El Niño Breaks Up the Tanana River ✎

Original: "El Nino signature in Alaskan river breakups"
· G. Boffetta
arXiv:1104.4038 · 2011-04-20 · arXiv non-exclusive · independent analysis · 3 min
Warm El Niño winters push the ice breakup on Alaska's Tanana River 3.4 days earlier.

Context

El Niño and the Southern Oscillation (ENSO) set the pace of interannual climate variability across the planet. In the tropics, its influence is obvious: warm water shifts rains and droughts. At high latitudes, the signal weakens, drowned out by local weather. In the archive next to this work lies a study on predicting El Niño a year and a half ahead, but that one deals with the phenomenon itself, while here we focus on its distant footprint. This article stands apart: it turns a folk lottery into a tool for testing long-range connections in the climate system.

Methods

Data from the Nenana Ice Classic lottery were used: since 1917, residents of the town of Nenana (population about 500) guess when a tripod frozen into the ice of the Tanana River will move and stop a clock on the shore. This yielded a 90-year series of breakup dates. El Niño years were identified using the BEST index (a combination of sea surface temperature in the Niño 3.4 region and the Southern Oscillation), requiring the index to remain positive for at least four months from October to April. The average breakup date in El Niño years (14 events) was compared with the overall average. To assess significance, a Monte Carlo method was applied: $10^6$ times, 14 random years were chosen, and we checked how often the mean deviated as strongly. Meteorological data (temperature, precipitation, snow) came from the Fairbanks weather station, 90 km upstream, for 1930–2006.

Results

The average breakup date of the Tanana River over 1917–2006 falls on May 5 (day 124.7 of the year, with a standard deviation σ=5.97 days). In El Niño years, the ice moves on average 3.4 days earlier: day 121.3. The deviation is 2.3 standard deviations from the mean, with a probability of chance occurrence of 1.1%, meaning the link is significant at the 98.9% level. Accompanying meteorological data explain the mechanism: in El Niño years, the average maximum temperature from January to April in Fairbanks rises from −7.6°C to −6.3°C, and total snowfall drops from 75 to 55 cm. Less snow and warmer air accelerate snowmelt on land, which is directly linked to river breakup: the correlation between the breakup date in Nenana and the first snow-free day in Fairbanks is ρ≈0.50.

Implications

The work shows that the El Niño signal is discernible even in subarctic Alaska, provided there is a sufficiently long observational record. What distinguishes it from the neighboring study on El Niño prediction is the subject: not the phenomenon itself, but its distant consequence. This is the first statistically rigorous demonstration of an ENSO teleconnection to river ice breakup. The result supports the idea that the observed trend toward earlier breakup (about 5.5 days per century) has a climatological, not random, nature.

Future development

The topic could develop toward searching for similar signals in other long records: lake ice phenology, flowering dates, historical documents. It would be interesting to test whether the response to El Niño changes over time, especially against the backdrop of global warming. Possibly, strengthening or weakening of teleconnections could serve as a marker of shifts in atmospheric circulation.

Impact

The result is important for climatology, hydrology of northern rivers, and paleoclimate reconstructions, where ice breakup dates are used as a temperature proxy.

Next steps

Test the reproducibility of the signal on other rivers in Alaska and northwestern Canada. Compare with model simulations of the atmospheric response to ENSO at high latitudes.

Key open problems

For climatology, the open question remains how far and how predictably tropical anomalies propagate. This work links a global index to a local phenomenon, but the mechanism of signal transmission through atmospheric circulation requires further study. The relative roles of thermal and mechanical (snow) contributions to ice breakup also remain unclear.

The Nenana Ice Classic lottery has existed since 1917, and its top prize in some years exceeded 300 thousand dollars. The tripod placed on the ice is connected to a clock on the shore: when the ice moves, the tripod falls and stops the mechanism. Residents guess not only the day but also the minute of the event.

Key numbers

  • shift in breakup date during El Niño years: 3.4 days earlier
  • probability of chance coincidence: 1.1%
  • correlation between breakup date and snow disappearance: 0.50
  • reduction in snowfall during El Niño years: from 75 to 55 cm
  • increase in maximum January–April temperature: from −7.6°C to −6.3°C
Concepts
La Niña El Niño El Niño-Southern Oscillation ENSO Southern Oscillation ERSST equatorial wave propagation spring predictability barrier Westerly wind burst Multivariate ENSO Index recharge oscillator model relative ONI NOAA PMEL / TAO NOAA CPC Teleconnection North Equatorial Countercurrent Warm water volume Niño 3.4 Kelvin wave