Home
The PHASE model enables the automatic derivation of Germany-wide raster data representing beginning phenological development stages (= phases) of natural and cultivated plants. The wiki provides information about the R functions' inputs, outputs and parameters. The functions are implemented within the programming environment of R. Figure 1 illustrates the relations of the functions, which are summarized in the wrapper file fPhaseR. After activating all necessary packages (function fLoadAndInstall), the procedure starts with the download of all crop-specific phenological observations (function fDownloadPhenObs). Then, the downloaded observations are imported (function fImportPhenObs) and coupled with phenological stations (function fPhaseStation) as well as daily temperatures (function fLoadTemp). For each phase-specific observation, a year-specific temperature sum is calculated (fTeffSum). By applying an optimization procedure, different temperature sum quantiles are derived. They are used for the estimation of so called critical DOY variants for each station, phase and year. The resulting DOY distributions are statistically filtered by different standard deviation intervals (F) and compared with observed DOY (function fDoyCrit). Excecuting function fFilterAssessment leads to observations, which are optimal regarding both observation number and comparison metric. The final optimal variant (function fOptShp) goes into an interpolation algorithm (fPhaseInterpolation), which results in a year- and phase-specific raster data set.
