Global Positioning System (GPS) radio occultations (ROs) are analyzed through regions of clouds to quantify properties of clouds and test the accuracy of the RO technique. Dry and moist occultation data from the aforementioned GPS satellites over the Americas and the Atlantic Ocean during May 2002 are obtained and filtered into ten categories based upon occultation latitude and GOES-8 satellite imagery analyses of clouds in the region. Comparisons of mean profiles and standard deviations from the mean for all data sets, fractional differences from the mean for the dry data sets and features observed in the moist data sets are performed. It is shown that there are noticeable impacts in data from bending angle, refractivity, temperature and vapor pressure retrievals owing to unique properties of clouds as sensed by the GPS RO technique. These impacts are the result primarily of the inherently different properties associated with each of the major cloud types of cirrus, convective, cumulus and stratus. These retrievals also serve to highlight the lack of variation found in the height of the tropopause in the tropics and the greater spread in the height of the tropopause in the midlatitudes.

It is also shown that inversion layers can be analyzed with a high degree of precision when taking into account cloud categorizations. In analyzing for inversions in vapor pressure data, it is shown that there are possible disagreements between moist temperature and humidity data; this is likely due in part to a combination of the methods used to the calculation of both sets of variables and the inherent error in moist temperature retrievals. From these findings, a method to improve upon the retrieval of temperature and vapor pressure data through regions of clouds is proposed utilizing the Clausius-Clapeyron relationship.