ATS Main Site

Satellite-based Convective Storm & Convective Initiation Studies

Convective Storm Intensity Estimation

The algorithm of Mecikalski and Bedka (2006) demonstrates how the spatial, temporal and spectral information from Geostationary Operational Environmental Satellite (GOES; Menzel et al. 1998) meteorological satellite data can be used collectively to identify, track and monitor growing convective clouds in their pre-convective initiation (CI) state, so to nowcast (0-1 h forecast) CI and most recently lightning initiation (LI; Harris et al. 2010). CI is formally defined as the first occurrence of a ≥35 dBZ radar echo at the lowest elevation tilt (Weckwerth et al. 2006), whereas LI is defined as the first flash of any kind from a convective cloud. The Mecikalski and Bedka (2006) method [hereafter, the SATellite Convection AnalySis and Tracking (SATCAST) system] processes a sequence of three, daytime 15-min images, using the cloud classification approach of Berendes et al. (2008) and the cloud tracking method of Bedka et al. (2005).

Research has been completed to determine the accuracy of SATCAST (Mecikalski et al. 2008; NOAA AWG CI ATDB 2009), enhance this CI and LI nowcasting system for use within the Corridor Integrated Weather System (CIWS; Wolfson and Clark 2006), and for GOES-R and the additional infrared (IR) channels the Advanced Baseline Sounder (ABI) will provide (Mecikalski et al. 2009). In Mecikalski et al. (2009), IR data from the Spinning Enhanced Visible and Infrared Imager (SEVIRI) aboard the Meteosat Second Generation (MSG) satellite were used to understand cloud-top signatures for growing cumulus clouds prior to CI. MSG IR fields (as proxies to ABI) were used to infer three physical attributes of growing cumuli, namely: (1) cloud depth, (2) cloud-top phase transition (or glaciation), and (3) updraft strength. All of these processes are highly related to CI (Browning and Atlas 1965; Balaji and Clark 1988; Ziegler et al. 1997), and as they can be monitored on 5-15-min time intervals by geostationary satellites, the ability to forecast CI within 1-2 hour timeframes is improved significantly. Through correlation and principal component analyses used to identify the non-redundant CI fields from SEVIRI, between 6 and 8 fields were found important per physical category. Specifically, data from the 3.9 (emitted), 6.2, 7.3, 8.7, 12.0, and 13.4 µm channels, along with the traditional SATCAST fields, provide multiple checks for nowcasting updraft strength, cloud-top phase (e.g., using the 8.7–10.8 µm difference and time trend), and cloud depth (e.g., via the 6.2–7.3 µm difference).

In addition, previous research has shown that satellite and lightning data can be combined to determine storm severity from cloud-to-ground flash rates (Zipser et al. 2006, Goodman et al. 1988, Roohr and Vonder Haar 1994), a basic concept this proposal will be developed upon. These studies suggest that the prediction of lightning is possible using satellite IR imagery, which would be very useful to operational forecasters for severe weather, aviation hazard, or forest fire prediction, and highlights the dramatically improved capability that will occur when both an Imager (ABI) is flown together with a Global Lightning Mapper (GLM) on GOES-R. For the CI/LI aspect of SATCAST, and what this proposal will build upon, use of the 0.6, 0.8, 1.6, 3.9, 8.5 and 12.0 µm channels together (as on SEVIRI) will be important for monitoring trends in cloud-top phase and the size (or effective radius) of hydrometeors as clouds evolve, which is related to updraft strength (Rosenfeld et al. 2008) and lightning production (Goodman et al. 1988).

Main Results:
The proposed research will involve coupling current GOES and MSG satellite spectral information used to determine cumulus cloud and growth with data that mimics that to be collected by the GLM [from TRMM’s Lightning Imaging Sensor (LIS)] towards identifying the correlation between strong updrafts, lightning intensity, and thunderstorm severity. Here, “storm severity” is defined as those aspects of a thunderstorm defined as severe (e.g., strong surface winds >25 ms-1). This project will produce a satellite-basedstorm severity index” product designed to substantially enhance our ability to isolate updrafts within thunderstorms, and the related hazards that accompany the occurrence of strong updrafts within active portions of convective storms, specifically heavy rainfall, frequent cloud-to-ground lightning, overshooting tops (and hence locations that generate severe turbulence aloft for aircraft), and subsequent ground-level heavy rainfall, and microbursts (caused by the collapse of updraft cores). The work will benefit two societal sectors, public and aviation safety. The main aviation safety linkage will come through demonstration of overshooting top identification and spatial location mapping, and henceforth, assessing regions of likely cloud-induced turbulence associated with gravity wave generation aloft.

Using MSG SEVIRI IR data over tropical Africa for 2-3 months, information gained from recent research (Mecikalski et al. 2009; Siewert et al. 2009), along with total lightning data from TRMM LIS via the “Global Total Lightning Flash” product, correlations in space and time will be developed between CI/LI “interest fields, ” strong cumulus cloud growth and intense lightning. Cumulus clouds will be identified using the algorithm of Berendes et al. (2008) and the GOES-R cloud-typing algorithm. Given the degree of correlation between IR fields (e.g., 10.8 µm TB cloud-top cooling rates; 6.2–7.3 µm, 6.2–10.8 µm to identify updraft depth; 8.5–10.8 µm and 3.9 µm reflectance for cloud-top glaciation), lightning frequency and cumulus clouds, we expect to be able to pin-point new and strong/severe storms on spatial scales of 3 km (the MSG footprint near the Equator). NWP data will assist by providing CAPE and CIN (to determine environments capable of supporting storm storms) and precipitable water (to estimate rainfall rates and microburst potential). The severity index product will be demonstrated as a GOES-R capability, one that takes advantage of the expected high temporal, spatial and spectral resolution the ABI and GLM will provide.