A Method to Monitor Seasonal Inundation States Using High Resolution ALOS PALSAR ScanSAR Observations in Alaksa

Friday, 18 December 2015
Poster Hall (Moscone South)
Marzieh Azarderakhsh, Fairleigh Dickinson University, Teaneck, NJ, United States, Kyle C McDonald, CUNY City College, Earth and Atmospheric Science, New York, NY, United States, Mahta Moghaddam, University of Southern California, The Ming Hsieh Dept. of Electr. Eng., Los Angeles, CA, United States and Jane Whitcomb, University of Southern California, Electrical Engineering, Los Angeles, CA, United States
Inland water extent is crucial to enhancing our knowledge about the biogeochemical transitions, carbon dynamics and prediction of boreal-arctic ecosystem. However, in high latitute regions inundation is directly affected by the timing of freeze and thaw conditions at the surface. In other words, these two features (i.e. freeze/thaw and inundation analysis) are interconnected and should be considered simultaneously in understanding hydrological and carbon cycles. One of advantages of L-band microwave satellite observation over optical observation is to penetrate to the clouds and vegetation to detect inundated areas. A recent study usese fine-beam mode of PALSAR observation Fine Beam Data (FBD) with resolution of up to 12m as well as ground based land cover using Random Forest Classification method to classify different types of wetlands over Alaska. However, these observations are not frequent and it makes the seasonal and inter-annual studies challenging. We employ PALSAR ScanSAR mode data with more frequent temporal coverage of up to 40 days along with the static map dervied from FBD data to study the timing of the inundation for these wetland classes for 2007- 2010 years period. This study has the advantage of providing freeze-thaw states to detect the timing of active season and then to provide timing of inundated surfaces. Traditional approaches normalize ScanSAR mode data with varying incidence angle to far angle using either histogram method or linear fit methods. We use a unique method that uses the angle dependency of different land cover types to detect inundated surfaces especially the ones that are mixed with vegetation. The seasonal and temporal inundation and freeze/thaw states maps are generated which agree well with previous studies, ancillary data, and ground observations over different land-cover classes. The results of this study benefits hydrological and carbon cycle studies as an important source of carbon uptake in arctic regions. Moreover, the results can be used for validation of newly launched Soil Moisture Active/Passive (SMAP) mission which has devoted an L-band active to predict seasonal freeze/thaw states in global scale.