Dr. Michael L. Larsen – Research Overview

Measuring and Quantifying Number Fluctuations of Environmental Aerosol Particles, Cloud Drops, and Raindrops

 

There are many things that are random but not “perfectly random”.  The distinction is subtle, however, and can cause a rather substantial amount of confusion.

 

Assuming youtube doesn’t take it down anytime soon, watch this famous clip from the movie “Singin’ in the Rain”.  Gene Kelly’s dancing is rather impressive, but everytime I watch the clip I’m drawn to looking at the rain.  According to the movie’s imdb trivia page, real rain was not used for the clip but rather water (mixed with milk for better appearance on film) was generated with overhead sprays.

 

The movie was made in 1952, so there weren’t CGI techniques available to generate the rain and I don’t want to criticize an obvious classic, but if you watch the rain you can tell it is fake.  Why?  It is too regular.

 

You’ve probably heard of old sayings like “when it rains it pours”.  Another (more vivid) linguistic phrase is “sheets of rain”.  This isn’t an empty phrase – you can see, hear, and (if you’re unfortunate to be caught in a downpour) feel this phenomenon.  It occurs when, already raining, the intensity of the rainfall abruptly increases.  If you look at the pavement, you can often see the edge between the weaker and stronger rainfall moving.

 

The point that I’m trying to make with these anecdotes is that everyday experience suggests that rainfall is variable.  Rainfall (usually) doesn’t behave like water coming out of a steady spray.

 

The people who analyze rainfall on relatively short spatial and time scales (e.g. smaller than a city and shorter than a day) are gradually acknowledging what common experience has told us for decades if not centuries.  Raindrops fall randomly, but statistically their fall cannot be described as perfectly random.

 

A similar debate has been going on among atmospheric scientists for the past few decades regarding cloud particles and aerosol particles.  For theoretical purposes, it would be rather convenient if one could argue that the positions of cloud droplets in a cloud could be considered perfectly random.  And, in fact, there aren’t many forces between cloud particles, so – in a classic confusion between correlation and causation – it has long been generally assumed that cloud particle positions are statistically independent since there are no substantial physical dependencies between them.

 

By now you probably realize that I believe the data supports the other viewpoint – cloud droplets (and, it turns out, aerosol particles) appear to have some tendency to be located closer to each other than you would expect if they were distributed perfectly randomly.  This is a statistical statement, so care must be taken in interpreting the implications of the statement.  However, with the advent of instrumentation that can identify the positions of individual particles in a cloud, the evidence for atmospheric particulates being distributed in a non-Poisson (not perfectly random) statistical manner is becoming increasingly convincing.

 

It appears that particulate spatial positions are most appropriately described by a “random but correlated” statistical framework.

 

***

 

Much more evidence needs to be gathered, however.  Because investigations used to assume perfect randomness, there isn’t much data or analysis yet completed to find “how far from perfect” the spatial distributions actually are.  The instrumentation needed to measure these things did not exist because nobody thought that an instrument that measured these things was necessary.

 

Now, such instruments are in development and more data is being taken to verify that departures from perfect randomness do exist for (nearly) all atmospheric particulates.

 

Dr. Larsen works on using new technology and modifying old technology in order to quantify how correlated these systems are and what influences these correlations.  Some investigators claim that the correlations are due to the particles responding to the turbulence in the air.  Others believe it may have to do with the nature and location of the sources of the particles in the first place.  (Cloud particles are created by vapor condensing on aerosol particles which are themselves created from a variety of sources.  However, many aerosol sources – for example dust coming from the surface of the earth – start out with the particles very close together.  It may not be that we need to ask “why are the particles getting close together” but rather “why would expect the particles to stop being close together”.)

 

The large amount of relatively recent (last 20 years or so) work on particle spatial distributions has developed more questions than answers, and Dr. Larsen’s research touches on a number of them.

 

This particular research project is very closely related to several of the other projects, including how clustering might influence:

 

Radiative transport

Estimates of total particle concentrations

Meteorological Radar Returns

Airborne Pathogen Inhalation Risk

How we simulate microphysical phenomena

How particles interact geometrically

How particles disperse in the frequency domain

 

Back to LARSEN RESEARCH

Back to MAIN PAGE

Send a comment to: larsenml@unk.edu

[Return to Physics Department] This web site courtesy of the Department
 of Physics and Physical Science
In cooperation with the [UNK Home]University of Nebraska at Kearney