Climatological data are essential to investigations of the five core themes (Van Cleve and Martin 1991) of the Long-Term Ecological Research (LTER) program. In the McMurdo LTER, research ranges from glaciology to soil nematology, and every component of the research is tied to meteorology/climatology in some manner. The need to monitor climate change and its impact on the local environment is central to the LTER theme, and the need to sample common parameters for intersite comparison is strong. At many LTER sites, records of climate variables are the only truly "long-term" records initially available (Greenland 1987). In like manner, one of the first objectives of the McMurdo LTER was to establish a meteorological network that would gather representative, year-round dry valley weather data.

Because of the harsh weather conditions, absence of sunlight to drive solar-powered systems for approximately 4 months of the year, and inaccessibility of the region in the winter, representative climate data collection in the McMurdo Dry Valleys is challenging. Furthermore, the heterogeneity of weather conditions in this region requires a weather network be used for data collection, as opposed to the single-point measurements common to many LTER sites (Greenland 1987). A sporadic history of manned and unmanned observations in the past provided crucial information on how to proceed with network construction. The first meteorological data to be reported for the dry valleys were summer-only data for the Lake Vanda region (Bull 1966, pp. 177-194). Manned winter observations were made by the New Zealand Antarctic Reserach Program (NZARP) in 1969, 1970, and 1974 (Thompson, Craig, and Bromley 1971; Riordan 1975, pp. 268-275; Bromley 1985).

The first automatic meteorologic station established in the dry valleys operated from February 1980 to December 1982 on Linnaeus Terrace, Upper Wright Valley, at an elevation of 1,750 meters (Stearns et al. 1993, pp. 1-22). This station (Asgard) was part of the ongoing University of Wisconsin automatic weather station (AWS) program, which operates a network of approximately 30 stations on the continent. Locations, affiliation, and operational history of this and other pre-LTER automatic weather stations are shown in figure 1.

At the start of the first McMurdo LTER field season (1993-1994), only two meteorological stations operated on the floor of the dry valleys. The McMurdo LTER took over control of the Lake Hoare station, and the U.S. Geological Survey continued to manage the Explorer's Cove station (both were replaced by new stations at the same sites that year). The Lake Fryxell lake-ice station was rebuilt and operated from 7 December 1993 to 6 January 1994 before being cannibalized to re-establish the Lake Fryxell shore station, which had been destroyed by wind the previous winter (the new Fryxell and Hoare stations are located a few meters above the old sites to avoid rising lake levels). New stations were built on the Commonwealth and Howard Glaciers and on the shore of Lake Bonney.

During the 1994-1995 field season, new stations were constructed at Lake Vanda (near the new Vanda Station at the mouth of the Onyx River) and the Taylor Glacier. Also, the McMurdo LTER, in collaboration with NZARP, established a station at Lake Brownsworth in Lower Wright Valley at the head of the Onyx River. Primary measurements made on all McMurdo LTER stations, instrumentation used, and time of initiation, are detailed in the table, and locations are shown in figure 1.

At present, all stations, with the exception of the two Wright Valley station, sample sensors every 30 seconds and send summary statistics (For example, averages and maximums) to solid-state storage modules every 20 minutes. a typical station will send about 20 values to final storage every output interval, resulting in a data collection rate of more than 500,000 data points per year (at 2 bytes per datum), per station. The two Wright Valley stations perform output functions every hour, resulting in one third of the data flux per year. An example of data collected at Lake Hoare in 1994 is shown in figure 2.

Two new stations will be added to the network in the 1995-1996 field season. A standard station will be established on the shore of Lake Vida in Victoria Valley. Lake Vida is one of the largest lakes in the dry valleys, and yet it is thought to be frozen throughout its 10-meter depth, making it an interesting anomaly for study among large dry valley lakes. The second new station will be established on Canada Glacier in taylor Valley. Year-round, this station will monitor all of the standard meteorological parameters, as well as net radiation and sensible and latent heat fluxes with the use of eddy correlation sensors during the summer.

Beyond the 1995-1996 season, we plan to connect the network by radio telemetry back to the McMurdo LTER base camp at Lake Hoare and eventually to McMurdo Station where data will be internet-accessible in real time. This will help in assessing problems before the field season begins and offers potential for remote sensing applicaitons. Our data flux is too large to use ARGOS satellites for data transmission, and the network is too far south for GOES satellite transmission. We will also be actively pursuing alternatives for long-term management of the McMurdo LAWN.

This research is funded by National Science Foundation grant OPP 92-11773. We'd like to thank Gary Clow (U.S. Geological Survey) and Dan Anderson (Campbell Scientific) for invaluable advice during network construction and Chris McKay for information on pre-LTER sites.

REFERENCES.

Bromley, A.M. 1985. Weather observations, Wright Valley, Antarctica (Information Publication 11). Wellington, New Zealand: New Zealand Meteorological Service.

Bull, C. 1966. Climatological observations in ice-free areas of southern Victoria Land, Antarctica. In M.J. Rubin (Ed.), Studies in antarctic meteorology (Antarctic Research Series, Vol. 9). Washington, D.C.: American Geophysical Union, Antarctic Research Serices.

Clow, G.D., C.P. McKay, G.M. Simmons, Jr., and R.A. Wharton, Jr. 1988. Climatological observations and predicted sublimation rates at Lake Hoare, Antarctica. Journal of Climate, 1(7), 715-728.

Friedmann, E.I., C.P. McKay, and J.A. Nienow. 1987. The cryptoendolithic microbial environment in the Ross Desert of Antarctica: Satellite-transmitted continuous nanoclimate data, 1984 to 1986. Polar Biology, 7, 273-287.

Greenland, D. (Ed.) 1987. The climates of Long-Term Ecological Research sites (Occasional Paper 44). Boulder: University of Colorado, Institute of Arctic and Alpine Research.

McKay, C.P., J.I. Nienow, M.A. Meyer, and E.I. Freidmann. 1993. Continuous nanoclimate data (1985-1988) from the Ross Deseart (McMurdo Dry Valleys) cryptoendolithic microbial ecosystem. In D.H. Bromwich and C.R. Stearns (Eds.), Antarctic meteorology and climatology (Antarcti Research Series, Vol. 61). Washington, D.C.: American Geophysical Union.

Riordan, A.J. 1975. The climate of Vanda Station, Antarctica. In G.Weller and S.A. Bowling (Eds.), Climate of the Arctic. Faribanks, Alaska: Geophysical Institute, University of Alaska.

Stearns, C.R., L.M. Keller, G.A. Weidner, and M. Sievers. 1993. Monthly mean climatic data for antarcti automatic weather stations. In D.H. Bromwich and C.R. Stearns (Eds.), Antarctic meteorology and climatology (Antarctic Research Series, Vol. 61). Washingotn, D.C.: American Geophysical Union.

Thompson, D.C., R.M.F. Craig, and A.M. Bromley. 1971. Climate and heat balance in an antarctic dry valley. New Zealand Journal of Science, 14, 245-251.

Van Cleve, K., and S. Marin (Eds.). 1991. Long-term ecological research in the United States (6th ed., rev.; LTER publication number 11). Seattle: Long-Term Ecological Network Office.