Past, Present, and Future Thermal Regime of Lakes in Western Alaska.

摘要

The temperature regime of surface waters and its phase (i.e., liquid or ice) are fundamental attributes of lake ecosystems that directly control energy balance and exchange with the atmosphere, strongly influences biological productivity and phenology, and in permafrost environments additionally influence thaw bulb (talik) development, thermal erosion, and ultimately surface permafrost degradation. This project focused on the past, present, and future thermal regime of lakes in Western Alaska by combining in situ observations, remotely sensed temperature observations, and hindcast/forecast model development using regional weather station data and global circulation models. In general, (1) shallow lakes in Yukon-Kuskokwin Delta (YKD), Seward Peninsula, and Kotzebue areas are warmer and more responsive to Air Temperature than deeper lakes on the Alaska Peninsula, Bristol Bay Lowlands, and Kodiak Island; (2) contemporary July lake temperatures average 14-17 degrees C for shallow lakes and average 8-11 degrees C for deep lakes in western Alaska; (3) in shallow lakes, temperatures (mean daily)commonly exceed 20 degrees C in the early summer (late June and early July) with highest modeled and observed values of 24 degrees C in smallest lakes; (4) in deeper lakes, temperatures rarely exceed 15 degrees C and summer peak values often occur late in the season (August-September); (5) generally, the summers of 1997, 2004, 2005, and 2013 produced the warmest average temperatures across all western AK lakes; and (6) the coolest years according to mean summer (open-water) and July temperatures were very inconsistent among lakes and regions. Comparing in situ observations with contemporary model output showed better performance for shallow thermokarst lakes in the central and northern portion of the study region (R2 range of 0.73 to 0.95) than for deep glacial/tectonic lakes in the southern portion and deepmaar lakes in the northern portion of the study region (R2 range of 0.27 to 0.67). This variation in model performance in terms of fit and overall temperature bias should be strongly considered when interpreting the results of hindcasts and forecasts for the lakes observed in this study. In general, (1) hindcasts model output using the NWS data from 1985 to present did not reveal any trend in lake surface temperatures; (2) The only significant water temperature trend noted during hindcast period was slight cooling in several YKD lakes; (3) forecasting lake surface temperature out to 2100 using GCM data showed that all lakes are expected to warm but at different rates and magnitudes; (4) most GCM driven lake temperature show increase during historic period, but not observed in NWS driven lake temperature modeled data output; (5) the lakes in Wood-Tikchik region are expected to warm the most relative to present conditions; and (6) the lakes on the YK Delta are expected to have the highest maximum future summertime temperatures, as observed during historic record. Future work on lake temperature should work to include continued monitoring on a few key lakes in western Alaska with additional measurements collected to account for other energy balance and meteorological forcing. Access, dedicated scientists, and committed funding are needed to maintain such lake observatories.