Landsat ETM+/TM/MSS

TM/ETM+

To be added: 1)water mask should be a lake ice output 2)other land mask specificity 3)fVEG by merging multiple TMSCAG runs (write matlab function) 4)use emissivity for each landcover type to determine actual temperature instead of brightness temperature 5)post-processing with option for subset to be saved in specific location 6)populate database Karl database with file locations for Landsat 5 and 7 7)pre-processing function that skips when the output already exists

Output 1: .mat -

   a) fsca raw - values are 0 to 1 
   b) fsca - values are 0 and 0.15 to 1. Product has had the following masks applied
   c) masks 7 bands:
               1) edge mask - to deal with bands having different spatial extents
                  a) NLDAS processing did a similar thing. All output from USGS is currently not trimming the edges. See current [1]
               2) "other land"  - always applied for 0 and 127 - optionally calculated and applied: herbaceous wetlands or other classes
               3) shadow mask (all bands < 0.10)
               4) temperature mask (if fsca>0.15, emissivity is 0.98, otherwise ASTER emissivity set used
               5) water mask (all bands < 0.10 & NLCD indicates water) - this is a subset of 2...inverse is snow covered lake
               6) optional cloud mask (calculated for scene only applied w/ flag - rarely used)
               7) fsca < 0.15
               
   i) snow covered lake (NLCD=Water,fSCA>0.15
   i) surface reflectance (6 bands)
   k) brightness temperature
   j) RGB - bands 5,4,2 (stretch version of surface reflectance)
   l) ndsi - (2-5)/(2+5)
   m) 3 band saturation mask indicating combination of saturation that resulted in different TMSCAG "runs"

MSS/TM crossover 1985-1987

[[2]]