Hewlett Gulch Fire Plume

On 16 May 2012, the Hewlett Gulch fire underwent explosive growth, from ~1000 acres to ~5000 acres in size in one day. During the 1947-2000 UTC period (1:47-2:00pm local) on that day the Colorado Lightning Mapping Array showed several lightning flashes within the plume at high altitude, ~10 km MSL or higher (look for the lightning in northeastern Larimer County).

The CSU-CHILL and Pawnee radars were scanning the plume in a dual-Doppler configuration during this time. Below I show a composite image of reflectivity (DZ), differential reflectivity (DR; a measure of the oblateness of the smoke particles), and vertical velocity (W) at 1935 UTC, about 12 minutes before the lightning began.

The plume was difficult to distinguish from surrounding cloud in reflectivity but featured very high differential reflectivity values, often higher than 5 dB. This is due to the highly oblate smoke and ash particles. Meanwhile, vertical velocity peaked as high as 7.5 m/s (~15 knots), a respectable value for normal convection, during this time. Below I show the same plot as above but now projected in the vertical.

Look at the altitude of that updraft - it goes above 10 km MSL! It is easy to see how this plume became electrified, but what is more interesting is why it took at least 10 minutes to start producing lightning after this growth period. CSU scientists will be collaborating with New Mexico Tech scientists to study the electrification of this smoke plume/pyrocumulus cloud in more detail. Indeed, this is the first time combined polarimetric, multiple Doppler, and lightning observations have ever been made in a fire plume.

UPDATE

I created an animated GIF for the vertical plots, 1905-2130 UTC. The growth above 10 km MSL is clearly evident during the early-to-mid 1900 hour (Click on this plot and others to see original size).

AMS Lightning Conference Presentation

You can find the abstract and a link to the recorded presentation for my recent talk on energetic lightning in winter storms right here. This talk was at the 5th Conference on the Meteorological Applications of Lightning Data, held as part of the AMS Annual Meeting in Seattle. Enjoy!

New research weblog location

Blogger discontinued FTP support, so goodbye to hosting my research info on radarmet for the time being. Enjoy the new surroundings, which should look suspiciously familiar. Also, here is a poster that will be up at the European Geosciences Union General Assembly next week. It is on sprite and sprite-parent lightning observations that we at CSU have been doing. Unfortunately, I won't be there - one of my colleagues will put it up during the Wednesday poster session.

EGU 2010 Poster

Radar Conference Presentations

Here are links to my presentations at the recent AMS radar conference:

• Talk on 3-D structure of convection in NAME
• Poster on statistical analysis of the STEPS lightning and radar dataset

Richard Heene et al. (2008)

Richard Heene, the father of "Balloon Boy" Falcon Heene, was the lead author of a study published in National Weather Digest, entitled "Electromagnetic Fields Recorded in Mesocyclones" (sorry, NWD does not provide online links to articles). While the non-existent balloon journey of Falcon is now potentially a hoax, I thought it would be interesting to examine his father's sole scientific publication to date.

Richard Heene is a proponent of an electromagnetic explanation for atmospheric vortices such as tornadoes, dust devils, and tropical cyclones. This is in contrast to the predominant view within the scientific community, which favors a fluid dynamics explanation (i.e., air pressure changes, inherent spin or vorticity of air, etc.). For example, in the Vortex 2 tornado-chasing campaign last summer, no electromagnetic observations of any kind were made. The scientific community does not need to invoke electromagnetism because, based on numerous experiments and observations, as well as theoretical modeling, fluid dynamics is seen as an adequate explanation for tornadoes. The same goes for things like dust devils and hurricanes. In essence, the electromagnetic folks have a "solution" in search of a problem. We already understand tornadoes pretty well in the context of wind, pressure, and temperature. We don't need to add electricity and magnetism to the mix.

In the paper, Heene and his colleagues show some magnetic field measurements near dust devils, rotating thunderstorms, and a hurricane. Unsurprisingly, they find enhanced magnetic fields near these phenomena. This is because we already know from decades of research that dust devils, thunderstorms, and hurricanes can be electrified. Due to winds in these storms, the electric charges move, thereby creating a magnetic field. What Heene et al. fail to show is that the magnetic fields have anything to do with the formation of the vortex phenomena they studied. The main problem is that they don't demonstrate that these magnetic fields are strong enough to cause vortices on their own. This has been a constant problem for electromagnetism proponents for decades.

Electromagnetism proponents like Heene and colleagues need to first demonstrate the weaknesses of fluid dynamics theory in explaining tornadoes, etc., and then they need to demonstrate how electromagnetism can fill those holes. Demonstrating that storms can have magnetic fields, and properly describing those fields, is scientifically interesting. But there is very little evidence to support the idea that magnetic fields cause tornadoes and like phenomena, especially in contrast to fluid dynamics.

Real-Time vs. Post-Processed Impulse CMCs

Here is a scatterplot of real-time vs. post-processed impulse CMC (iCMC) values for 9 May and 20 June 2007. There is some decorrelation between the two, and a negative bias in the real-time values. However, the relationship is robust enough that we may be able to use bias-corrected, real-time data quantitatively with modest error.

As a bonus, here are real-time iCMC and post-processed continuing current CMC vs. total CMC for available 9 May and 20 June flashes. Real-time iCMC bears little relationship to total CMC, which is mainly driven by continuing current.

Total Charge Moment Change Statistics: 20 June vs. 9 May 2007

I have processed the total charge moment change (CMC) data for the 20 June and 9 May 2007 storms. Here are some preliminary statistics on the differences between sprite-producing flashes between the two storms. Besides the obvious flash-altitude differences, which have been established long before now, there are some interesting charge and charge moment change differences. First of all, 20 June has larger impulse CMC and impulse charge; while 9 May features larger continuing current duration and charge, but also weaker current amplitude. In the sum, 9 May has the smaller total CMC but greater total charge. The latter result was inferred from previous work, prior to receiving the CMC data. However, these differences are not particularly significant, based on a rank-sum test. I am still thinking about the implications of these results.

(I apologize for the formatting; Blogger appears to not handle HTML tables seamlessly.)

Variable	20 June	9 May 2007	Significance
Impulse CMC (C km)	493.2	268.1	>99%
Height CMC (km AGL)	7.5	5.7	>99%
Impulse Charge (C)	66.1	44.7	97%
Continuing Current (CC) Duration (ms)	56	129	>99%
CC Amplitude (kA km)	33.2	13.0	>99%
CC Height (km AGL)	7.8	5.7	>99%
CC CMC (C km)	1687.7	1558.9	52%
CC Charge (C)	221.5	278.3	91%
Total CMC (C km)	2180.9	1801.6	84%
Height CMC (km AGL)	7.7	5.7	>99%
Total Charge (C)	285.3	322.6	76%