In 2002, Manilla had a 6-month drought with one of the most extreme rainfall shortages on record. In nearly fifty years since 1966 there have been no other shortages like it.
I have discussed this drought in two posts: “Profile of an Extreme Drought”, and 3-year trends to August 2004 (An extreme 1-year drought).
[For an update on the longer and more extreme drought of 2018-19, see the note below.]
This post is about the rainfall record only. It compares the percentile values of rainfall totals for groups of months: one month, two months, and so on. The graph shows how the drought began, developed and faded. Other droughts may go through similar stages. I have plotted the pattern of rainfall shortages month by month, showing only even-numbered months. I have plotted them in different colours, with matching “Call-out” labels.
April 2002 (Red): no drought yet.
In April, the monthly rainfall was slightly below average: in the 40th percentile. In this month, nearly all rainfall totals up to the 42-month total were also below average. Only the 6-month total was above average. This set up the conditions for a drought. Notice that rainfall totals for periods longer than 42 months were all well above average. This hardly changed at all in this year. There had been a lot of rain in previous decades.
June 2002 (Orange): 2, 3, and 4-month droughts.
When May rainfall was in the 1st percentile and June rainfall in the 25th percentile, the June 2, 3, and 4-month totals became serious or severe shortages (below the 10th percentile).
August 2002 (Green): 2, 3, 4, 5, 6, and 9-month droughts.
With July rainfall again in the 1st percentile, and August rainfall in the 26th, the drought became extreme. The 4, 5, and 6-month totals were in the 1st percentile: few months had ever had such low figures.
October 2002 (Blue): 3, 4, 5, 6, 9, 12, 15, and 18-month droughts.
September and October both had rainfall in the 18th percentile. That relieved the short-term shortages somewhat, but not those in the medium term. Shortages in the 4, 5, and 9-month totals were in the 1st percentile, but the 6-month total was very much worse. At 76 mm, this 6-month total was the third driest on record, beaten only by August 1888 (43 mm) and September 1888 (69 mm).
December 2002 (Purple): only 9- and 12-month droughts remain.
November rainfall that was near average (40th percentile) and high December rainfall (84th percentile) broke the drought. Only some longer-term effects persisted as severe rainfall shortages in 9- and 12-month totals.
Note added 2019.
Later such graphs in this blog have a logarithmic scale to distinguish the extreme rainfall shortages. Here is the one for the even more extreme drought of February 2019.
climate by surly 
I must say i found this graph quite difficult to get my head around. It took a lot of effort to stop trying to interpret the x axis as a time axis, especially as it is scaled in months! Even harder was to comprehend that each coloured line is not in fact a series, but represents a set of facts about a single date. I think part of my problem was the lines linking each point, which to me reinforced the illusion of sequentiality. I would be interested to see the data represented with time on the x-axis, and a separate series for each drought period, e.g 1 month, 2-month, etc.
Thanks for your helpful comment, Allan.
I’m sorry to have caused confusion. There are many ways to present data on droughts. I had intended to post an introduction on the logic of this particular way; perhaps I still should.
I intended my graph to show a key feature of drought: in any given month, some effects, such as the state of the pasture or crop, have accumulated only for weeks or perhaps a month or two. Others, such as the state of farm dams and streams, may have accumulated over several months. The state of large reservoirs and groundwater storages may depend on the total shortfall of rain over many years. The severity of the drought is a combination of severities on all time scales.
That is why I have chosen to make a line of a given colour represent all the kinds of shortage accumulated at a given date. Then the details of the state of the drought can be traced through from one month to a later one.
I think droughts may typically progress like the 2002 Manilla drought did. That is:
Before the onset (like April 2002), simply a mild shortage of rain totals when counted over periods from a month to a year or two.
At first (like June 2002), severe shortages have accumulated for only a few months.
Next (like August 2002), shortages in 4, 5, and 6-month rainfall totals become not merely “severe” but extreme.
Then (like October 2002), if the drought has continued, rainfall shortages (still accumulating over a period less than a year) become catastrophic.
When more normal rain has fallen (like December 2002), short-term shortages abate, but longer-term shortages persist.
Perhaps you have checked the other two posts on the 2002 drought? They deal with the timing of the drought, and the non-rainfall features. However, they deal with it as a smoothed several-month phenomenon, without considering long-term and short-term features.
My other posts about the history of drought in Manilla do have time on the x-axis. The kind of graph you suggest is this one.
We bought Mulcaster Downs (114 acres on the Manilla River 3km North of Manilla) in 2002. The ground was bone dry and there was very little grass. Only 8 head of cattle instead of the normal 35. A storm in September turned a dirt track into a mud skating rink causing us to install a gravel road which is still there. The storms brought the weeds – tiger pear, devils rope pear, saffron thistles and African Box thorn. From 2004 to 2010 we lived on the farm full time, drinking rainwater and using river water for everything else. We never ran out of rain water but it was close at times. Two or three good storms would refill the tanks. My impression is that significant rain events a Manilla are linked with monsoonal lows bringing Indian Ocean moisture to the Western slopes. Otherwise frontal thunderstorms. I miss our coffee mornings with Neely.
Thanks for this Supeito.
Thanks for a landholder’s perspective on droughts, Phil.