BARROW SNOWMELT DATE
ACKNOWLEDGEMENTS: This study was supported by NOAA's Climate Monitoring and Diagnostics Laboratory (CMDL), the International Arctic Research Center (IARC) and the Cooperative Institute for Arctic Research (CIFAR) - University of Alaska. We also wish to thank G. Divoky, J. Foster, C. George, D. Kane, G Maykut, R. McClure, and T. Zhang for providing various data sets.
BACKGROUND
To understand global climate change more fully it is necessary to assess the variability of earth's cryosphere in response to other climatic factors. In particular, the timing of the disappearance of snow each year can influence the net energy budget for an entire season. Feedbacks involving the change in surface albedo may enhance or diminish any response, which may be manifested in the regional temperature regime. At the NOAA/CMDL Barrow Observatory (BRW) the date of snowmelt has been observed since the late 1940s. A continuous record, on the basis of visual observations, of this event in Barrow has been kept by the National Weather Service. In recent years, NOAA has also made a determination of the snow disappearance date on the basis of objective, radiometric measurements made over open tundra. This is illustrated in Figure 1 which shows a typical annual cycle of the solar, or shortwave SW, components of the net radiation balance on a daily average basis and also the derived surface albedo.
Figure 1. Daily average downward and upward shortwave irradiance measured at the NOAA/CMDL Barrow Observatory during 1994 (top panel). The date of snow melt is determined radiometrically as the date when the surface albedo drops below 30%. [adapted from Stone, et al., 1996]
NOAA has selected a threshold of 30% albedo, the ratio of upward-to- downward SW irradiance, as a good indicator of final melt out. The date is determined to be that of the first daily average below 30%. Once this occurs each spring albedo seldom increases again until autumn except for an occasional late snowfall of brief duration that can occur even during summer. What is remarkable is that this annual event occurs over only a few days as Figure 1 indicates. However, snow disappearance each spring is highly variable from year to year, which in turn can influences the annual energy budget quite significantly. The timing of seasonal snow melt at high latitudes is potentially one of the most important but least understood processes that affects global climate through the "temperature - albedo feedback" mentioned above. Any long-term, regional trend in the distribution and melt of the snow pack may be interpreted as a manifestation of climate change. Therefore, we are examining the history of the Barrow snow melt date in great detail to understand why it varies and to determine if it is occurring earlier in response to global warming.
HISTORICAL SNOWMELT TIME SERIES
Foster (1989) claimed that the disappearance of snow in spring at Barrow showed a trend manifested by a progressively earlier melt since the 1950s, speculating that this was an indication of global warming. Dutton and Endres (1991), however, took issue with Foster's conclusion, suggesting that the apparent trend was, in large part, attributable to local urbanization effects. Their argument was based on more objective, radiometric measurements made over the open tundra, upwind of Barrow at the NOAA/ CMDL Observatory (BRW). These were certain not to be influenced by urban effects, but the analysis was based on only a few years of overlapping data.
Figure 2. The date of snow melt in the vicinity of Barrow has apparently occurred earlier in recent years. In town, the disappearance of snow is occurring earlier than at the NOAA CMDL Observatory located at a remote tundra site, possibly because of the effect of urbanization (e.g., Dutton and Endres, 1991).
The issue is revisited in the Figure 2 above, which shows the respective time series updated through 1996. It now appears that the radiometric estimates of the snowmelt date out on the tundra are also tending to occur earlier in recent years. Moreover, there is fair temporal correlation between the two independent observations giving credence to both data records. It is hypothesized that earlier melting of the snow pack, on average, each spring may result from a) less than normal accumulation of snow throughout the winter months, b) warmer spring temperatures that accelerates ablation/melt or c) a combination of a and b.
EVIDENCE OF A REGIONAL TREND IN THE DATE OF SNOW DISAPPEARANCE ON THE NORTH SLOPE OF ALASKA
Figure 3a shows time series of melt dates and proxy observations from six other North Slope sites that are correlated with the 1966-2000 BRW record. The locations of these sites are indicated on the map, Figure 3b. Yearly data were analyzed to determine trends and correlation coefficients, but only 5-year-smoothed data are presented. The Sagwon SAG (69.4°N, 148.8°W; elev. 351 m) and Franklin Bluffs FBF (69.9°N, 148.1°W; elev. 76 m) melt dates were determined using a 0.30 albedo threshold (e.g., Figure 1 ). Both sites are within the Kuparuk River Watershed located southeast of Barrow. Barter Island BTI (70.1°N, 143.6°W; elev. 15 m) was another NWS station where, until 1987, melt dates were determined from snow depth data. The series labeled satellite SAT was derived from visible satellite images of a strip of tundra about 150 km south of Barrow, Alaska [Foster et al., 1992]. The upper two curves in Figure 3a are proxy records. Cooper Island CPI (71.7°N, 155.7°W; elev. 3 m) is a time series of dates when a species of Arctic bird, the Black Guillemot, first lays an egg. Each spring Guillemots nest on the island but only after the snow melts and they have access to nest cavities do they breed, producing their first clutch of eggs about two weeks later. Isaktoak ISK is a time series of dates when ice has melted completely off of the Isaktoak Lagoon, which is located in the village of Barrow. These proxy records are correlated with the BRW time series suggesting that snowmelt and ice melt are influenced similarly by variations in climate. The dashed curve in Figure 3a represents an ensemble average of all observations normalized to the average timing of the BRW melt date. A linear fit of this 142 station-year record shows an advance in the spring melt of 8.0 days over 35 years 4.0 at a confidence level of 95%. However, the correlated variations of time series shown in Figure 3a are more indicative of climatic shifts than of a monotonic change. Note also that the melt tends to progress from the more southerly locations of the Kuparuk River Watershed, northward toward the coast (e.g., Barter Is.), and last in the vicinity of Barrow. Also, the dates of snow disappearance at all sites were moderately late in 1999 and 2000 and again in 2001 (not shown) indicating a recent shift in the conditions or factors that influence the timing of the spring melt in northern Alaska. These factors, which include snowfall amounts, temperatures, and sky cover that vary with synoptic conditions are discussed in Stone et al. [2001a, http://www.arm.gov/docs/documents/technical/conf_0103/stone-rs.pdf; and 2001b].
Figure 3. (a) Analyses of six independent time series of melt dates compared with the 1966-2000 BRW melt date record (updated from Figure 2). 5-year smoothed time series, and linear fits are shown. Each is correlated with the NOAA/CMDL-BRW record (red) with coefficients indicated [in brackets] for each of the sites described in the text. The dashed analysis (unlabeled) is for an ensemble average of the 142-station years, normalized to the BRW time frame, (b) map view of northern Alaska showing the approximate locations of the seven sites for which time series analyses shown in Figure 3a were made. Each is color-coded for cross-reference. [adapted from Stone et al., 2001b]
REFERENCES
Dutton, E. G. and D. J. Endres 1991: Date of snow melt at Barrow, Alaska, USA, Arctic and Alpine Research, 23,(1), p.115-119
Foster, J. L., J.W. Winchester, and E.G. Dutton, 1992: The date of snow disappearance on the Arctic tundra as determined from satellite, meteorological station and radiometric in situ observations, IEEE Transactions on Geoscience and Remote Sensing, 30(4), 793-798.
Foster, J. L. 1989: The significance of the date of snow disappearance on the arctic tundra as a possible indicator of climatic change, Arctic and Alpine Research, 21,(1), p. 60-70
Stone , R.S., E.G. Dutton, J.M. Harris, and D. Longenecker: THE ADVANCING DATE OF SPRING SNOWMELT IN THE ALASKAN ARCTIC, Proceedings of the Eleventh Atmospheric Radiation Measurement (ARM) Science Team Meeting, March 19-23, 2001, Atlanta, Georgia, 2001a.
Stone, R.S., E.G. Dutton, J.M. Harris, and D. Longenecker: Earlier Spring Snowmelt in Northern Alaska as an Indicator of Climate Change, J. Geophys. Res., in press 2001b.
RELATED LINKS
The Cooperative Institute for Arctic Research
International Arctic Research Center
![[GRAPH AS DESCRIBED BELOW]](graphics/snomelt2.gif)
![[GRAPH AS DESCRIBED BELOW]](graphics/melthist.gif)
![[GRAPH AS DESCRIBED BELOW AND MAP OF SITES]](graphics/jgrfig3ab.jpg)