LAUSR

The paper describes experience of analysis of the high time resolution satellite image series aimed on detection of growing season framing dates. Satellite data use for growing season detection were discovered as observation gap filling technique in areas with sparse network of meteorological stations. We discovered the possibilities of growing season detection using annual series of 1-day Moderate Resolution Imaging Spectroradiometer (MODIS) composites instead of the ground-based monitoring of surface air temperature. Particularly, the study was aimed on comparison of growing season detection results based on ground and satellite observations. We attracted Google Earth Engine platform to process initial (Terra/MODIS) satellite data series, derived map series of Normalized Difference Water Index (NDWI) with a 1-day time resolution, produced annual graphs of the index in the reference points of meteorological stations, detected growing season framing dates at the reference points for the period of 2000–2017, and compared detected growing seasons with their estimations made upon the analysis of surface air temperature dynamics at the same meteorological stations. Involvement of satellite data gives the possibility to monitor growing seasons not only at the points of meteorological stations but through the whole studied area highlighting heterogeneity in spatial distribution in growing season dates and regional differences of growing season dynamics. Currently, we concluded weak quality of growing season detection based upon the 1-day NDWI series, due to the noise presented on annual graphs of NDWI. Mean observed accuracy of detection in comparison with the ground observations of surface air temperature was ~ 11.3 days, while median accuracy value was ~ 10.5.