Atmospheric boundary layer (ABL) plays an important role in weather and climate. Most of the water vapor is contained in the bottom 3-km of the troposphere. Water vapor and its variations are directly tied to hydrological processes that are important to weather and climate. For example, the intensity of a hurricane is directly tied to the amount of water vapor in the tropical ABL that can support its development. The turbulently mixed convective ABL is characterized by rapid mixing, and thus negligible vertical gradients of conserved scalars, in contrast to the overlying stably-stratified free atmosphere which has at most intermittent turbulence and can support large scalar gradients. The depth of the tropical marine ABL is also dynamically linked to the large-scale atmospheric overturning circulation via subsidence in the free atmosphere. Thus, variations in the depth of the tropical marine ABL can provide information about the strength of the large-scale tropical atmospheric circulation. Despite its importance, there has been a lack of analysis of the ABL depth and an assessment of its spatial and temporal variations over the entire tropics because there is no observing system that can provide adequate data for such an analysis. The quality of the ABL depth from global re-analysis products is of lower accuracy over the tropics, compared with mid-latitudes, simply due to the lack of observations. The top altitude of the ABL is a crucial parameter for understanding and modeling of transport processes in the troposphere, weather prediction and climate models.

The six-satellite COSMIC mission launched on 14 April 2006, will provide ~ 2500 GPS radio occultation (RO) soundings per day. The COSMIC RO data recorded in OL mode shows that more than 60% of GPS RO soundings penetrate below 1 km over the tropics and the magnitude of the negative refractivity bias inherent to RO in the lowest 1-2 km compared to other observational data is substantially reduced. This provides an opportunity to examine the depth of ABL and its diurnal, seasonal and geographical variations. In this study, we develop an algorithm that automatically detects the tropical atmospheric boundary layer height using COSMIC RO data and validate this algorithm using high-resolution radiosonde data over the tropic island. We focus our analysis over the tropics, from 30 N to 30 S, over a one year period, and examine the mean pattern and the spatial and temporal structures of ABL depth variations. We also use the analysis of ABL depth based on the COSMIC RO data to validate the ABL depth as depicted by global analyses from the ECMWF global forecasting systems. This paper represents an innovative use of the GPS radio occultation data, and it will have significant implications to weather and climate communities.