The experimental, upgraded LAMP convection predictand is defined as occurrence/non-occurrence of ≥ 40 dBZ radar reflectivity and/or ≥ 1 TL flashes in 20-km gridboxes spaced 10 km apart over a 1-h period. The corresponding lightning predictand is specified as the TL component of the convection definition. Since the convection specification is dominated by the radar component, redundancy between the two predictands is small. Also, replacement of CG flashes embodying the currently operational convection or lightning predictand with (more frequent) TL flashes in the corresponding upgraded predictand provides a more complete specification of convection or lightning events and accommodates higher temporal resolution.
The predictor base for the multiple-regression-based convection or TL probabilities consists of a blend of the most recent and advected radar and lightning parameters, HRRR products, and the National Centers for Environmental Prediction Global Forecast System (GFS) - and North American Mesoscale Model (NAM) -based model output statistics (MOS) convection or TL probabilities. The observational predictors yield fine detail and high skill in the first few LAMP hourly valid periods, HRRR predictors skillfully extend detail out to 17 hours, and MOS probabilities specify large scale forecast patterns (with modest skill) for all LAMP forecast projections.
The upgraded LAMP convection and TL guidance products are presently being issued experimentally in real time for user evaluation and feedback (early convection feedback has been quite favorable). The conference presentation will describe how the probability and potential products are formulated, their performance, and the anticipated implementation schedule.