Random Forest fused MODIS and Landsat snow cover from spectral mixture analysis in the Sierra Nevada, USA

This data is snow cover fraction from the Snow Covered Area and Grain Size (SCAG) model for Landsat OLI and Terra MODIS and well as a 2-stage random forest model to fuse the 2 datasets for improved temporal/spatial resolution. There are 170 scenes in 2001 to 2012. It was used in the a publication for Remote Sensing of the Environment titled: Multi-sensor fusion using random forests for daily fractional snow cover at 30 m, doi: to be assigned.

Inputs: [YYYYMMDD is year month day of month, $num is 5 or 7 for Landsat platform, $sens is sensor TM or ETM+]

Landsat.zip:

Snow cover from Landsat: SSN.p042r034_YYYYMMDD.Landsat$num-$sens.canopyadjusted_mask.v01.tif 

MODIS.zip:

Snow cover from MODIS: SSN.SN_W$YYYYMMDD_$YYYYMMDD.Terra-MODIS.snow_cover_percent.v01.tif

Predictors.zip 

Static predictors (see RSE publication Table 2): SouthernSierraNevada*.tif [* here is the variable name]

Outputs [ [YYYYMMDD is year month day of month]

ProbabilityNot0Not100.zip

SSN.prob.btwn.YYYYMMDD.v3.tif - from classification random forest, probability of being between 0 and 100

Probability100fSCA

SSN.pro.hundred.YYYYMMDD.v3.tif - from classification random forest, probability of being 100

RegressionResult.zip

SSN.regression.YYYYMMDD.v3.tif - from prediction random forest

Final_Downscaled.zip

SSN.downscaled.YYYYMMDD.v3.3e+05.tif - final product (combination of classification and prediction)