The Use and Evaluation of Split-Window Techniques for NOAA/AVHRR Surface Temperature Extraction over Different Surface Covers: case study (Perak Tengah & Manjong) area, Malaysia

Split window technique has been applied in this study to
extract Land Surface Temperature (Ts) as a step leading to the
retrieval of surface soil moisture and evapotranspiration over
four study areas in Perak Tengah& Manjong, where the areas
were selected according to the difference in surface cover type.
Multiple NOAA/AVHRR images throughout June-August (2009)
were acquired over the study area and processed; measured
ground air temperature data (Ta) for the targeted areas were
collected as well. Normalized Difference Vegetation Index (NDVI)
and land surface Emissivity (ε) were calculated from image visible
bands 1&2. The brightness temperature from bands 4&5 was
then used with ε to derive Ts using three different split window
algorithms. The results showed a good correlation between
satellite estimated Ts and measured Ta especially when Uliveri
algorithm was applied over the agricultural area with R
reaching
87%. The application of the three formulas produced results that
are in agreement with one another with accuracy ranges between
90 – 99%. This ensures the usefulness of using Uliveri, and
Becker over the study area. A reasonable negative relationship
also was found between NDVI and Ts over the uniform vegetation
covers, indicating the applicability/suitability of the so called
“Triangle method”.

The Use and Evaluation of Split-Window
Techniques for NOAA/AVHRR Surface
Temperature Extraction over Different Surface
Covers: case study (Perak Tengah & Manjong)