HR: 0830h
AN: NG51A-0823    [PDF]
TI: Multifractal Large particle limit in rain: implications for radar rain measurements
AU: Desaulniers-Sou, N
EM: Desaulniers-Soucy.N@ems-t.ca
AF: Physics, McGill University, 3600 University st., Montreal, Que H3A 2T8 Canada
AU: Lilley, M
EM: lilleym@physics.mcgill.ca
AF: Physics, McGill University, 3600 University st., Montreal, Que H3A 2T8 Canada
AU: * Lovejoy, S
EM: lovejoy@physics.mcgill.ca
AF: Physics, McGill University, 3600 University st., Montreal, Que H3A 2T8 Canada
AU: Schertzer, D
EM: Daniel.Schertzer@cereve.enpc.fr
AF: CEREVE Ecole Nationale des Ponts et Chauss‚es, 6-8, avenue Blaise Pascal,Cit‚ Descartes, MARNE-LA-VALLEE, 77455 France
AB: Raindrop size and position data - from eighteen scenes of a = 8 m**3 region each containing 5,000 to 15,000 rain drops - obtained for 5 different storms in the HYDROP (HYdrometeor Detection and Ranging using stereO-Photography) experiment - was systematically statistically analyzed in spheres ranging from 10 cm to 2m in diameter. In four of the five storms, we found convincing evidence for the convergence to a multifractal scaling large N limit; the observed scaling exponents were quite close to those reported in the rain literature at much larger scales. By randomizing the positions of the drops, we could compare directly the true fluctuation statistics with those of the classical theory; the latter were significantly smaller. By carefully considering the meteorological conditions, in particular the turbulence intensity and the drop size distributions, we could explain the observed variations in the inner scale. In the multifractal large N limit, the drop size distribution is of secondary importance, it is the drop number, liquid water content, and other densities which tend to universal multifractal scaling limits. These results are related to the often invoked log-normal behavior of rain, but with important differences. We also show that if the densities are multifractal with the same universal multifractal index a, that we recover power law relations between their statistics, for example mean(R)=a*Mean(Z)**b (c.f. the usual deterministic relations R=a*Z**b where R, Z are random variables).
DE: 1854 Precipitation (3354)
DE: 3220 Nonlinear dynamics
DE: 3250 Fractals and multifractals
DE: 3354 Precipitation (1854)
SC: Nonlinear Geophysics [NG]
MN: 2003 Fall Meeting