Summary of issues concerning T2m


In some regions, Spring and Fall 2 m temperatures are strongly affected by the the snow
cover that was used by Reanalysis.  

        The snow cover in N/N Reanalysis had these issues.

1948-1973:  Used model generated snow cover with contraints (limits) based on 1973+ period.
      Expect to see differences with later period.
1973: NESDIS snow cover used (weekly, NH only), poor climo used in SH
1974-1994: bad snow cover  http://www.cpc.ncep.noaa.gov/products/wesley/ek.snow.html
1995-Oct 1998: good NESDIS snow cover
Oct 1998-present: USAF snow cover used
     USAF snow is higher resolution, daily (vs weekly) and is global (vs NH)
     Many people have noticed differences because of the snow cover's have different climatologies.
     Note: the NESDIS product was discontinued

        In addition, the 2 m temperature had this additional problem.

1/1986-12/1992 old boundary layer parameterization (see problem.tmp) has problems with 
       near-zero surface winds

I would use the NCEP/DOE Reanalyses as it only had the problem with the change to USAF snow cover (WNE).