The interaction of clouds and dry air in the Eastern tropical Pacific

Cloud radar observations of eastern Pacific intertropical convergence zone cloud vertical structure are interpreted in light of soundings, 100-km-scale divergence profiles calculated from precipitation radar Doppler velocities, and surface rain gauge data, from a ship at 10 degrees N, 95 degrees W during the 2001 East Pacific Investigation of Climate ( EPIC) experiment. The transition from convective to stratiform rain is clear in all four datasets, indicating a coherence from local to 100-km scale. A novel finding is dry air intrusions at altitudes of 6-8 km, often undercutting upper-level ice clouds. Two distinct dry air source regions are identified. One is a relatively dry area overlying the cooler waters of the Costa Rica oceanic thermocline dome, centered approximately 400 km east-northeast of the ship site. The other is the even drier near-equatorial subsidence zone south of 6 degrees-7 degrees N. The former source is somewhat peculiar to this specific ship location, so that the ship sample is not entirely representative of the region. The 20-25 September period is studied in detail, as it depicts two influences of the dry air on cloud vertical structure. One is the modulation of small-scale surface-based convection, evident as a weakening and narrowing of cloud radar reflectivity features. The other springs from intense sublimation cooling as differential advection brought snowing anvil clouds over the dry layers. During one half-day period of strong sublimation, the cooling rate is inferred to be several tens of degrees per day over a 100-hPa layer, based on a heat budget estimate at 100-km scale involving the horizontal wind divergence data. This is consistent with fluxing ice water contents of 0.05-0.10 g m(-3) derived from the cloud radar reflectivities. The temperature profile shows the dynamically expected response to this cooling, a positive-negative-positive temperature anomaly pattern centered on the sublimating layer. A buoyancy-sorting diagnostic model of convection indicates that these upper-troposphere temperature anomalies can cause premature detrainment of updrafts into the lower part of the cloudy layer, a feedback that may actively maintain these long-lasting dense anvils. Middle-troposphere southerly dry air inflow is also evident in large-scale analysis. Given the proximity of the dry equatorial subsidence zone to the eastern tropical Pacific, the differential advection of dry and cloudy air, the ensuing sublimation, and its dynamical aftereffects may play a role in establishing the region's climate, although the extent of their significance needs to be further established.