Methodology of spatialization of energy balance components using MODIS and ERA Interim data

Abstract

This work aims at proposing a methodology to spatialize the Energy Balance (EB) components from the
combined use of MODIS sensor products and ERA-Interim reanalysis data, and to evaluate the performance of the OSEB and
METRIC models for estimating EB components in the predominantly humid subtropical climate of Rio Grande do Sul, Brazil.
The research was conducted for 3 consecutive years from 2009 to 2011. The temporal variation patterns of the EB components
were analyzed by comparing the values estimated by the models against the reference measurements of a micrometeorological
tower (Sulflux network). For the spatial variability analysis, reference evapotranspiration (ETo) was calculated at a local scale
(data from Weather stations of the National Institute of Meteorology) and a regional scale (ERA-Interim reanalysis data) while
the consistency between these estimates was also verified. Afterward, using only the pixels in which the sensible heat flux
represented more than 50% of the radiation balance, correlation maps between ETo ERA-Interim and latent heat fluxes were
constructed. The estimated BE components are consistent regarding both, the order of magnitude and variation throughout the
year, when compared to the reference measurements of the micrometeorological tower. The spatial variability analyses also
reveal satisfactory results for both models, however, the METRIC model results showed greater coherence since it provided a
greater number of dates with consistent results, thus being indicated as the most suitable model for building the time series.