Laserfiche WebLink
<br />2176 <br /> <br />JOURNAL OF THE ATMOSPHERIC SCIENCES <br /> <br />VOLUME 3S <br /> <br />60 <br /> <br />50 <br /> <br />- Unrimed Plane Dendrites <br />-- LGT-MDT Rimed Plane Dendrites <br />........... HVY Rimed Plane Dendrites <br />-..- LGT- MDT Rimed Spatial Crystals <br />-.- Graupel <br /> <br />I <br /> <br />../;././ <br />/....,/ <br />/.:........../ <br /> <br />/.. ......./.. --- <br /> <br />../ ././ ~.;...-~ <br />~ ......... ...--;::;.---"'- <br />....,.,......- ~'l'::"-----_"- <br />.--" .... <br /> <br />40 <br /> <br />z 30 <br /> <br />20 <br /> <br />10 <br /> <br />o <br />10-5 10-4 10-3 1()2 10-' 100 <br />l:.M (gm-cm/sec) <br /> <br />FIG. 6. Composite fragment generation functions for all five crystal types studied showing <br />the number N of fragments produced per collision as a function of the change of momentum <br />!!.M in the collision. <br /> <br />depending on the initial crystal type and size distribu- <br />tions, on the size distributions of the fragments which <br />are produced in the cloud and on the rate of accre- <br />tion and diffusion which takes place on the original <br />crystals and fragments. Since these factors change in <br />a time-dependent manner, a general analytic fnnction <br />for K (t) is not obtainable. Therefore, a numerical <br />model was constructed to approximate the change <br />in K (t) and integrate (17) in time. <br />The model was used in two modes: 1) to determine <br />the best crystal combinations and size distributions <br />for generating secondary particles and 2) to determine <br />the magnitude of secondary particle generation for <br />a specific time-dependent case. In the first mode the <br />model was run for one time step for each of 252 dif- <br />ferent crystal types and size distributions combina- <br />tions. Ko was determined for each combination. The <br />largest Ko's indicated the best crystal types and size <br />distributions for generating secondary particles. In the <br />second mode the model was run for fifteen time steps <br />for special cases. In these computations K (t) varied <br />with time and, consequently, C(t) had a different form <br />from that when K (t) was' equal to a constant K o. <br />The magnitude of secondary particle generation was <br />found to be less than the case for K(t)=Ko with some <br />assumptions and greater for other assumptions. <br /> <br />b. Determination of optimum rate const(Jnts <br /> <br />The numerical model was run with combinations <br />of the five crystal types. Including the combinations <br />of each crystal type with itself, fifteen possible com- <br />binations existed. Within a given combination of two <br />crystals, four size distributions for each type of crystals <br /> <br />were studied giving sixteen values of K 0 for each of <br />the fifteen combinations or a total of 240 values of Ko. <br />The following features of crystal-crystal collisions were <br />found from the distributions of K 0: <br /> <br />1) The broader the size distribution the larger the <br />value of K o. Reviewing from Section 2 we recall that <br />the larger Ko, the greater the generation of secondary <br />particles for a given crystal concentration. Thus, we <br />can say, the broader the crystal distribution the <br />greater the secondary particle generation. This effect <br />is most likely due to the greater concentration of <br />crystals at the larger sizes for broad distributions. <br />The larger the crystal size the larger the relative <br />velocity and the more fragments produced in a col- <br />lision. The collision frequency increases as the square <br />of the concentration so that more crystals at large <br />sizes rapidly increases the secondary particle generation. <br />2) Unrimed plane dendrites should not generate <br />secondary particles among themselves. Ko was equal <br />to zero for all sizes distributions studied. Since den- <br />drites appear to be the most fragile unrimed type of <br />crystal, this finding would seem to imply that no <br />unrimed crystals may produce secondary particles by <br />mechanical fracturing. However, plane dendrites also <br />have the lowest terminal velocity for a given crystal <br />size and other crystals which appear less fragile may <br />still produce fragments because of their greater fall <br />velocities. We find; however, that the magnitude of Ko <br />for even rimed crystals' is so small that a statement <br />to the effect that unrimed crystals' cannot generate <br />secondary particles is probably correct. <br />3) The greater the amount of rime the larger Ko <br />and consequently, the greater the secondary particle <br />generation. This finding is similar to that obtained <br />