Using available time series of Passive and Active Microwave to develop SMAP Freeze/Thaw algorithms adapted for the canadian subarctic

Seasonal terrestrial Freeze/Thaw cycle in Northern Quebec Tundra (Nunavik) determined and evaluated with Passive and Active Microwave Observations. SMOS time series data were analyzed to examine seasonal variations of soil freezing, and to assess the impact of snow cover and land cover on freeze-thaw cycle. Further, the soil freezing maps derived from SMOS observations compared to microwave active images in the region near Umiujaq and Field survey data. The objective is to develop algorithms to follow the seasonal cycle of freezing and thawing of the soil in the Tundra and Boreal forest. Field data shows that freezing and thawing dates vary much spatially at the local scale in the Boreal Forest and the Tundra. Therefore, the field validation of the F/T state maps at the regional scale will be very important. Agreement Factor derived from comparison of SMOS FT maps with daily in-situ data shows low values which does not seems to be acceptable. New parameters such as lake and pond as well as vegetation type and height present on surface have to be introduced in the algorithm to obtain more realistic estimations.