9568

• UCREFRP Catalog Number: 9568
• Author: Resource Consultants Inc.
• Title: Sediment Transport Studies of the Little Snake, Yampa, and Green River Systems.
• USFW Year: 1991.
• USFW - Doc Type: Fort Collins, CO.
• Copyright Material: NO

I
I
I
I
I
I
I
I
I
I
I

I

4.

Pearson product-moment correlation coefficient was determined for the relationship
between the annual data in each division and the annual data for the area-averaged
composite variable. These correlation coefficients show the strength of the relationship
between the individual climate divisions and the area-wide composite values.

5.

Linear regression, with year as the independent variable, was conducted to determine the
strength of any linear changes over the study period and/or for subperiods within the
record.

6.

A Student's t test was conducted to determine if any significant changes in mean
conditions had occurred before and after the construction of Flaming Gorge Dam (the
first period extends from 1895-1962, the second period from 1963-1989); the t value is
computed as:

td =

Xl - X2 - (Ill - 112)

[ Nl S12 + N2 S; (~ + ~)]1/2
N + N - 2 Nl N2
1 2

where (Ill and 112) is the expected difference in the two subperiod means (set equal to 0)
and s1 and s2 are the standard deviations within the subperiods. Absolute values of t
above 1.99 comrrm a significant change in means at the 0.95 level of confidence.

7.

The Bartlett test for inconsistency of dispersion was computed to determine if changes in
climate variance had occurred before and after the construction of the Flamin~ Gorge
~am. With only two subperiods involved, this Bartlett test reduces to a ratio s max /
s min where s2max is the larger of the two subperiod variances and s2min is the smaller
of the two variances. The ratio, M, is made positive when variances have increased and
negative when variances have decreased. When the absolute value of M exceeds 1.84, a
significant difference in subperiod variances is confrrmed at the 0.95 level of confidence.

4.5.4 Results

The results from these analyses are summarized in Table 43 and time series plots are presented in
Figures 4.8-4.10. Findings for each variable include:

I
I
I

I
I

I
t

1. Generally, mean annual temperatures in the study area have been rising at a statistically
significant rate. Examination of the time series plot of mean annual temperatures
(Figure 4.8) reveals a highly variable pattern from year to year. However, the 5-year
running means show a tendency for cooling from 1895 to the late 1910s, sharp warming
from the mid-to-Iate 1910s to the mid-1930s, and warming from the early 1970s to the
present.

At four of the six climate divisions, and for the area-averaged data, the Student's
t test results (Table 4.3) indicate a highly significant increase in the temperatures of the
1963-1989 subperiod when compared to the 1895-1962 subperiod. A linear trend fitted to
the 1895-1989 area-averaged annual data reveals a statistically significant rise in
temperature of 0.0140p yr-1 over the 95 years of record. The rise in temperature during
the most recent subperiod (1963-1989) is O.051'F yr-1; this rate of temperature increase is
more than four times larger than the rate for the entire 95-year period. No significant
changes in variance were determined in any of the temperature arrays.

4-11