bullet  EUMETSAT   bullet  SAFs   bullet   NWP SAF   bullet  Monitoring Reports   bullet  AMV Monitoring Report  bullet   May 2008

Impact of height assignment method - May 2008

Back to investigations index page

Density Map Vector Zonal Conclusions

Conclusions
General
  • The plots are useful for better understanding where different height assignment methods are applied; for example most high level winds are assigned heights using the CO2 slicing and WV intercept techniques.
  • In general the CO2 slicing plots exhibit the best statistics at high level and the EBBT/cloud base at lower levels.
  • In some cases the plots can be used to better understand the origin of a speed bias or poor statistics.
Meteosat
  • The WV intercept and EBBT plots show larger slow speed biases than the CO2 slicing plots for high level extra-tropical Meteosat-9 IR and WV winds.
  • The IR CO2 slicing zonal plot shows a marked slow speed bias below 450 hPa in the extra-tropics (Feature 2.9 in the 3rd analysis report).
GOES
  • There is little difference in the statistics for winds assigned EBBT or cloud base heights, although it is worth noting that the latter are only used over the sea.
  • The density plots show there are a few GOES-12 low level winds assigned CO2 slicing heights. These have a slow speed bias of ~6 m/s.
  • The GOES winds assigned WV intercept heights show a large slow speed bias at mid level in the extra-tropics (zonal plot example shown below, but also see map plots). This was described under Feature 2.9 in the 3rd analysis report
    Unedited GOES-11 IR O-B speed bias
    EBBT WV intercept
     
  • The zonal plots of GOES winds assigned EBBT heights show a marked fast speed bias at 300-500 hPa in the tropics (e.g. above). This only affects a small number of winds. The unedited GOES-11 IR high and mid level EBBT map plots show the geographic location of this fast speed bias. Also notice a slow speed bias off the coast of California (mid level plot).
  • Similar to Meteosat-9, the GOES-12 CO2 slicing zonal plot shows two distinct regions of extra-tropical slow bias (one above 300 hPa and one below 500 hPa).
Polar
  • There is little difference in the statistics for different height assignment methods.
  • It is worth noting that comparitively few MODIS winds are assigned WV intercept heights. This method is not available for the AVHRR winds.