERA40 not as much stratus as CFSRR
 But more clouds in convective areas

  ISCCP problems with polar clouds?
 (ISCCP problem telling snow and ice
    from cloud or model problem??)
Both reanalyses less in stratus regions
(CFSRR displaces stratus in east
Pacific, south Atlantic)

ERA40 more over tropical oceans than

ISCCP more clouds than either in
subtropics, mid-latitudes, tends to
have less than either in higher latitudes
  CFSRR stratus decreases
  After first few months

 Total clouds 25N
April 1998-July 2000
Boundary cloud
Two year mean
Era40 doesn’t distinguish
Between low cloud and
Boundary layer cloud

Low-level cloud
Two year mean
                         CFSRR has three layers of clouds in
                         tropics—900, 700, and 250 hPa

                         From pgbh00 file

                         GFS tends to decrease stratus clouds
                         In first few hours of forecasts

Zonal mean cloud liquid water
        2 year mean
Stratus clouds supposed to be thin

Highest amounts of cloud liquid water in
east Pacific stratus in CFSRR
  CFSRR too much downward short wave over ocean,
      causing too much net heat flux into ocean

More sensible heat, less latent heat over land than ERA40

 Note differences between Kiehl and Trenberth and SRB

         COADS community could not balance
             ocean surface energy budget
    —latent heat parameterization produces too little
         evaporation to balance net short wave
    Global surface energy balance Aug. 1998-July 2000
          cfsrr    Era40    SRB      Kiehl and
lh        83.49    81.93             78          78-90
sh        15.66    15.46             24          16-27
dsw       193.06   176.59   183.46   198
usw       25.93    22.38    22.72    30
nsw       167.12   154.21   160.73   168         142-174
dlw       340.84   344.59   341.89   324
ulw       397.25   397.83   395.41   390
nlw       -56.41   -53.24   -53.52   66          40-72
Net rad   110.71   100.97   107.22   102         99-119
nhf       11.57    3.58
Global land surface energy balance Aug. 1998-July 2000

               cfsrr     Era40      SRB
     lh        40.45     43.71
     sh        33.46     24.20
     dsw       192.60    176.52     184.60
     usw       49.14     42.54      41.29
     nsw       143.45    133.98     143.30
     dlw       304.61    308.11     307.17
     ulw       371.16    372.69     372.30
     nlw       -66.55    -64.58     -65.14
     Net rad   76.91     69.39      78.17
     nhf       3.00      1.48
Global ocean surface energy balance Aug. 1998-July 2000

                   cfsrr     Era40     SRB
         lh        101.74    98.19
         sh        8.13      11.79
         dsw       193.48    176.79    183.22
         usw       16.22     13.89     14.92
         nsw       177.26    162.89    168.3
         dlw       356.33    360.20    356.75
         ulw       408.46    408.64    405.38
         nlw       -52.13    -48.44    -48.63
         Net rad   125.13    114.45    119.67
         nhf       15.26     4.48

            Opposite biases

    SRB tends to have less
    downward short wave
       in polar regions

CFSRR too much downward short
   wave over tropical oceans

  ERA40 too little over tropics
CFSRR East Pacific minima
    displaced to west
    from SRB minima

ERA40 minima much less
            Bias in DSW from SRB

Difference in NHF        Difference in DSW SFC
        CFSRR tends to have too much dsw
              except in stratus regions
      (and higher latitude northern continents)

   ERA40 too little tropics, too much midlatitudes

  Both reanalyses more dsw than SRB south of 45S

         CFSRR may have too thick stratus,
                 too much clouds
          over higher latitude continents

Difference in downward short wave radiation dominant
          difference between CFSRR, ERA40
                 in surface heat budget
               over lower latitude oceans
    CFSRR too much
  downward shortwave
except in stratus regions
 Where there is too little
CFSRR Stratus too thick
      DSW SFC
 Apr. 1998-July 2000
     DSW SFC
Apr. 1998-July 2000

Reanalyses higher albedo   Ocean
    Over snow, ice

   Zonal Mean Albedo

     Two-year mean

Two-year mean
 CFSRR higher albedo
       than either
  ERA40 or SRB over
high latitude continents
    and over Tibet
CFSRR more evaporation
      In trades and
   in higher latitudes,
 less over lower latitude
       than ERA40

CFSRR higher sensible heat
over low latitude continents,
   lower over ocean and
  high-latitude continents
        than ERA40

CFSRR much more heat
  Into tropical ocean
      than ERA40

   Zonal mean
   net heat flux
 CFSRR more heat flux into ocean
      In convective regions,
Less in stratus regions than ERA40
Global mean radiation balance top of the atmosphere
                   Aug. 98-Jul00

                cfsrr     Era40     AVHRR
      OLR       243.22    243.77    230.16
      NSW       244.19    236.16
      Net rad   +.97      -7.61
AVHRR OLR lower than ERBE
        Differences in
tropical divergent circulation,
        Stratus clouds
    explain differences in
        TOA radiation
ERA40 slightly stronger
   Surface stress
Hawaii apparent in ERA40,
       Not CFSRR

 Zonal surface stress
   Two year mean
Differences in atmospheric circulation
      between CFSRR and ERA40

 --differences in large scale tropical
    divergent circulation apparent

 --differences in vertical structure of
       lower tropical troposphere

    --differences in stratosphere

   --differences in two-year mean
        August 1998-July 2000
Stratus cloud regions apparent
     CFSRR more low-level
  vertical divergent circulation

Upper level divergent circulation
      less vertical extent

  Zonal mean horizontal
     Two year mean
    Tropical east Pacific
Horizontal divergence
Two year mean
   Zonal mean
   vertical velocity
   Two year mean

CFSRR vertical motion
    more confined
To lower troposphere,
 Tropical east Pacific
30S-30N cross-section
   vertical velocity
   Two year mean
 vertical velocity 7.5N

longitude cross-section

     Two year mean
CFSRR has rising motion in ITCZs more
   confined to lower troposphere,
 Weaker centers of deep rising motion
CFSRR much noisier near
Weaker vertical motion in
   tropics at 500 hPa
CFSRR slightly deeper
Tropical moisture profile

Shared By: