SeasonLook is a tool used to determine growing seasons in agricultural areas all over the world. The information of SeasonLook also aids crop classifications, as crops differ in growth start and duration.
Traditionally, the Fourier analysis is applied to extract information on the growing seasons. However, the disadvantage of the Fourier analysis is that growing seasons need to have fixed durations, e.g. 12, 6 and 3 months, which hampers the detection of, for example, a 5 month growing season.
SeasonLook is based on NDVI time series. These time series are fitted to a growing curve that is specifically designed to detect growing seasons. A succesful fit provides a number of parameters that allow distinction of start and duration of growing seasons. The growing curve is also capable of detecting sequential growing seasons within a year. The result is that yield predictions are no longer based on constant values for the entire study area, but that start and duration of the growing season are determined on pixel level.
Yield predictions and crop classifications are improved greatlywith the use of SeasonLook. Accurate and reliable information on the start and duration of growing seasons is of great importance, especially for yield prediction.