The relationships between vegetation, irrigation, precipitation, and atmospheric water vapor were investigated using a vegetation index (NDVI) and vertical profiles of atmospheric water vapor derived from the Moderate Resolution Imaging Spectrometer (MODIS). In order to reduce noise in the water vapor and NDVI data, Empirical Orthogonal Functions (EOFs) were computed and correlations were tested between the main principal components of NDVI and water vapor in each of 19 atmospheric levels. In 13 of 19 vertical layers, land use was significantly correlated with water vapor content. These effects were largest near the surface, up to 850 hPa (1.6 km above sea level), and weak but detectable between 780 and 150 hPa. Meteorological data support the correlation between NDVI and water vapor as seen in positive vapor pressure anomalies from 1940 to 2001 in irrigated areas. Cross-correlograms identified a spatial lag in the correlations between vegetation and water vapor at medium altitudes of 70 km. Movement may be related to prevailing wind patterns or changes in local convection as a result of increased water vapor flux.
