Ignorance scores as a function of the number of observations per grid cell. The curves for the half-ignorance algorithm (red lines) are calculated for three values of O_0.5 = 1, 5, and 10 (i.e. enough number of observations to reduce the ignorance score by half).