WMOD00511

• Title: Precipitation Guage Testing on the Wasatch Plateau, Utah, During Early 1993
• Date: 11/1/1993
• State: UT
• Weather Modification - Doc Type: Report

...

688

B. SEVRUK AND L. ZAHLA VOVA

t

inhomogeneities of precipitation time-series and corrections for wind-induced losses. Consequently, the
effects of historical and recent changes in the GSE, such as moving the gauge to a new site, tree planting
and removal, or the construction of new buildings near the gauge site, can be quantified to produce estimates
of the wind-induced error and the possible errors in the amount of precipitation measured. Figure I shows,
for instance, how much the annual amount of precipitation would change as a consequence of a change
from protected to open GSE and vice versa. Thus, the selection of homogeneous precipitation time-series
for the aims of climate change studies and the adjustment of inhomogeneous time-series can be based on
the physics of the phenomenon rather than on statistics alone. In this manner Sevruk and Zahlavova (1992)
established a chronology of changes in GSE classes for 398 Swiss precipitation time-series. The average
record length was 80 years, spanning 10 to 173 years. The results showed (i) that there is, on average, one
change in the GSE every 27 years, (ii) that the frequency of change increases with the observation period
length, and (iii) that the frequency of change markedly increases for periods longer than 100 years, as is
shown in Figure 6. The most frequency change in degree of exposure was by one class and more (130 out
of 398). For 90 sites a change of 0,5 of a class was recorded, and for 23 sites a minimum two-class change
was recorded. The total number of gauge sites where no change appeared was 69. If the stations with a
90-year record, for the period 1901-1990, are considered, the number of gauge sites without change decreases
to 19. This increases to 23 for stations with a 50-year record for the period 1941-1990. The analysis showed
that for stations below 2000 ill a.s.1. the most frequency classes of GSE are 2 and 2'5, but the proportion
of classes I and 1.5 increases rapidly with increasing altitude, so that above 2000 m a.s.1. almost all gauge
sites are in these classes. A continual increase in the proportion of stations with GSE classes I and 1.5
throughout this century was also obvious, probably because some of the new gauge sites are situated out
of villages and at higher altitudes.
In another study, by Sevruk and Kirchhofer (1992), the classification of GSE was used to estimate the
wind speed at the level of the gauge orifice. The aim was to compute the wind-induced losses and to map
them together with precipitation losses over the territory of Switzerland, at a scale of I: 500000. In total,
340 gauge sites, including the shielded and non-shielded Hellmann gauge and the shielded Swiss storage
gauge, were taken into consideration. The altitude range was from 198 to 2540 m a.s.1. for the Hellmann
gauges and up to 3800 m a.s.1. for the storage gauges. Using the principle of similarity, the average annual
wind speed was transferred from measured to unmeasured sites, taking into consideration the regional wind
fields and the GSE class. Further, it was reduced to the level of the gauge orifice according to equation
(I). The corrections were calculated from the difference between corrected and uncorrected values of
precipitation, and interpolated on a grid I km x I km. They increased absolutely and relatively with
increasing altitude, from approximately 100 mm to 800 mm; this corresponds to 5-30 per cent of the

! 6
t
n
.
-; 4
~
~
2 .
.

o

o

40

80

L

120 (year)

Figure 6, The number of changes in gauge site exposure as a function of the length of the observation period. Average values, 398
gauge sites

""-""'~ll-1'i~~a
""'~'" .':,f.'.' :. :::'.;';;;1' ~b):;~