Skip Nav

Cryosphere - Fractional Snow Cover

We are currently investigating various remote sensing approaches related to snow, including fractional snow cover (FSC) mapping. FSC mapping overcomes the problem of low spatial resolution of images acquired by global satellites with daily coverage such as Moderate Resolution Imaging Spectroradiometer (MODIS). Various FSC algorithms exist, but a common challenge is mapping snow in forested areas due to the effects of tree canopies.

Using Artificial Neural Networks for FSC mapping is one of the approaches that our group is investigating. A multilayer perceptron is trained with backpropagation learning using MODIS surface reflection as input to the network. Target output is provided by moderate resolution Landsat ETM+ binary snow cover maps. The performance of the method is tested with Landsat snow maps as well. Estimates derived by ANN are compared to the MODIS FSC (MOD10) product.

Geosciences TAMU Logo

Aggies can change the world. Geoscientists lead the way.