Recent Improvements and Observations of Super Rapid Scan Mesoscale Atmospheric Motion Vector Flow Fields over Deep Convection

Over the last two years super rapid scan data operations for GOES-R (SRSOR) mesoscale atmospheric motion vectors (mAMVs) have been used to objectively identify flow fields over mature deep convection. Recent findings have suggested that flow features unique to supercell thunderstorms are objectively seen with current GOES-14 SRSOR 1-min data (Apke et al. 2016). Features such as derived large, persistent cloud top divergence (CTD) maxima and cloud top vorticity (CTV) “couplets” (adjacent maxima in cyclonic and anticyclonic vertical vorticity) were observed over supercells that were not identifiable in 5-15 min satellite scanning rates. Apke et al. (2016) suggested that SRSOR derived CTD may have use in observing variables such as updraft strength without the need for ground based multi-Doppler radar information. Multi-Doppler radar information is typically limited to 5-min time scales over narrow domains, thus SRSOR derived CTD can reveal updraft characteristics of storms over large domains not previously observable at a 1-min time scale. Since updraft characteristics are useful in the immediate nowcasting of severe weather such as hail and the initiation and intensification of cloud lightning production, an accurate calculation of CTD and CTV flow fields is valuable to forecasters interested in monitoring and predicting the occurrence of severe weather and the intensification of thunderstorm lightning production.

While Apke et al. (2016) did show utility in flow fields derived with a Barnes (1973) objective analysis, they were limited in their ability to apply background flow fields to the data, and further discussed accuracy problems with the magnitudes of CTD and CTV when large upper level mAMV observation spatial data gaps were present or the data were not uniformly distributed (so called “ballooning”). The original approach also used basic brightness temperature mAMV height assignment techniques to truncate low level vectors from the anvil level cloud analysis, however Apke et al. (2016) discussed a need to improve this truncation method as low level vector contamination was noted to cause large changes in the two-dimensional derived CTD and CTV fields. This presentation will show an update of the Apke et al. (2016) approach to a recursive filter analysis (RFA; Hayden and Purser, 1995). The RFA is not sensitive to changes in data uniformity and spatial density, and allows for the inclusion of background model data into the derivation of the upper level flow field. Using CTD derived from the RFA, analyses will be presented on 6 deep convective case studies with comparison to hail size. This presentation will also go in depth into comparison of RFA analyzed flow fields and total lightning information from ground-based lightning mapping arrays.

Preliminary results show that the “ballooning” problems present in Apke et al. (2016) are effectively removed with the application of the RFA without sacrificing CTD maxima and CTV “Couplets” observed in the well sampled flow fields observed over deep convection using the original Barnes (1973) analysis. Problems of low-level vector contamination in the upper level flow field calculation are reduced using the observation quality control variables present in the RFA when including a background numerical flow field without the need for computationally expensive cloud masking procedures. Using CTD also helps to identify lightning jumps, when they occur in relation to changes in updraft strength and updraft volume, and will help to separate these jumps in lightning flash counts from lightning jumps that may occur due simply to graupel mass increase alone. Observations of both localized flashes and maximum flash density counts are found to match well with derived CTD using SRSOR datasets. Given results that correlate temporal CTD modulations with severe weather reports (e.g., large hail), methods for combining GLM data with SRSOR CTD will be discussed with respect to forecasting severe weather events, and may prove useful to severe weather forecasters over regions that do not have sufficient radar coverage. Further updates will be provided on the development of a real time GOES-R mAMV flow field product for forecaster use and experimentation upon the launch of the new satellite.