Page History

Taking Acount of previous ENS solutions - Grand or Lagged Ensembles

Occasions when there is a clear divergence between the latest and previous developments in the EM and the latest HRES ensemble or ensemble mean occur increasingly rarely -  but if it does the forecasters is in a difficult position.  The HRES may even lie outside the ENS plume although maybe  Nevertheless, there  can be some agreement between the spread and jumpiness.

Image Removed

Fig6.3.4A:  Schematic illustration of the relation between the ENS (blue lines) and the three latest HRES (latest HRES shown by the thickest, longest red line, earliest HRES shown by the thinnest, shortest red line) for the case where the latest few HRES are not contained in the ENS.  The HRES results consistently lie outside the ENS plume.

Fig6.3.4B:  Schematic illustration of the relation between the latest ENS ensemble (blue lines) , a and two previous ENS ensembles (green lines), and the three latest HRES (latest HRES shown by the thickest, longest red line, earliest HRES shown by the thinnest, shortest red line).  Here the three latest HRES solutions are more similar to a previous ENS solution suggesting the latest ENS is deficient in some way.  and red lines) 

An ensemble of NWP model runs starting from different data times is known as a lagged ensemble.  Forecasters might consider combining the ENS and the HRES in an extended ensemble including previous ENS results in a current and recent ensembles into a larger lagged ensemble (as in Fig6.3.4B).  It is inappropriate to rely on the results of any individual ENS or HRES result, but ensemble member or ensemble run itself.  But if there is some indication of extreme or hazardous weather the threat should be passed on to users, but with a very low probability. Note that the y-axis refers to 'flow regime'.  Scenarios such as this are more commonly seen on meteograms, for local weather parameters, when resolution differences between HRES and ENS (e.g. affecting the representation of islands) can be pivotal. The discussion here does not relate to that aspect.