Semi-Objective Method For Forecasting Cloud-to-Ground Lightning
Les Colin, NOAA/NWS, Boise, ID
A Semi-Objective Method For Forecasting Cloud-to-Ground Lightning
NOAA NWS, WFO Boise, ID
A method for forecasting “scattered” cloud-to-ground lightning (LTGCG) is presented as it might pertain to issuance of an NWS Fire Weather Red Flag Warning. The term “scattered” is arbitrarily defined as at least three LTGCG strikes within seven miles of a given point within three hours. Data sets of scattered LTGCG in a given area are collected hourly, together with corresponding initial boundary layer lifted index (LI) and surface-based CAPE model grids, which are used as predictors. The hourly lightning data are combined into 12 hour groups labeled 06Z-18Z and 18Z-06Z. Since most LTGCG in the western U.S. occurs in the 18Z-06Z time frame only that group is used, and will be associated with 00Z initial model LI and CAPE data.
Next, the forecast area is partitioned into LI, CAPE solenoids of 2 degree intervals in LI and 100 J/KG in CAPE. In each solenoid the number of grid points having scattered LTGCG is divided by the total number of grid points in that solenoid, resulting in a percentage of scattered LTGCG for that LI, CAPE combination. Then a two-dimensional plot of scattered LTGCG percentage is produced on the NWS GFE grid domain, with CAPE replacing latitude (y-axis), and LI replacing longitude (x-axis). Separate 2-D plots are made corresponding to 700 hPa RH of 10%, 30%, 50%, 70%, and 90%. A time-interpolator is applied to these five plots to produce a hundred 2-D LI, CAPE plots (one for each value of 700 hPa RH, from 0% to 99%). The process is applied to roughly 20 days near the current forecast date, and averaged. The forecast method is now complete.
To use it in an actual forecast in GFE one needs current model LI, CAPE, and 700 hPa RH grids. The values of LI and CAPE at each forecast point are referred to the 2-D LI, CAPE plot that has that point’s 700 hPa RH value. A forecast grid is thus formed. This is the percentage of scattered LTGCG.
Examples of output graphics are shown, and a series of forecast vs observed scattered LTGCG are shown side by side as visual verification.