This is an old revision of the document!
Effective local roughness length was determined from averaging the land-use-grid related roughness length over a square area of 4000 x 4000 m2.
Due to change of turbulence from step M to M+1 (either by change of stability class or by change of roughness length) the sigma parameters of a puff can only increase, never decrease. This increase was applied both to the sigma parameters at puff head and puff end.
puff end stay the same as the sigma parameters at the head of the puff at the time step before. This makes a more realistic smooth transition of the puff size.
The effective transport velocity of a puff during release from the source was the wind velocity at source height. Especially at low release heights this could cause a too short initial puff length because the increase of the vertical size of the puff was neglected. During the following time step the end of this puff again was transported with that speed, but the head was transported with a vertical mass profile weighted enhanced speed.
modified: it is 50% urel plus 50% u100 for stability A to E, and 70% urel plus 30% u100 for stability F. (urel is wind speed at release height).
In cases of sideward transport of a long puff with strong difference of displacements at puff head and puff end it could happen that the geometry of the parallelogram needed for calculation of time-integrated concentrations failed.
difference of displacements of puff head and puff end would exceed 20 %, then transport of the puff is calculated with the average of wind vectors at puff head and puff end.