Map of Boulder showing the location of the Foothills Laboratory on Mitchell Lane northeast of the Foothills Pkwy and Valmont Road intersection.

Note: The groups referred to in Carlson's email are the Observing Facilities Advisory Panel (OFAP) and the Facilities Advisory Council (FAC). The NSF Deployment Pool covers funding for the following facilities that STEPS requested: the CSU-CHILL and NCAR/S-Pol radars, the SDSMT T-28, and the two NCAR mobile GLASS. Any other instruments such as the UND Citation would be funded from the grants programs. Further information about the facilities request process can be found at NCAR/ATD.

Brief review of scientific goals and experimental design presented in the STEPS Scientific Overview Document:
    1. Elements of STEPS to be considered.
    2. Field instruments that were proposed.
    3. Radar network including coverage areas for Doppler and dual-polarimetric measurements.
    4. Lightning mapping network within the area of triple-Doppler coverage.

Radar networks consisting of the original STEPS radar network near Goodland Kansas and three possible networks near Greeley Colorado. Characteristics of the networks, especially areas of coverage for the research dual-Doppler, dual-polarimetric radars are shown in the attached table. The plots of all radar networks are centered on Akron Colorado and at the same scale for ease of comparison. The areas of dual-Doppler coverage are defined by either 30-deg or 20-deg angles between views from the separate radar locations. The dual polarimetric, dual Doppler areas for the two research radars (CSU-CHILL and NCAR/SPol) are shown by circles (red outlines) passing through the locations of these radars. Other dual-Doppler lobes outside the inner area of triple-Doppler coverage are shown with black outlines. The NMIMT lightning mapping system would be located within each of the inner triple-Doppler coverage areas. Topographic heights are contoured every 1000 ft, with a 3-peat pattern of thickness starting at 3000 ft just east of Goodland. (Networks provided by L. Jay Miller.)

In addition there is a map of Colorado state property in the vicinity of Bethune Colorado where we might consider placing the CSU-CHILL radar. Note: More on this and possible ways of defraying the cost of moving CSU-CHILL will be a topic of discussion led by Steve Rutledge.

If the reviewer's used my plot to decide that Greeley was better than CO/KS border for STEPS, then they are mis-interpreting (mis-using) that plot. The small bullseye is actually the northeastern suburbs of Denver and surrounding area. This bullseye in severe weather events is highly modulated by population. That is not to say that severe weather does not occur there with some frequency; I just think it is misleading. Second, the storms are certainly different. This would be more difficult to show with the severe reports (perhaps size of hail and intensity of tornadoes would show this); but clearly supercells and certainly LP supercells are more common by Goodland. Thirdly, my plot clearly shows that predominately +CG severe storms are > 2x more likely by the KS/CO border. Again, not that Greeley and vicinity does not get them (e.g., 7 June 95 and others) but they are more frequent by Goodland. Lastly, the coveted +CG monster event (high +CG flash density in addition to high %+CG) rarely occurs around Greeley; there is a much better chance by Goodland. So, that's four reasons right there. From the perspective of studying severe storms that produce predominately +CG lightning, Goodland is far superior to a domain closer to the Front Range. The following is a 10-year (1989-1998) climatology of +CG lightning in severe storms (in particular, note Figs. 2 and 5).

Four panel image of NLDN and OTD data analyses. The upper right panel shows the IC:CG ratio computed from an analyses of OTD satellite and NLDN data for the period May 1995 - June 1999. This figure was provided by Larry Carey (CSU) with the permission of Dennis Boccippio (NASA/MSFC).

Results from a three-year study of CG lightning activity over the CONUS. The attached images show a strong west-to-east transition in CG lightning activity/behavior from the foothills of the Rocky Mountains in Colorado to the High Plains of Kansas. Moving east from the foothills, the production of positive and negative CG lightning is seen to separate by convective type and timing: isolated storms tend to dominate the production of +CGs during the early evening while MCSs tend to dominate the production of -CGs during the nocturnal period.

It is important to point out that this "separation" process is observed from the CO/KS border northeastward to Wisconsin. Over this southwest-to-northeast oriented region, the correlation between severe weather (large hail and/or tornadoes) and positive strike dominated (PSD) storms is a maximum; greater than 50% of large hail and tornado reports are associated with an anomalously high fraction of positive CG lightning (Larry Carey, CSU/Atmos). For this reason, I believe that the impact of the STEPS project extends well beyond the borders of Colorado and Kansas to include large portions of the central and northern Great Plains. As an example, the 3 May 1999 tornado outbreak over Oklahoma comprised multiple storms producing predominantly positive CG lightning (see attached image). It is crucial that the STEPS program be established in a location to sample the greatest number of PSD storms. Research by Larry Carey suggests that this occurs over western Kansas between Dodge City and Goodland. From Zajac and Rutledge (1999; ICAE).

The attached figure shows the number of transient luminous events (TLEs) by storm for the period 1996-1999. The STEPS study region within 200 km range of the center of the lightning mapping system (LMS) is outlined in blue.

The relocation recommendation is not appropriate for a variety of reasons, all related to the dramatically decreased probability of observing the specific meteorological phenomena that are the foci of the STEPS program.

These include:

1. Greatly reduced numbers of LP supercells in the greater Greeley region. Dutring seven summers of nearly continuous monitoring from the Yucca Ridge Field Station, we are aware of only two LP supercells within 100 km of YRFS (and rather poor specimens at that).
2. Not only do the number of +CGs increase significantly as one moves eastward, so do the number of large peak current (>75 kA) events associated with sprites and elves (see Lyons et al. 1998).
3. Storms rarely produce sprites and elves until they are 150-200 km east of YRFS. Moving the STEPS domain so close to the mountains would reduce by more than 50% the number of events occurring above the Lightning Mapping System (LMS) coverage area (Lyons et al. 1999).
4. Blue jets are thought by some to occur above hail-producing supercells. When such storms occur at night in eastern Colorado (permitting blue jet detection by low-light imaging), they are much more likely to be in the Kansas border region than further west.

References:

Lyons, W.A., T.E. Nelson, R.A. Armstrong, E.R. Williams, D.M. Suszcynsky, R. Strabley, M. Taylor, and L. Gardner, 1999: Characteristics of thunderstorms and lightning flashes which produce mesospheric transient luminous events. Procs., Intl. Conf. on Atmospheric Electricity, Lakes Guntersville, AL, 4 pp.

Lyons. W.A., M.A. Uliasz and T.E. Nelson, 1998: Large peak current cloud-to-ground lightning flashes during the summer months in the contiguous United States. Mon. Wea. Rev., 126, 2217-2233.