Walker circulation and tropical upper tropospheric water vapor

Maps of the boundary layer divergent wind component derived from velocity potential for two 35-day periods in 1991 and 1994 exhibit a pattern of convergence in the lower levels in the tropical western Pacific, following the Sun between hemispheres, and divergence in the eastern Pacific, these two features agreeing with schematics of the Walker circulation. Vertical motion maps show rising motion in the western Pacific and sinking motion in the eastern Pacific; water vapor maps derived from the microwave limb sounder on the Upper Atmosphere Research Satellite show a similar picture with moist plumes evident at 215 hPa in rising regions and dryness in sinking regions. While these three indicators of vertical motion are internally consistent and fit with accepted views of the Walker circulation, it is shown that there is not equally clear evidence that the eastern and western regions are directly linked by horizontal motions, particularly in the upper troposphere. Mass flux studies based on the divergent winds show that the southeastern Pacific anticyclonic region is fed by upper tropospheric inflow from all sides; in October 1991 the inflow from the west, east, south, and north contributed almost equally, while in February 1994 the inflow from the west was the largest contribution, but flow across the eastern and southern boundaries was also important. The water vapor maps suggest that a substantial fraction of air in the upper troposphere to the west of South America originates over the continent. Lidar measurements of cirrus show good evidence of diverging flow into the northern hemisphere from the rising motion in the west Pacific near 10 degrees S during February 1994. The situation in the southeastern Pacific cell in October 1991 is somewhat analogous to that in the northwestern Pacific cell in February 1994 in that flow from a nearby landmass is entrained into the circulation. From daily water vapor maps coupled with potential vorticity maps and cross sections, it is suggested that direct exchange between the stratosphere and the troposphere in the form of almost vertical plumes may occur in subsiding regions in the tropics.