The Colorado Surface Water Supply Index (SWSI) Performance Since 1981

1
The Colorado Surface Water Supply Index (SWSI)
Performance Since 1981

• Nolan Doesken
• Colorado State Climatologist
• Colorado Climate Center
• Colorado State University

2
So whose idea was this anyway???
3
(No Transcript)
4
(No Transcript)
5
Background

• WATF (Water Availability Task Force) was created
  within the Drought Response Plan to objectively
  assess water supply information
• Initially used the Palmer Drought Severity Index
  to trigger drought response actions.
• Soon discovered that the Palmer Index was too
  slow to respond to changes and seemed to perform
  poorly in seasonal snowpack environments
• SWSI was later developed to aid in water
  availability monitoring in mountainous areas.

6
SWSI Usage
7
SWSI Introduction

• Purpose to be an indicator of basin-wide water
  availability for the mountain water dependant
  sector, be predictive and permit comparison of
  water supply conditions between basins to assess
  relative drought severity (Shafer and Dezman,
  1982)
• SWSI seemed to be more robust and responsive to
  changing conditions than the Palmer index for the
  mountainous Western U.S. It accounts for known
  components of water supply -- reservoir storage,
  streamflow, precipitation and snowpack. SWSI
  seemed more understandable than the black box
  PDSI model

8
SWSI Calculation

• SWSI (aPsnow bPprec c Pstrm dPresv -50) /
  12
• a snowpack weighting factor
• b precipitation weighting factor
• c streamflow weighting factor
• d reservoir weighting factor
• abcd 1
• P probability of non-exceedance (percent) for
  component i
• snow, prec, strm and resv are the snowpack,
  precipitation, streamflow and reservoir
  components, respectively.
• Subtraction of 50 centers the distribution around
  zero
• Division by 12 bounds the values between -4.17
  and 4.17 to mimic range of Palmer Drought Index.

(Shafer and Dezman, 1982)
9
SWSI Calculation

• Normalization was accomplished by using
  non-exceedance probabilities -- allowing for
  comparison across varying climates.
• Requires representative long-term data or
  accurate estimates appropriately representing
  both wet and dry extremes.
• Each of the components is weighted by overall
  impact in the basin.
• Weighting factors change with season (winter
  versus summer -- snowpack or streamflow)

(Shafer and Dezman, 1982)
10
Weighting Coefficients

• May seem arbitrary, but were thoughtfully
  developed based on known reservoir capacity as
  percent of average annual streamflow.
• Specific process was followed for each basin
• Authors knew that coefficients might change
  monthly, but stuck with two-season approach for
  simplification

11
(No Transcript)
12
(No Transcript)
13
Input Data

• Snowpack (SWE)
• 6-7 Snow course/SNOTEL stations per basin.
• Streamflow monthly total volume flows
• Gages on major streams above diversions were
  preferable.
• 1-4 gages per basin chosen by location,
  regulation and data availability.
• Reservoir Storage
• Only managed non-random variable
• Both irrigation and municipal reservoirs used to
  reduce dependence of previous values.
• South Platte uses 18 reservoirs, others 4-6, none
  on Yampa/White/N.Platte
• Precipitation
• 4-6 National Weather Service stations per basin.
• Cumulative water-year values used Dec May
  (mimic snowpack), individual monthly amounts for
  remainder of year.

(Shafer and Dezman, 1982)
14
Example Snowpack non-exceedance probability for
the Rio Grande basin
(Shafer and Dezman, 1982)
15
Things to consider

• Discontinuance of stations used in the
  calculation.
• When one or more extreme events occur and go
  beyond the frequency distribution.
• The time series must be reanalyzed.
• Changes in water management activities.
• New or expansion of existing reservoirs.
• Changing reservoir operation rules

(Shafer and Dezman, 1982)
16
Additional Consideration

• Have you looked at reservoir data lately?
• Probability distributions can be quite odd

17
What has 28 years of SWSI computation shown us?
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
References

• Shafer, B.A. and L.E. Dezman. 1982. Development
  of a Surface Water Supply Index (SWSI) to assess
  the severity of drought conditions in snowpack
  runoff areas. In Proceedings of the Western Snow
  Conference, pp. 164175. Colorado State
  University, Fort Collins, Colorado.

27
My personal concluding thoughts

• SWSI was developed to be a practical tool
• SWSI must be worth something or we would have
  quit using it long ago
• SWSI values below -3 for consecutive months do
  correspond well to known drought conditions
• WATF intended to test SWSI by comparing it to
  observed drought impacts dont think we ever
  did that, but always perceived it to be practical
  and useful

28
More Thoughts

• WATF participants always felt that SWSI could be
  improved -- just werent sure it was worth the
  effort just to satisfy WATF
• Fundamental questions
• How many basins? How localized?
• Monthly coefficients?
• Realities of inhomogeneous data
• Predictive or Diagnostic?