...
Due to computational constraints these These two variables represent maximum values in the past 6 hours derived from a combination of (A) standard hourly model output (not at each model time step, and (B) values of CAPE and CAPE-shear computed every time step (450 s for HRES and 720 s for ENS) at gridpoints where the convection scheme is active. The method of computation for (A) is the standard ECMWF method used for some time. However the way in which CAPE is computed in (B) is physically different (and is done within the convection scheme, with zeros applied outside convectively active areas). These new parameters may help users to better assess the potential for convection than their instantaneous counterparts.using instantaneous values. As the plots below illustrate, the new fields tend to be smoother and more continuous. This is because model CAPE at a given location on a given day can increase and decrease quite sharply, over timescales much less than 6 hours.
In IFS cycle 46r1, implemented on 11 June 2019, mxcape6 and mxcapes6 have changed – maximum values are derived only from hourly model output and (B) is not considered any more. This addresses the ambiguity that may arise by combining (A) and (B). Now for example T+24h mxcape6 includes CAPE values from the following forecast steps: T+19h, T+20h, T+21h, T+22h, T+23h and T+24h.
ECMWF is considering adapting the method of computation of these new fields in the next cycle, to achieve greater consistency.
Maximum CAPE versus instantaneous CAPE (left) and Maximum CAPE-shear versus instantaneous CAPE-shear (right).