Paper No. 10
Presentation Time: 8:00 AM-12:00 PM
APPLICATION OF MODIS ENHANCED VEGETATION INDEX AND LAND SURFACE TEMPERATURES TO PREDICT RIPARIAN EVAPOTRANSPIRATION ACROSS THE SEMI-ARID SOUTHWEST, USING MICROMETEOROLOGICAL TOWER DATA FROM THE MIDDLE RIO GRANDE, SAN PEDRO RIVER, AND THE LOWER COLORADO RIVER
Near-real time, validated estimates of riparian evapotranspiration (ET) using remote sensing and ground methods could provide improved estimates of riparian water use, which would lead to more accurate water balance models. Ground ET measurements (eddy covariance method on field towers), were correlated with micrometeorological variables, Enhanced Vegetation Index (EVI) and Land Surface Temperature (LST) data from the Moderate Resolution Imaging Spectrometer (MODIS) sensor. Tower data was available on the Middle Rio Grande (four towers), San Pedro River (four towers), and Lower Colorado River (two towers). Middle Rio Grande sites include two saltcedar (Tamarix ramosissima) and two Rio Grande cottonwood (Populus deltoides ssp. wislizennii) dominated stands. San Pedro River sites include two mesquites (Prosopis glandulosa), one Sacaton grass (Sporobolus wrightii), and one desert scrubland of mixed species. Lower Colorado River sites include one cottonwood (Populus fremontii) and one willow (Salix gooddingii). EVI was more closely correlated (r=0.76, P<0.001) with ET than the Normalized Difference Vegetation Index (r=0.68, P<0.001). While cottonwood ET rates were higher than saltcedar rates, these differences were reflected in EVI values, and the combined data set had r values close to those for individual species. Canopy air temperature was the meteorological variable that was most closely correlated with ET (r=0.82, P<0.001). MODIS Land Surface Temperatures (LST) (5 km resolution) were also closely correlated with ET (r=0.74, P<0.001) and could serve as a surrogate for canopy air temperature. A multivariate regression equation for predicting ET from EVI and canopy air temperature had a coefficient of determination (r2) of 0.84 (P<0.001) across sites, species and years. The finding that ET predictions did not require species-specific equations is significant, since these are mixed vegetation zones that cannot be easily mapped at the species level. Ancillary data such as vegetation maps, to create validated, 16-day estimates of ET along the southwestern rivers (where there are towers) can be used to create new estimates of riparian ET to adjust water balance flux models for selected river stretches, which will enhance our ability to manage it for multiple uses.