Paper No. 7
Presentation Time: 3:00 PM


LEE, Gregory1, GRANITTO, Matthew2, WANG, Bronwen3, SHEW, Nora3, MAUK, Jeffrey L.1, HAYES, Timothy S.4, JONES III, James V.5, SCHMIDT, Jeanine M.3, TODD, Erin3 and WERDON, Melanie B.6, (1)U.S. Geological Survey, PO Box 25046, Denver Federal Center, Mailstop 973, Denver, CO 80225, (2)U.S. Geological Survey, Mineral Resources Division, PO Box 25046, MS 973, Denver Federal Center, Denver, CO 80225-0046, (3)Alaska Science Center, U.S. Geological Survey, 4210 University Dr, Anchorage, AK 99508, (4)U.S. Geological Survey, E.N.R. Building, Tucson, AZ 85719, (5)U.S. Geological Survey, Alaska Science Center, 4210 University Drive, Anchorage, AK 99508, (6)Alaska Division of Geological & Geophysical Surveys, 3354 College Rd, Fairbanks, AK 99709-3707,

Alaska has an area of 1.7 million km2, and regional to local geochemical surveys have been undertaken by state and federal agencies. We present a new compilation of geochemical data that have been collected since the early 1960s until the present, including (1) 65,993 sediment samples from the National Uranium Resource Evaluation (NURE) program, (2) samples from the U.S. Geological Survey including sediments (90,360), soils (6,869), rocks (108,966), heavy-mineral concentrates (48,096), and minerals (7,470), and (3) samples from the Alaska Division of Geological & Geophysical Surveys including sediments (10,919), soils (100), rocks (12,437), and heavy-mineral concentrates (1,063). Sample density varies significantly on a local scale, but averages 1 site per 11.6 km2 for the entire state.

Legacy data in our compilation have highly variable accuracy, precision, and associated lower determination limits due to progressive improvements in analytical methods during the last 50 years. Due to the relatively high detection limits for many older analytical methods, many geochemically scarce elements have highly censored data distributions. For these elements, the Kaplan–Meier estimator was used to provide estimates for mean, median, standard deviation, and percentile values. For data that are lightly censored, however, most statistical parameters can be calculated directly. Robust statistical methods and percentiles allow identification of samples that are anomalously enriched in various elements, although the highly censored data preclude identification of samples with anomalously low concentrations. In many places, areas that show element enrichments correlate with the geologic substrate; for example, sediments that are enriched in Ni and Cr commonly occur in belts of mafic to ultramafic rocks. These results suggest that careful treatments of legacy data can provide significant and revealing geochemical maps for large areas that may be useful in mapping through cover, as well as for environmental and mineral resource applications.