Assessing the suitability of sentinel 2 imagery for the validation efforts of modis fractional snow cover product

This study mainly focuses on investigating the suitability of Sentinel 2 data for the continous validation efforts of fractional snow cover (FSC) product of moderate resolution imaging spectroradiometer (MODIS) over Turkey. In the first stage of the study, Sentinel 2 binary snow cover maps obtained by applying three different binary snow mapping algorithms were tested against in-situ snow depth measurements. In total, 205 Sentinel 2 images taken between November 2017 and April 2018, 286 snow depth measurement from 75 ground observation stations were used to assess the performance of Sentinel 2 binary snow maps. All three binary snow mapping algortihms were in good aggreement with in-situ snow depth data with POD ≥ 0.82, FAR ≈ 0.70 and ACC ≥ 0.80. In the second stage, 200 MODIS FSC images for the 2017- 2018 snow season over Turkey were tested by using reference FSC maps generated from Sentinel 2 binary snow maps. Overalll results indicated that MODIS FSC product, namely, MOD10A1 exhibited quite good performance with RMSE = 0.13 and R = 0.88. When the monthly-based performance metrics were anlysed it was observed that the accuracy of MOD10A1 product degraded in April and this behaviour was mainly attributed to the patchy snow cover during the melting period. Additionally, MOD10A1 was found to exhibit poorer performance over forested areas, whereas its accuracy was high over areas dominated by savannas and mixed agricultural lands.