In arid and semi-arid regions, as well as in parts of the tropics where wet and dry seasons alternate, periodic moisture shortages represent one of the most critical constraints to rainfed farming. Rainfall variability exacerbates this, influencing the onset and end of the growing season, or introducing dry spells that fragment the growing period. Early studies from Tanzania, Ethiopia, Namibia (e.g., Publications  9, 10, 11, 14, 15, 28, 30), and other countries demonstrated that growing period durations can vary significantly, ranging from less than 10% to over 40%, depending on location and year.

High rainfall variability, which affects the onset, end, and duration of the growing period, means that conclusions about crop suitability cannot rely on average rainfall. Instead, a probability-based assessment is necessary. This led to the development of the concept of the ‘dependable growing period,’ which refers to the growing period likely to be exceeded in 4 out of 5 years. The concept can also be applied with different risk levels, such as 3 out of 4 years for higher risk for crops, or 9 out of 10 years for lower risk.
The dependable growing period is calculated through a simple frequency analysis of growing periods over a sufficiently long time series, ideally spanning 30 years or more. Each growing period is determined following the procedure detailed in this document.

NOTE
Looking back with a touch of nostalgia at how far computer technology has come, the first water balance calculations and statistical analyses for assessing growing period variability were conducted on an HP 9825 microcomputer, now a museum piece. I wrote custom software in Hewlett-Packard’s cryptic High-level Programming Language (HPL) specifically for this purpose. These early efforts formed the basis for this set of maps of Tanzania prepared in 1983.