Forecast Description
Motion Vectors
Much work has gone into improving the motion vectors used in NCWF-2 over those used in NCWF-1. A key difference between the NCWF-1 vectors and NCWF-2 vectors is that NCWF-1 vectors were tied to individual storm centers while NCWF-2 motion vectors are provided on a 4 km grid.
(a) RUC steering level winds (white) with TITAN storm detection and motion vectors (black contours and line segments) and (b) same as (a) except with merged motion vector field (black vectors).
The gridded motion vector field is produced by blending RUC-derived steering level flow with TITAN-derived motion vectors (see image above). Winds from sigma level 20 (~700 mb) from the 1 hour RUC forecast (updated hourly) are interpolated to a 4 km grid. These winds were found to be well correlated with the motion of small storms. TITAN vectors are computed for storms that have been tracked for 20 min or more. The TITAN motion vectors then go through a rigorous quality assessment that checks for consistency of the storm motion with time. The QC’d TITAN vectors are then expanded around the area of detected convection by 240 km and are used to replace the RUC steering-level winds where available. In areas where TITAN vectors from two or more tracked storms overlap, the motion vectors are determined using a distance-weighted average.
Final motion vector field including spatial and temporal smoothing.
Toggle image to see NCHD and the 120 min extrapolation of the trended NCHD.
The final merged field undergoes temporal smoothing so that there is a smooth transition in time between RUC and TITAN vectors in successive forecasts.