The base product is convection occurrence probability, which is produced from geographically-regionalized regression equations. The upgraded predictand is defined as the occurrence of either ≥ 40 dBZ radar reflectivity or one or more total lightning flashes [TL, consisting of in-cloud (IC) and cloud-to-ground (CG) flashes] or both in a 20-km gridbox during a 1-h period (2-h period previously). The grid boxes are spaced 10 km apart, which provides increased spatial resolution and yet precludes reduction in predictand occurrences that would result from non-overlapping 10-km grid boxes. The replacement of presently used CG flashes with TL flashes also enhances convection occurrence frequency, as the number of IC flashes is about six times the number of CG flashes in the TL archive furnished by Earth Networks, Inc. The predictand is further upgraded by the replacement of legacy coarse-resolution coded radar reflectivities with high resolution reflectivity data from the Multi-Radar Multi-Sensor System (MRMS), developed at the NOAA National Severe Storms Laboratory and operationally implemented in September 2014.
Increased spatial resolution and forecast probability skill is expected from new predictors based on gridded MRMS and TL observations and forecast model output from the High Resolution Rapid Refresh model (HRRR) developed at the NOAA Earth System Research Laboratory and implemented in September 2014. Additional predictors obtained through simple advection of the MRMS and TL variables are being tested, as they may supplement HRRR predictors at the shortest LAMP forecast projections. These diverse fine-scale predictors are blended with larger-scale predictors comprised of MOS (Model Output Statistics) convection probabilities based on the National Centers for Environmental Prediction North American Mesoscale model and Global Forecast System.
The multi-scale blend of observational and model-based predictors is a signature attribute of the LAMP convection probabilities. Convection “potential”, specified by post-processing the probabilities is retained in the model upgrade as this derived product may aid user interpretation of the probabilities. Experimental production of this convection forecast guidance in real time is expected to begin by October 2015. In the conference presentation we will address new aspects of the model and provide an initial evaluation of forecast performance.