As a follow-up to a study in Geophysical Research Letters, a journal of the American Geophysical Union, I decided to research the data for Houston’s official reporting site, George Bush Intercontinental Airport (KIAH) only. Data there goes back to 1969.
Based on precipitation data from KIAH, the data since 1969 suggests that half of Houston’s annual rainfall occurs on the wettest 12 days of the year. That is, if you sorted daily rainfall amounts in a given calendar year from greatest to least and added up those greatest, or wettest, days in a cumulative sum, then, on average, half of that year’s rainfall accumulated as a result of the top 12 wettest days. Well, 11.7 days, to be a bit more precise.
Since 1969, Houston’s KIAH receives rain (at least a trace amount) on 147 days of the year, as showing in Figure 1 below:
A word on averages and normals and such. Using precipitation data from KIAH beginning with the first full year of data (1970) through 2017, the average annual rainfall total at KIAH is 49.84″. If you look at the climate data at the NWS Houston website, you will see that they use an average annual “normal” as 49.77″. Why the discrepancy? Because the NWS uses data from the 30-year period 1981-2010 to calculate their normal. For my purposes here, I decided to use the average data over the entire period 1970-2017.
Looking back at Figure 1, we notice some interesting findings. For example, on average, Houston receives at least one day a year with a daily rainfall total in excess of 5″. Actually, the average works out to be 0.5 days in excess of 5″, but I rounded up for a cleaner graph. If we look at the data in Figure 1 in tabular format, we get the following result:
Figure 2 is an attempt to find the number of days it takes to accumulate half, or 50%, of our annual average.
- Column A is an arbitrary set of daily precipitation ranges.
- Column B shows the number of times, on average, that a certain range of rainfall has been reported at KIAH. For example, in 18,099 daily reports since 1969, KIAH has reported greater than 5″ of rainfall on 26 occasions. That works out to about 0.5 events each year, as shown in column B.
- Column C is a factor that is, essentially, a mid-point figure of the range of figures in Column A. For example, the 1.00-1.50 range in column A corresponds to a 1.25″ factor (the halfway mark between 1.00 and 1.50) in column C.
- Column D multiplies the results in Column B and Column C to calculate a rainfall amount associated with each range of rainfall totals.
- Column E is a summation of those calculated rainfall amounts in Column D. It is a running total that ends up with a total rainfall amount of 50.50″. While this figure is slightly larger than the average rainfall (49.84″) and the normal used by the NWS (49.77″), it is within 1.5% of each of those figures. And since this is a calculation based on an arbitrary factor, the summation of 50.50″ is reasonably close in approximation to the average and normal.
- Column F is a summation of those calculated events in Column B. It is a running total that helps show the number of days that it takes to accumulate a certain amount of rainfall in a given year, as shown in Column E. For example, in a typical year, a total of 15.35″ (see column E) will accumulate on the wettest 4.7 days of the year.
Assuming average rainfall is 50.50″ as shown in Figure 2, then half of that (25.25″) falls somewhere in between 8.1 and 14.6 days. A graph of this relationship shows this more clearly:
Figure 3 is a cumulative plot of rainfall using the table in Figure 2. Notice the 50% line at 25.25″ falls somewhere between 8.1 and 14.6 days. As it turns out, that simple average is 11.7 days, as shown in Figure 4:
The simple average of all years shown in Figure 4 is 11.7 days. The interesting thing about the data is that the tendency for lesser days (8 days, as an example) has increased over the decades. In the 1970s, there was only one year (1976) where it only took 8 days to reach half of that year’s annual rainfall total. Same for the 1980s: just one occurrence (7 days in 1989). Not once in the 1990s did it ever occur on fewer than 11 days. But, by the 2000s, half of the year’s rainfall fell on 8 or fewer days three times (2002, 2008, & 2010)! And, in the 2010s, this has again happened twice so far (2011 & 2017)!
Conclusion from this simple look at the data: more extreme and heavy rainfall events have been occurring with greater frequency in recent decades.
- When considering the number of days as combined with a cumulative total of rainfall in a given year, I based my decision to cut-off at the 50% mark once cumulative rainfall exceeded 50% of the yearly total. In other words, I did not use an average or normal figure. I used that year’s total rainfall. For example, in 1970, 48.19″ of rain fell at KIAH. Half of that is 24.10″. A sum, or accumulation, of the total rainfall from the greatest daily amounts to the least mounts from 1970, we cross the 24.10″ mark at day 10, as shown below: