McMASTER UNIVERSITY STATISTICS SEMINAR

Week of January 25 - 29, 1999

SPEAKER:	Dr Harry Pavlopoulos Department of Statistics, Athens University of Economics & Business, and Center for the Study of Earth from Space, University of Colorado at Boulder
TITLE:	"Scaling Properties of Rain Fields"
DAY:	Wednesday, January 27, 1999
TIME:	3:30 p.m. [Coffee in BSB-202 at 3:00 p.m.]
PLACE:	BSB-108

SUMMARY

Rainfall is a highly complex physical phenomenon, manifesting its complexity via dynamic spatial distributions of rain rate fluxes. I will first explain an empiricism regarding a relationship between instantaneous spatial averages of rain rate and the fractional area where rain rate exceeds a given threshold level over a region of fixed size. This relationship was obtained by the need to measure rain from space orbiting satellites. One possible explanation for this relationship is presented, leading to the so called "optimal threshold method" for estimation of spatially averaged rain fluxes from satellite data. The change of fractional area, where rain rate exceeds zero threshold level, with spatial scale exhibits 'statistical scaling'. An overview of scaling properties of rainfall will be given, along with an introduction to the theory of discrete multiplicative cascade processes. They provide a new mathematical framework within which some of the contemporary approaches to modelling spatio-temporal statistics of rain fields is being pursued.

ABOUT THE SPEAKER

Dr Haralabos (Harry) Pavlopoulos obtained his B.Sc. in Mathematics from University of Patras, Greece, and his M.A. and Ph.D. in Mathematical Statistics from University of Maryland, USA. He is presently on a sabbatical leave at the Center for the Study of Earth from Space / CIRES (Cooperative Institute for Research in Environmental Sciences), University of Colorado at Boulder. He is a Lecturer in the Department of Statistics at Athens University of Economics and Business, Greece. He is an active member of the Committee on Probability and Statistics in the Physical Sciences of the Bernoulli Society and of the European TMR Network on Spatial and Computational Statistics. Dr Pavlopoulos's research interests include spatial and temporal modeling and inference for intermittent non-negative random fields, random cascades and their multiscaling properties, environmental statistics, time series analysis of chaotic processes and non-parametric regression.

REFERENCES

The following references have been provided by Dr Pavlopoulos to be used as background for his talk. They are on reserve at Thode Library (STATS 770: Statistics Seminar).

[1] P Freidlin, M. & Pavlopoulos, H. (1997), "On a Stochastic Model for Moisture Budget in an Eulerian Atmospheric Column," ENVIRONMETRICS 8, No. 5, pp. 425-440.

[2] Pavlopoulos, H. & Makatis, G. (1998), "Spectral Multiscaling on Spatially Averaged Rain Rate: A Hint for Spatio-Temporal Modeling," ENVIRONMETRICS 9, No.6, pp. 689-713.