Our goal is to retrieve land-surface temperatures (LST) at an accuracy better than 1K from the Moderate Resolution Imaging Spectroradiometer (MODIS) data. MODIS, the key instrument aboard the Terra (EOS AM-1) satellite, will scan a swatch approximately +/- 55 degrees from nadir at an altitude of 705 km and will view the entire Earth's surface every 1-2 days. This instrument was designed and built by Santa Barbara Remote Sensing (SBRS). The Terra satellite was launched on December 18, 1999, from Vandenberg Air Force Base, California. MODIS has 36 bands with instantaneous field-of-view (IFOV) of 250m to 1km, respectively, at nadir, with bands 1-19 and 26 in the visible and near-infrared range, and the remainder in the thermal infrared (3-15 microns).
We developed two LST algorithms, one is the generalized split-window algorithm (Wan and Dozier, 1996), and another is the physics-based day/night LST method (Wan and Li, 1997).
The first LST algorithm uses MODIS data in bands 31 and 32 in the split-window (at 11 and 12 microns, respectively), is suitable for land-cover types such as dense evergreen canopies, lake surfaces, snow, and most soils, that have stable emissivities known to within 0.01. The emissivities in MODIS bands 31 and 32 are inferred from the land-cover types based on TIR BRDF models (Snyder and Wan, 1998; Snyder et al., 1998), that simulate the scene emissivity from the proportions, surface structures, and spectral emissivities of the components in the scene. The examples of spectral emissivities of terrestrial materials can be found in the MODIS UCSB Emissivity Library.
The day/night LST method retrieves land-surface temperature and band emissivities simultaneously from pairs of daytime and nighttime MODIS data in seven TIR bands.
More about these algorithms can be found in the MODIS LST Algorithm-Based Document ( MODIS LST ATBD).
We have validated the MODIS LST algorithms with ground-based and airborne measurement data from our field campaigns [Validation].
[Publications] [Related Links]Return to MODIS LST Group Home Page