Application of Passive Microwave Data in Estimating Freeze-Thaw Dates of a Small Lake

摘要

Our method uses coarse resolution passive microwave data to estimate freeze-thaw dates of soil in the region surrounding the lake. Using the soil freeze-thaw dates we then determine a high probability 'window' of days during which the lake freeze-thaw dates may occur. This paper mainly focuses on using passive microwave data to estimate this high probability window of days for the freeze and the thaw seasons of a lake. The importance of this project in our broader research goal of estimating freeze-thaw dates for small lakes can be explained as follows. Given a classifier that can classify SAR images of small lakes into 'ice' and 'no-ice' images (a parallel research project that we are involved in), we demonstrate that using passive microwave data to estimate the 'window' will improve the practical use of the classifier in detecting freeze-thaw dates of the lake, by reducing the number of datasets required (using only the datasets that fall within the window). By reducing the number of SAR datasets required for the algorithm we reduce both financial and computational cost of the classification algorithm. This in turn will make the method more affordable and improve the usability of the classification algorithm to keep track of lake freeze-thaw dates for a larger number of lakes. The passive microwave data collected, results obtained till date and related literature will be discussed in this paper.