ABSTRACT
Application of a Split-Window Algorithm to Estimate Land Surface Temperature from NOAA-AVHRR Data

Juan Parra1*, José Sobrino2, Luis Morales3, Giorgio Castellaro4, Juan Uribe4 y Nelba Gaete5
 

The estimation of land surface temperature, in partial to completely cloudless clear days, was estimated using data from the Advanced Very High Resolution Radiometer (AVHRR) sensor aboard the satellites of the series National Oceanic and Atmospheric Administration (NOAA) for application of an algorithm based the Split Window's method. The utilized algorithm presupposes corrections for emissivity and the total atmospheric water vapour, and it was validated with field measurements of surface temperature. To this end an LI-1000 data logger was installed at the Agrometeorology Station dependent of the National Institute of Agricultural Research, Carillanca Regional Research Center Carillanca, Araucanía Region of the Araucanía (38º41' S lat; 72º25' W long; 200 m.o.s.l.). The measurements of temperature were realized at time intervals of one hour. Obtained results showed that the application of the algorithm permits obtaining the land surface temperature with a high grade degree of confidence, having a root mean square error (RMSE) inferior to 1%.

Keywords: emissivity, land surface temperature, water vapour, radiometer, satellite.
1 Universidad de La Frontera, Facultad de Ingeniería Ciencias y Administración, Casilla 54 D, Temuco, Chile. E-mail: jparra@ufro.cl *Autor para correspondencia.
2 Universidad de Valencia, Facultad de Física, Dr. Moliner 50, 46100, Valencia, España.
3 Universidad Tecnológica Metropolitana, Facultad de Ciencias Naturales, Matemáticas y del Medio Ambiente, Casilla 9845, Santiago, Chile.
4 Universidad de Chile, Facultad de Ciencias Agronómicas, Casilla 1004, Santiago, Chile.
5 Instituto de Investigaciones Agropecuarias, Centr