This paper derives statistical models for predicting wintertime subseasonal temperature over the western US. The statistical models are trained on two separate datasets, namely observations and dynamical model simulations, and are based on least absolute shrinkage and selection operator (lasso). Surprisingly, statistical models trained on dynamical model simulations can predict observations better than observation-trained models. One reason for this is that simulations involve orders of magnitude more data than observational datasets.

K–Ar and 40Ar/39Ar ages have been measured on nine mafic volcanic rocks younger than 1 myr from the Snake River Plain (Idaho), Mount Adams (Washington), and Crater Lake (Oregon). The K–Ar ages were calculated from Ar measurements made by isotope dilution and K2O measurements by flame photometry. The 40Ar/39Ar ages are incremental-heating experiments using a low-blank resistance-heated furnace. The results indicate that high-quality ages can be measured on young, mafic volcanic rocks using either the K–Ar or the 40Ar/39Ar technique. The precision of an 40Ar/39Ar plateau age generally is better than the precision of a K–Ar age because the plateau age is calculated by pooling the ages of several gas increments. The precision of a plateau age generally is better than the precision of an isotope correlation (isochron) age for the same sample. For one sample the intercept of the isochron yielded an 40Ar/36Ar value significantly different from the atmospheric value of 295.5. Recalculation of increment ages using the isochron intercept for the composition of nonradiogenic Ar in the sample resulted in much better agreement of ages for this sample. The results of this study also indicate that, given suitable material and modern equipment, precise K–Ar and 40Ar/39Ar ages can be measured on volcanic rocks as young as the latest Pleistocene, and perhaps even the Holocene.

A sediment core spanning ∼ 7000 cal yr BP recovered from Stella Lake, a small sub-alpine lake located in Great Basin National Park, Nevada, was analyzed for subfossil chironomids (non-biting midges), diatoms, and organic content (estimated by loss-on-ignition (LOI)). Subfossil chironomid analysis indicates that Stella Lake was characterized by a warm, middle Holocene, followed by a cool “Neoglacial” period, with the last two millennia characterized by a return to warmer conditions. Throughout the majority of the core the Stella Lake diatom-community composition is dominated by small, periphytic taxa which are suggestive of shallow, cool, alkaline, oligotrophic waters with extensive seasonal ice cover. A reconstruction of mean July air temperature (MJAT) was developed by applying a midge-based inference model for MJAT (two-component WA-PLS) consisting of 79 lakes and 54 midge taxa (rjack2 = 0.55, RMSEP = 0.9°C). Comparison of the chironomid-inferred temperature record to existing regional paleoclimate reconstructions suggests that the midge-inferred temperatures correspond well to regional patterns. This multi-proxy record provides valuable insight into regional Holocene climate and environmental conditions by providing a quantitative reconstruction of peak Holocene warmth and aquatic ecosystem response to these changes in the Great Basin, a region projected to experience increased aridity and higher temperatures.