Global greenhouse warming is predicted by Global Climate Models (GCMs) to change temperature and precipitation in the Arctic to a greater extent than in temperate and tropical regions. Within a hundred years or less, the temperatures in the Arctic may be several degrees Centigrade warmer than current averages. As a result, scientists expect to see changes in snow accumulation and melting, in the depth of summer soil thaw and in summer precipitation patterns. Terrestrial plant communities will probably change in stature and in species composition in many regions with concomitant changes in evapotranspiration which returns water to the atmosphere. All of these factors will affect the water balance of drainages across the circumpolar arctic watershed with potentially large changes in the discharge of freshwater and water-borne materials to the coastal seas of the Arctic Ocean. Societal impacts of changes in freshwater runoff to the ocean include feedbacks to global climate change through alterations in sea ice distribution and Arctic Ocean circulation as well as through impacts on the plants and animals of arctic wetlands, rivers and coastal seas.
Freshwater inputs exert a surprisingly large impact on the water circulation of the Arctic Ocean and on the Global Ocean circulation as well. Freshwater inputs promote stratification, a layering of waters of differing salt content and hence density, of the arctic shelf seas and contribute to the formation of sea ice. As sea ice forms it excludes salt and leaves behind very salty waters called brines. These dense brines form descending plumes which travel offshore and contribute to the mid-depth and deep water masses of the Arctic Ocean basin. Since the water masses beneath the Arctic Ocean surface waters are warmer than the surface waters in contact with the polar ice, the maintenance of the physical separation provided by the stratification provided by saltier layers below is essential to the existence of the permanent polar ice cover. Without this permanent stratification, the ice cover would melt. This would greatly affect the climate of the polar regions because the ice reflects much more of the incoming solar radiation than seawater.
The impacts of freshwater inputs extend even further than the Arctic Ocean basin because the brackish sea ice and the low salinity surface waters of the Arctic Ocean flow southward into the North Atlantic. This is the region famous for the production of the very cold and dense North Atlantic Deep Water (NADW). NADW sinks and flows southward at depth forming the beginning of the dominant pattern of circulation of the World Ocean. However, during periods of high freshwater export from the Arctic, these bouyant freshwaters inhibit the process of NADW formation and thus cause a slowing or even a stagnation of the ocean circulation with potentially large impacts on global climate. For example, the Gulf Stream which moderates the climate of northern Europe would diminish and plunge Europe into a cold period as happened during the historical ``Little Ice Age''.
For these reasons, Global Climate Model simulations which predict a warmer Arctic with higher precipitation are stimulating the research community to intensify the study of freshwater inputs to the Arctic Ocean. These inputs may be a keystone aspect of the global climate system yet we cannot say with confidence what the changes in freshwater export from land to the Arctic Ocean would be under altered climate. Futhermore, we cannot yet predict what would happen if the polar ice cover melted or the NADW formation slowed for an extended period. We do know, however, that the impacts would be large and global in extent.
A research project led by Bruce Peterson is studying the pan-arctic water balance with particular emphasis on a Geographic Information System (GIS) approach to the estimation of freshwater export via rivers to the arctic coastal seas. Water balances for all the watersheds of the Arctic are being determined and annual variability in river discharge from 1960 to 1990 is being assessed. Water balance and river routing models will be combined with river water nutrient chemistry data to develop statistically-based models of the fluxes of nutrients from watersheds to the Arctic Ocean. The product will be baseline estimates of contemporary water and nutrient fluxes from land to water for the entire pan-arctic drainage. These estimates are needed now to help assess the likely impacts of predicted climate change on water and nutrient balances for the pan-arctic watershed. The baseline discharge data are also needed to formulate, test and validate water balance and routing models required for prediction of future drainage basin to ocean fluxes when climate and the arctic ecosystem are very different from today.
The research is interdisciplinary and international in scope. Areas of expertise and contributing U. S. scientists include arctic river biogeochemistry (Bruce Peterson), water balance modeling (Charles Vorosmarty, University of New Hampshire, Durham, NH ), permafrost modeling (Steve Frolking, University of New Hampshire), precipitation distribution estimation (Cort Willmott, University of Delaware, Newark, Delaware), and atmospheric water vapour dynamics (Mark Serreze, University of Colorado, Boulder, Colarado). International contributions and collaborators include controls of riverine biogeochemical fluxes (Michel Meybeck, University of Paris, Paris, France), river discharge and hydrometeorological data sets (Igor Shiklomanov, State Hydrological Institute, St. Petersburg, Russia), monitoring of water and nutrient fluxes: AMAP program (Vitaly Kimstach, Arctic Monitoring and Assessment Program, Oslo, Norway), and Russian riverine nutrient fluxes (Viatoheslav Gordeev, Shirshov Institute of Oceanology, Moscow, Russia). A Russian postdoctoral scientist (Alexander Shiklomanov, formerly of the Arctic and Alpine Research Institute, St. Petersburg, Russia) has recently joined the project to assemble unpublished data on precipitation and discharge for the Russian Arctic and to perform hydrology modeling.
Each member of this team of investigators provides a unique contribution to the long-term goal of securing a quantitative and predictive understanding of how runoff and associated biogeochemical fluxes are linked between the pan-arctic land mass and the arctic ocean coastal zone. For example, the atmospheric water vapor budgets for large drainages provide checks on the precipitation and runoff measurements for the same watersheds. Likewise, the ability to interpret and map the arctic precipitation distribution based on a sparse network of measuring stations depends critically on Cort Willmott's expertise in spatial interpolation routines which account for topographic and wind field effects. The involvement of Russian scientists is needed because the bulk of the data on precipitation and discharge for the vast Russian arctic region is not available in electronic format and must be retrieved, documented, checked and formatted for this project.
Data for drainage basins throughout the Arctic is archived at the University of New Hampshire in the Global Hydrological Archive and Analysis System for use in drainage basin modeling. This GIS-based system allows the efficient development of relationships between watershed characteristics, climate and river discharge. More than 40 data layers are currently available in GIS format on a pan-arctic and global scale including river networks,elevation, vegetation, geology, permafrost, soils, fertilizer use, population, lakes and reservoirs, runoff and climate. Our project is now using this analysis system and the discharge data in two ways. First we produce estimates of current discharge for all the watersheds flowing to the Arctic Ocean as illustrated in Figure 1. These estimates are based on the actual long-term mean annual discharge for the period of record where gauging data exists and on the water balance model where records do not exist. The second use of the measured discharge data is to improve the calibration and to validate the water balance and routing models for the arctic regime which is dominated by snowmelt and soil water thaw during the relatively short summer season. A validated model will provide the basis for prediction of monthly runoff from all arctic rivers to the Arctic Ocean under conditions predicted by climate and land surface models for future decades. These are among the model outputs required for prediction of Arctic Ocean stratification and circulation over the coming decades as greenhouse warming effects increase from barely detectable to readily measurable.