Date of Award

January 2013

Document Type


Degree Name

Master of Science (MS)


Atmospheric Sciences

First Advisor

David J. Delene


The research goal is to determine if a relationship exists between Aerosol Optical Depth (AOD) and radar-derived precipitation in Mali, a country located in West Africa. Precipitation parameters derived from 2006, 2007 and 2008 C-band radar measurements are compared against Noon to 2 p.m. averaged AOD measurements. The AOD and Angstrom Exponent (AE) are from an Aerosol Robotic Network (AERONET) site located 240 km northeast of Bamako, and radar cell (20 dBZ threshold) tracks are generated using the Thunderstorm Identification, Tracking, Analysis and Nowcasting radar software (TITAN). Only isolated TITAN cells are used in the analysis. In addition, environmental filters are applied so that only cases with large Convective Available Potential Energy (CAPE) and weak 0 to 6 km Above Ground Level (AGL) vertical speed shear are included in the analyzed dataset to increase the likelihood of isolated convective cells. Additionally, only cases with low Lifting Condensation Level (LCL) are used to increase the likelihood of surface-based convection. Days with an Angstrom Exponent (AE) less than or equal to 0.35 are removed to reduce the influence of desert dust. With the exclusion of low AE days, an increase in AOD is assumed to imply larger aerosol concentrations in the lower troposphere.

Analysis of the filtered data set shows a significant negative relationship between AOD and number of cells per day; a p-value of 0.009 and R2 of 0.20 indicate a one percent probability that the slope occurred by chance and that 20 % of the decrease is due to AOD, respectively. In contrast, linear regressions of each of the following are not statistically significant: AOD versus precipitation flux per cell, AOD versus duration per cell and AOD versus precipitation flux per day. Based on linear regression analysis between non-dust AOD (aerosol burden) and radar data, the conclusion is made that increased concentrations of aerosols result in fewer storm cells (perhaps because it is more difficult for initial clouds to produce precipitation-sized drops); however, the aerosol burden does not affect the amount of precipitation produced by individual detected cells that are able to form. These results, as well as previous studies, suggest that Mali's operational weather modification program should concentrate its resources by conducting hygroscopic seeding on days with high AE and AOD because it is those days that clouds would benefit the most from a broadening of the cloud drop size distribution.