Spatial and temporal properties of water vapor and latent energy flux overa riparian canopy

A scanning, volume-imaging Raman lidar was used in August 1997 to map the w ater vapor and latent energy flux fields in southern Arizona in support of the (Semi-Arid Land Surface Atmosphere) SALSA program. The SALSA experiment was designed to estimate evapotranspiration over a cottonwood riparian cor ridor and the adjacent mesquite-grass community. The lidar derived water va por images showed microscale convective structures with a resolution of 1.5 m, and mapped fluxes with 75 m spatial resolution. Comparisons of water vapor means over cottonwoods and adjacent grasses show similar values over both surfaces, but the spatial Variability over the co ttonwoods was substantially higher than over the grasses. Lidar images supp ort the idea that the enhanced variability over the cottonwoods is reflecte d in the presence of spatially coherent microscale structures. Interestingl y, these microscale structures appear to weaken during midday, suggesting p ossible evidence for stomatal closure. Spatially resolved latent energy flu xes were estimated from the lidar using Monin-Obukhov gradient technique. T he technique was validated from sap-flow flux estimates of transpiration, a nd statistical analysis indicates very good agreement (within +/-15%) with coincident lidar flux estimates. Lidar derived latent energy maps showed th at the riparian zone tended to have the highest fluxes over the site. In ad dition, the spatial variability of 30 min average fluxes were almost as lar ge as the mean values. Geostatistical techniques where used to compute the spatial lag lengths, they were found to be between 100 and 200 m. Determination of such spatially continuous evapotranspiration from such a c omplex site presents watershed managers with an additional tool to quantify the water budgets of riparian plant communities with spatial resolution an d flux accuracy that is compatible with existing hydrologic management tool s. (C) 2000 Elsevier Science B.V. All rights reserved.