Dr. John Christy: “no-significant-trend” in S. Sierra snowfall since 1916

While there’s always lots of worry in California and Nevada over water supplies driven by the Sierra snowpack, and wailing in the MSM over what global warming will do to the snowpack, there doesn’t seem to be any trend, up or down.

John Christy has provided me with his latest paper, just published in E&E. I’m been authorized by him to present it here. In this case, no news is good news.

CHANGES IN SNOWFALL IN THE SOUTHERN SIERRA

NEVADA OF CALIFORNIA SINCE 1916

John R. Christy

Justin J. Hnilo

Earth System Science Center

University of Alabama in Huntsville

ABSTRACT

A time series (1916–2009) of annual snowfall totals for Huntington Lake (HL, elev.

2141 m) in the southern Sierra Nevada of California is reconstructed. A

reconstruction is (a) necessary because HL data after 1972 are mostly missing and

(b) possible because nearby stations reveal high correlations with HL, two above

0.90. The results show mean annual snowfall in HL is 624 cm with an insignificant

trend of +0.5 cm (+0.08%) ±13.1 cm decade−1. Similar positive but insignificant

trends for spring snowfall were also calculated. Annual stream flow and

precipitation trends for the region again were insignificantly positive for the same

period. Snow-water-equivalent comparisons, measured on 1 Apr since 1930 at

26 sites and since 1950 at 45, show similar small, mostly positive, and insignificant

trends. These results combined with published temperature time series, which also

reveal no significant trends, form a consistent picture of no remarkable long-term

changes in the snowfall of this area and elevation of the southern Sierra Nevada of

California since the early 20th century.

INTRODUCTION

Paleo-reconstructions of western U.S. precipitation indicate significant periods of

drought and surplus with relatively high multi-decadal variability (e.g. Meko et al.

2007). Could the region be entering a period of reduced precipitation, with a reduction

in snowfall in the mountains, perhaps as dry as that estimated from 12th century treerings (Meko et al 2007)? In terms of recent trends, Mote et al. 2005 found mostly

upward trends in snow water equivalent in the southern Sierra for the period limited to

1950–1997 (48 years, or about half of the current study). They found positive trends as

well in the southern Rocky Mountain region, while poleward of approximately 38°N

there were widespread declines.

Barnett et al. 2008 indicate that for 1950–1999 most of the Western U.S. snowyregions show warming temperatures and earlier peak runoff, suggesting a trend towardless snow and more rain. This could be an ominous development for water resourceplanners as the mountain snowmelt, both its quantity and timing, provides a majorresource on which municipal, industrial and agricultural systems rely.

We shall examine snowfall itself because it is a vital metric to understand since it is critical for businesses and operations related to snow (winter sports, road clearing, etc.) as well as

snow-dependent ecological systems.

The question we will examine is whether a tendency in snowfall in the Southern

Sierra Nevada (So. Sierra) is detectable. The So. Sierra are important for many

reasons including their location as one of the most southern mountain ranges in the

U.S. with significant water resource impacts and thus potentially an early indicator of

climate change since modeled changes show significant warming here due to

enhanced greenhouse gas concentrations (e.g. Snyder et al. 2002). Mote et al. 2005

examined only 48 years of data and Barnett et al. 2008 only 50 years, but both found

a slight upward trend in water-resource availability in the So. Sierra. In an earlier

study of snow water equivalent (SWE) measured on 1 Apr of each year, Howat and

Tulaczyk 2005 found no trend in SWE for 177 snow courses.

However, by subtracting 1 Apr from 1 Mar SWE there appeared to be a small gain (loss) in Δ SWE for 1950–2002 at the higher (lower) elevations along with insignificant increases in water volume for Nov–Mar. The implication here is that over a shorter period of time,

the SWE contours on 1 Apr have risen in elevation. However, while extremely

valuable as a water resource index for late-spring and summer runoff, SWE on 1 Apr

often misrepresents the actual total snowfall during the cold season as early snows

may have melted by this time and later snows are not included (see examples later).

We shall look at annual snowfall as a different, though obviously related, climate

metric relative to SWE.

Has snowfall changed over a longer period in the mid-elevation (∼2000 m) of the

So. Sierra? This question has links to our previous study of the So. Sierra in which

seasonal maximum (TMax) and minimum (TMin) temperatures were produced

(Christy et al. 2006). The wet-season (Dec-May) temperature trends for 1910–2003

were not significantly different from zero (TMax +0.08, TMin −0.01 °C decade−1),

suggesting that if precipitation trends were near zero, then snowfall might also show

little change. Indeed, an examination of annual “water year” (Jul – Jun) precipitation

totals for this region’s climate division indicates a trend of +0.2% decade−1 (1916–2009)

while that of the nearest long term station (Fresno) shows +2.7% decade−1. Thus a

look at a longer snowfall record, and attendant variables such as runoff, is one way to

examine consistency, at least obliquely, to the temperature record.

CONCLUSION

With the available data from six mid-elevation stations in the Southern Sierra region of

California we reconstructed annual snowfall totals for 36 missing years of the

Huntington Lake record to complete the time series (1916–2009). The standard error

of the missing years is calculated to be ±36 cm, or 6% of the 94-year annual mean of

624 cm in the most robust estimation method (though we utilized the average of six

methods which reduces the standard error further.)

The results of both the annual and spring snowfall time series indicate no

remarkable changes for the 1916–2009 period in the basins drained by the Merced, San

Joaquin, Kings and Kaweah Rivers. In the six reconstructions the range of trend results

varied only slightly from −0.3% to +0.6 % decade−1. With a consensus trend of only

+0.5 cm (+0.08%) decade−1 ±13.1 cm decade−1 there is high confidence in the

“no-significant-trend” result. The corroborating information on temperature trends

(Christy et al. 2006), stream flow, precipitation and shorter period snow water

equivalent trends presented here are consistent with “no-significant-trend” in So. Sierra

snowfall near 2000m elevation since 1916.

==================================

Paper (PDF) is here:

2010_ChristyH_snow