Did the El Ni

1
Did the El Niño of 2002-03 break the drought in
Colorado or elsewhere in the West?
Klaus Wolter NOAA-CIRES Climate Diagnostics
Center klaus.wolter_at_noaa.gov
With Kudos to Jon Eischeid, Roger Pielke, Sr.,
and Robin Webb
Drought in the Western U.S. 1999 - 2003
Expected precipitation anomalies in 2002-03
Observed precipitation anomalies in 2002-03
Role of individual precipitation events
Preliminary Conclusions
2
Western U.S. Drought from 1999-2002
3
During the winter of 2002-03, the West Coast got
wet first followed by interior states from
Arizona to Montana that recovered at least
partially from a severe multi-year drought. This
fall, drought has reaffirmed its grip on states
like New Mexico and Montana.
Where did the West recover from drought
conditions in 2002-03?
4
Stateof ENSO mid-2002 into early 2003 based on
MEI a moderately strong El Niño event
MEI ranks in JA02 11, SO 13, ND 10, JF03 9,
MA 20
5
Percentage of Significant Departures During El
Niño 18.0 10 cases 1957, 63, 65, 72, 82, 86,
87, 91, 94, 97 If the MEI were used, 1977 would
replace 1963.
6
Percentage of Significant Departures During El
Niño 20.5 10 cases 1957-58, 65-66,
68-69, 72-73, 82-83, 86-87, 87-88, 91-92, 94-95,
97-98 If the MEI were used, 1977-78 would replace
1968-9.
7
Percentage of Significant Departures During El
Niño 6.1 10 cases 1958, 66, 69, 73, 83, 87,
88, 92, 95, 98 If the MEI were used, 1978 would
replace 1964.
8
Percentage of significant composite precipitation
anomalies in the Western U.S. vs. three different
ENSO indices Niño 3.4, Southern Oscillation
Index (SOI), and Multivariate ENSO Index (MEI).
Highest values in each season are underlined.
Niño 3.4, the official measure of ENSO, is not
optimal for anticipating regional impacts.
9
Recent precipitation trends for the U.S. (left)
SON, (bottom left) DJF, and (bottom right) MAM.
Correlations above 0.3 can be considered
significant. According to this analysis, Montana
has had the most pervasive drying trend in the
West over the last 30 years. This comes on top of
the expected El Niño signal, or does it express
the preference for El Niño conditions during that
period? Compared to recent temperature trends,
most of the precipitation trends appear
insignificant.
10
How representative are climate divisions?
Biggest problems exist with precipitation in the
interior West, and Climate Divisions do NOT
include SNOTEL data.
11
New Climate Divisions
Official CDs for Colorado (left) and Experimental
CDs (right) based on multivariate statistical
analysis of climate data that also include SNOTEL
data. Such new CDs are being derived for the
entire U.S.A. Let us focus on 1 (red), 3 (black),
6 (purple), and 8 (green) for this El Niño
impact study.
12
El Niño in Colorado

• Seasonal precipitation amounts (in percentiles
  from 1950-99) for 10 El Niño cases 1957-8, 65-6,
  72-3, 77-8, 82-3, 86-7, 87-8, 91-2, 94-5, and
  97-8, based on the MEI. If 8 or more cases out of
  10 reside above or below the median, the
  distribution is shifted significantly. If 4 or
  more cases reside in the upper (80) or lower
  (20) quintile, there is only a 10 chance that
  this result is by accident. The 2002-3 El Niño
  event ended up ranked 10th, pushing 1977-78 into
  11th place.

Wet
Wet
Wet
Dry
WET
13
Observed precipitation rankings in 2002-3
During the fall season of 2002, increased western
U.S. precipitation matches expectations quite
well, in particular over Utah, western Colorado
and northern New Mexico. During the following
winter, the overall El Niño foot print remains
visible, with Washington receiving more than
expected and AZ/NM less. Dry conditions over the
central high plains are the least consistent with
El Niño expectations.
14
Observed precipitation rankings in 2002-3
During the spring of 2003, California continues
to receive good moisture consistent with
lingering El Niño conditions. The Pacific
Northwest into Montana appear lucky to receive
more moisture than expected. Over the interior
southwest, northeast Colorado hit the jackpot
due to its March snowstorm, while AZ/NM continue
to under perform. A more detailed look at
Colorado follows below, revealing that the spring
moisture bonanza is actually very consistent with
MEI-based expectations.
15
What a difference a year makes, The Big Picture
September 2001 through August 2002 maximizes the
impact of the 12-month running drought in the
southwestern U.S. Compared to the record-setting
recovery in the eastern U.S., our return to
near-normal in 2002-03 appears rather bland. It
was most welcome nevertheless, and, together with
prudent water management, it allowed for at least
a modest recovery in the western water supply
situation.
16
What a difference a year makes, Colorado-style
WY 2002 was the driest on record in much of
southern Colorado, and close to that in the
north. WY 2003 ended up close to or wetter than
normal along the Front Range and in most of
northern Colorado. Dry pockets persist in the
southeastern plains and the San Luis Valley. The
biggest positive anomaly (around Summit County)
overlaps with massive cloud-seeding efforts last
winter
17
Impact of the 2002-3 El Niño in Colorado
18
Did the 2002-3 seasonal totals in Colorado match
expectations (high hopes)?

• Using the simplest possible classification of the
  data (above or below the median), lets compare
  the expected WET/DRY/WET sequence in most of
  Colorado (expect for the Arkansas Valley that
  tends to get WET in all 3 seasons from fall to
  spring) with preliminary observations
• The northern two climate divisions (1 and 8)
  received WET/DRY/WET, while the south (3 and 6)
  ended up DRY during the spring. This means that
  10 out of 12 forecasts were correct in their
  sign, and two were not. How unusual is that?
• If you classify each season since 1950 into above
  and below median precipitation totals, and
  compare against the WET/DRY/WET scenario (WWW for
  region 6), there are only two cases (out of 42)
  in which there is a match this good (10 out of
  12) during non-El Niño years (1984-5 and 98-99),
  while there are three such cases (out of 10)
  during El Niño years. IOW, the a priori odds for
  such a good match were 30, given the El Niño
  setup, and only 5 during the other years.
• Another way of quantifying this the median
  number of matches of this type is only FIVE out
  of 12 during non-El Niño years, and between EIGHT
  and NINE during El Niño years. IOW, 2002-3 did
  not only match, but exceed expectations in
  Colorado.

19
The importance of individual events

• Big storms battered California on and off during
  the winter and spring of 2002-3. One can argue
  that the siege of storms in mid-December made all
  the difference for the Sierra Nevada snowpack,
  while the biggest rain in L.A. came right in the
  middle of the annual AMS meeting. Higher odds
  for big storms are typical of El Niño, but this
  one broke some records!

20
Snowstorm of the century?!
Speaking of records, the mid-March 2003 storm was
easily one of the biggest three snowstorms of the
last century in Colorado, rivaling or exceeding
totals last seen in April 1921 (24h U.S. record
at Silver Lake, Boulder County), and December
1913 (the analog case in terms of the synoptic
set-up?). Liquid totals reached 7-9 H2O at
numerous COOP and SNOTEL sites along the Front
Range, or 20-30 of the average annual
precipitation at these locales.
21
The role of individual storms in Colorado

• The March 17-20 snow storm turned a lackluster
  winter around, and refilled Front Range
  reservoirs. It is easily the biggest single snow
  storm (in terms of SWE) in Front Range SNOTEL
  history (since 1978).
• A near-record breaking snow storm in late April
  hit West Slope ski resorts right after many of
  them had closed down for the season. As can be
  seen in this SNOTEL diagram for Vail, this one
  contributed only about 10 to the annual total,
  while the March storm added 25 to Lake Eldoras
  totals.

22
Preliminary Conclusions

• As discussed last year, multi-year drought
  conditions in the West reached record proportions
  by late summer 2002, matching or exceeding the
  severity of one of the worst droughts on record
  along the eastern seaboard.
• Recovery since then has been not nearly as
  decisive for the West as it has been for the
  East. In fact, if it had not been for a few key
  events (that did not produce for everyone), the
  drought would have continued virtually unabated.
• The El Niño of 2002-03 provided for a rare
  opportunity to combine the typically skilful
  forecasts during such events with increased
  public interest in the West. Much was riding on
  this one!

23
Preliminary Conclusions

• Probably not unrelated to its only moderate size,
  the El Niño of 2002-3 did not provide for
  universal relief in the southwestern U.S.
  Nevertheless, beneficial moisture during the fall
  and late winter season in particular helped turn
  things around from California to Colorado.
• Of course, this preliminary assessment does not
  prove that the observed precipitation anomalies
  were related to El Niño. However, I hope I was
  able to point out areas of agreement and
  disagreement.
• While the tier of states from Washington to
  Montana fared better than expected, unchanged or
  worsened drought conditions in New Mexico and
  parts of Arizona attest to the worst
  disappointments of this El Niño year in the West.