GPS observations of precipitable water and implications for the predictability of precipitation during the North American Monsoon

I. Introduction: The semi-arid to arid North American Southwest has experienced, in recent decades, tremendous population growth despite limited water supplies. Water availability is highly susceptible to climate variability and change and is therefore of great concern. Climate models generally predict the region will experience warmer and drier winters in the future. During summer, incursions of moist tropical air associated with the North American Monsoon (NAM) bring critical convective precipitation (eg. Adams and Comrie, 1997; Higgins et al., 2006). Representing deep convective regimes in tropical regions, however, has been problematic in large-scale numerical models (e.g., general circulation models (GCMs)). The NAM, no exception, is represented poorly, if at all, in GCMs (e.g. Lee et al., 2007). As a result, the future behavior of the monsoon is uncertain but may intensify if the land-sea temperature contrast increases. To improve upon this situation, the North American Monsoon Experiment (NAME) was initiated to better understand convective precipitation and its predictability during the NAM and ultimately improve its modeling. This goal is challenging given the complex interplay between moisture, diurnal heating, large and small-scale dynamics and complex terrain in convective rainfall. Results from data gathered during the 2004 NAME field campaign are briefly summarized here and new results assessing precipitation sensitivity to precipitable water vapor in high resolution model simulations are reported. Since precipitation condenses from water vapor, improving precipitation prediction requires better knowledge of the quantity and distribution of water vapor. As such, water vapor was the focus of the International H 2 O Project (IHOP) campaign to gain insight into moist convection in the U.S. Southern Great Plains (Weckworth et al., 2004). Like other remote tropical regions, observing water vapor in the NAM region, particularly Mexico, is challenging. During the monsoon, satellite IR observations of water vapor are limited due to frequent cloudiness and microwave water vapor observations in the region are limited due to uncertainties in land surface emissivity. Radiosondes, somewhat intermittent, are found only in Empalme, Mazatlan and La Paz along the Gulf of California coast, Chihuahua, east of the Sierra Madre Occidental (SMO) mountains and Tucson to the north.