Rainfall in Eugene Oregon

Документ взят из кэша поисковой машины. Адрес оригинального документа : http://zebu.uoregon.edu/raining.html
Дата изменения: Thu Feb 10 20:49:07 2000
Дата индексирования: Mon Oct 1 20:07:56 2012
Кодировка:
Поисковые слова: южная атлантическая аномалия

The Problem:

In 1995 Eugene set an annual rainfall record by recording, officially, 65.56 inches of rain.

In 1996 Eugene broke this record by a large amount by recording 77.17 inches of rain. ( But see below )

A Statistical Look

The mean annual rainfall for Eugene from the 25 year period 1970-1994 was 51.5 inches with a dispersion of 8 inches. This 25 year period included several wet years (1971, 1982,83,84) and thus has a mean which is somewhat higher than if another 25 year period were selected. A time sliced look at the mean annual rainfall can be found right here . However, I suspect that most people would be surprised that, over the last 25 years, the average indeed is as high as 50 inches. Below I list the 3 wettest and driest years during this period just to give you an idea of the variation:

Wettest Years:

1983: 64.01

1971: 60.67

1982: 60.02

Driest Years:

1985: 33.83

1976: 34.78

1978: 39.50

If we use standard statistical techniques and ask what is the probability that it would rain 65.56 inches in one year (the 1995 total) here is what you get:

As stated earlier, the mean annual rainfall and its dispersion is 51.5 +/- 8 inches. To a high degree of approximation, rainfall data are well fit by a bell curve. This fit produces the quoted mean and dispersion.

Compared to this distribution, a total of 65.56 inches is 14.06 inches above the mean value or 14.06/8 = 1.75 standard deviations above the mean. This is not a rare event and has approximately a 3% chance of happening. Indeed the driest year in this 25 year period is 17.67 inches or 2.2 standard deviations below the mean. This is a significantly rarer occurrence which is expected only 1.5% of the time. However, in either case, over a 100 year period, these wet and dry rainfall totals would be expected to occur.

But what about the 1996 total of 77.17 inches? Well this is 25.67/8 = 3.2 dispersion units above the mean. This is extremely rare and would happen, on average once every 1000 years!

The first conclusion from this is simple: In no way can 77.17 inches be considered as a statistical fluctuation which occurs within the context of the normal weather patterns that produce precipitation in Eugene. Last years rainfall record of 65.56 inches could be considered in this way. That total broke the previous record by about 1.5 inches. However, this year broke last years record by a full foot!

The primary reason why we exceeded last years record by a foot of rain was due to three significant rain events. But we can examine this on a month-by-month basis:

January: 14.74 inches of rain - this is substantially above normal but not a record. The record was set, of course, in 1995 with 15.86 inches which broke the January 1990 record of 15.09 inches. So 1996, like 1995 got off to a wet start.
February: Everyone remembers those floods . We got 16.85 inches of rain that month which easily set the record. Combined Jan/Feb rainfall is hence 31 inches and we are well on the way to set a new rainfall record in 1996.
March: 2.45 inches of rain - relatively dry
April: 4.56 inches --> fairly typical
May: 3.5 inches --> above normal but not unusual
June: 0.39 inches
July: 0.54 inches
August: 0.05 inches
September: 2.5 inches --> this is a bit above normal
So now we are at the end of September with a total rainfall for the year of 45 inches of rain. If the remainder of year has normal rainfall amounts we project a final rainfall for the year of about 65 inches, which will be close to last years record. Of course, this total is entirely driven by the huge combined Jan/Feb rainfall.
October: 4.50 inches --> not particularly wet. This now brings the yearly total to 50 inches.
November/December --> yep 27 inches of combined rainfall driven mostly by large rain events, associated with tropical moisture feeding the jet stream, on Nov 19-20 and Dec 26-31.

Another way to look at this is to say we got approximately 58 inches of rain in Jan,Feb,Oct,Dec. So many large rain events in one year is what is unprecedented. Is this statistical fluctuation, part of a return to a wet cycle (where in the data can you identify events like these in the past?) or a signal of climate change? I will leave this as a question for now and will return to adding to this page later.

Corrections to the 1996 data

In 1996 a new automated rain gauge was installed at the Eugene airport. During heavy rains, apparently, the rain gauge does not empty itself fast enough to properly record the data. Below are the official NWS corrected rain totals for the year 1996. Although these are now "official", I go on record here by saying there is no way I believe that Eugene, Oregon received 101 inches of rain in 1996.

Here are the approximate "corrected" data, month by month.

January: Old = 14.74; corrected = 14.48
February: Old = 16.85; corrected = 17.08
March: Old = 2.45; corrected = same
April: Old = 4.56; corrected = 5.15
May: Old =3.50; corrected = 3.50
June: Old=0.39; corrected = same
July: Old = 0.54; corrected = same
August: Old = 0.05; corrected = same
September: Old = 2.5; corrected = 3.1
October: Old = 4.5; corrected = 7.1
November: Old = 12.10; corrected = 18.42
December: Old = 14.95; corrected = 26.28(!!!)

The official total is 101.93 and yes the corrected values don't quite add up to that. Clearly the correction is largest (I would say extreme) for the two wettest months November and December.

As a comparison, for Salem, 60 miles to the North, the official 1996 rainfall total, which set a new record, was 66.71 inches. Does anyone really believe that it rained 35 inches more in Eugene than in Salem? - I don't.

Bottom line however, 101.93 is now our official rainfall record and its gonna take a helluva greenhouse effect to break this!

Finally, for those of you interested in decadal weather cycles, go here