I'm simulating a long-term fallow irrigation, so the upper boundary should represents irrigation for 4 short periods and 'atmosphere' conditions in-between these periods and in the following months.
Variable flux could suite, but then, it seems i loose the evaporation component ??