Laserfiche WebLink
<br /> <br />'IJ <br />, <br />, <br /> <br />" <br />'" <br />'" <br />/ <br />I <br />\ I <br />....--...." <br /> <br />"," <br />/ <br />,/ <br />I <br />I <br />,/ \ I <br />/ \..,~-_.1 <br />I <br /> <br />IEzM,Jf\ .-. <br />..-7 :::"~~'_'-:-'",-,. <br /> <br />I \ <br />J \ /.. <br />, ............... /' <br />I '"'' J <br />I ., <br />J <br /> <br />" <br />I <br />I <br />f <br />~ ~ I <br />w I <br />ii: 4 I <br />~ f <br />~ . ....- /::.' .~~.==='7.:":.-:::.""r,-"_ <br />d IEl-4 ,\ '." <br />I \.. /'~\' <br />/ .......~........./ <br />J <br />J <br />J <br /> <br />~ <br />.2: t <br /> <br /> <br />Figure 3 - A) Maximum positive (solid - control, <br />dash-dot - seeded) and negative (dashed - control <br />and dotted - seeded) vertical electric field <br />component for early seeding. B) Same for late <br />seeding. Chaff corona threshold 30 kV/m. <br /> <br />conclusion of CCOPE and we decided that the 19 <br />July storm would be a good case to test the <br />model. Up to this point all other electrical model <br />simulations had been run on generic soundings <br />and compared with generic results. We simulated <br />the 19 July storm and made a direct comparison <br />between the model results and observations both <br />electrically and non electrically. The first results of <br />this work were presented at the Vllth International <br />Conference on Atmospheric Electricity (Helsdon et <br />a/., 1984). This was the last paper concerning <br />thunderstorm electricity that had Dr. Orville as a <br />co-author. <br />Although Dr. Orville's direct participation in <br />the modeling of thunderstorm electrification ended <br />at this point, his interest continued and his legacy <br />was established. This legacy is three-fold. First, <br />he, along with M. H. Smith, outlined the require- <br />ments for including electrical calculations in a <br />multidimensional, coupled dynamic' and micro- <br />physical model. Second, along with J. Pringle and <br />T. Stechmann, he was the first to actually <br />incorporate a rough parameterization of electri- <br />fication into such a model and begin to study <br />electrification mechanisms. Third, he mentored the <br />next generation of modelers (Chiu and Helsdon) <br />whose primary focus would be the electrical <br />aspects of thunderstorm modeling. So, despite <br />the fact that his name is not prominent in the <br />electrical modeling literature, his influence and <br />inspiration have been paramount in making <br />electrical models an accepted tool in research <br />concerning thunderstorm processes. <br /> <br />40 <br /> <br />3. PROGRESS <br />In the same time frame, results from the first <br />three-dimensional (3D) model including electrical <br />parameters were published (Rawlins, 1982). He <br />included bulk-ice microphysics and a simple non- <br />inductive ice/ice scheme along with an inductive <br />scheme. The main problem with this model was <br />that the uniform 1-km grid spacing made the re- <br />suits of limited use (electrical features of observed <br />storms often vary on length scales less than this). <br />Takahashi (1984) also expanded his warm-rain 2D <br />model to include the ice phase and the nonin- <br />ductive mechanism. He used a 200-m grid length, <br />but the 8 km depth of the grid limited the results to <br />shallow clouds. Both of these simulations were <br />generic in nature. <br />We refined our simulations and published our <br />results in 1987 (Helsdon and Farley, 1987a, b), <br />concluding that combined inductive/noninductive <br />charging produced strong electrification whereas <br />either process on its own did not. Figure 4 shows <br />the results of the 2D SEM CCOPE simulation at <br />the time of the observed (and simulated) lightning <br />flash. Shown are: (A) the net total charge density, <br />(B) the electric potential, (C) the vertical, and <br />(D) the horizontal electric field components. <br />Super-imposed on each plot is a representation of <br />the lightning channel, as calculated by the light- <br />ning parameterization scheme that was und~r <br />development. Unfortunately; there was an error In <br />the code for the non inductive-only calculation that <br />caused the calculated electric field to be too low <br />by a factor of about 4. Given that our parameter- <br />ization of the non inductive scheme was crude, we <br />did not pursue revised calculations, but worked on <br />making the scheme more quantitatively accurate <br />based on laboratory results. <br />The next phase in model development in- <br />volved the representation of lightning. Charging <br />schemes can produce strong electrification, as is <br />observed in real storms. However, in simulations <br />the build' up of charge and the resultant electric <br />field will continue unabated. In models that include <br />small ions, this leads to high ion velocities that <br />become the limiting element in the determination <br />of the time step that must be used to maintain <br />numerical stability (ion speeds of 50 to 100 m/s <br />are common in fields over 100 kV/m). More impor- <br />tantly, in real clouds fields do not continu~ to b~i1d <br />up, but are limited by the presence of lightning. <br />Because of this, models without some repre- <br />sentation of the lightning process are only useful <br />for examining the early electrification of clouds (up <br />to the time of first lightning). Several early at- <br />tempts were made to approximate the effects of <br />lightning in electrical models. <br /> <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br /> <br />38 <br />