piloting of the PMS

  • Print

The piloting of the precipitation treatment is dynamic in the annual course, which is based on the 10 driest years in the dataset (1964, 1966, 1980, 1988, 1989, 2001, 2004-2007, all < 600 mm annual precipitation, Fig. 1d). A normalised curve was derived from the cumulative precipitation.

A linear relationship with zero offset is fitted from Julian day 1 to Julian day 365. Two characteristic deviations (minima of the residuals, dry periods) from this linear relationship are approximated with Gaussian functions. The standard deviation is fitted with a 3rd order polynomial function, and is used as a decision tool, to better interpret deviations of the current precipitations in relation to the mean. The reduction of precipitation in the treatment is illustrated on a monthly basis in Fig xxx.  A schematic of the automated program that runs the PMS is given in Fig. xxx.