This graph is one of a series. All are copied here.

Twenty-five months to May 2019.

The above graph is described here.

Twenty-five months to April 2019.

The above graph is described here.

Twenty-five months to March 2019.

The above graph is described here.

Twenty-two months to January 2019.

The above graph is described here.

Nineteen months to October 2018.

The above graph is described here.

Drift of percentile values 

Percentile values are a stable measure of relative rainfall shortage when the sample is large: over 1600 monthly observations in this case. However, percentile values must be re-calculated at each new record low value of rainfall. The effect can be seen in the series of contour graphs above.

Driest 15-month totals

The October graph above shows that a new low record for the 15-month rainfall total was set in September 2018. It was 400mm (beating the previous record of 404 mm set in 1912). Being the driest of 1600 such rainfall observations, the percentile value for September 2018 was 0.04%. It plots in the black zone below the 0.1th percentile, containing only the driest and second-driest observations.
The April graph includes a new 15-month record value (397 mm in April 2019) leaving September 2018, as the second-driest, still in the black zone.
The May graph shows that this record was broken yet again, to 388 mm.The September 2018 15-month total was now only third-driest. Its percentile value of 0.12% no longer qualified it for inclusion in the black zone below the 0.1th percentile.

Twenty-two months to May 2003.

This is the 2002 drought, for comparison with the 2018-19 drought. The above graph is described here.

Twenty-two months to November 1995.

This is the 1994 drought, for comparison with the 2002 and 2018-19 droughts. The above graph has not yet been described.

---

Method

This kind of graph simply displays the time sequence, month by month, of rainfall shortages that I have displayed on line graphs prepared for each month. In the post for a recent line graph (April 2019) I have described my method of analysis and its limitations.