• Canada
• Other research productsclose
• National Science Foundation close

We investigate the power of voting among diverse, randomized software agents. With teams of computer Go agents in mind, we develop a novel theoretical model of two-stage noisy voting that builds on recent work in machine learning. This model allows us to reason about a collection of agents with different biases (determined by the first-stage noise models), which, furthermore, apply randomized algorithms to evaluate alternatives and produce votes (captured by the second-stage noise models). We analytically demonstrate that a uniform team, consisting of multiple instances of any single agent, must make a significant number of mistakes, whereas a diverse team converges to perfection as the number of agents grows. Our experiments, which pit teams of computer Go agents against strong agents, provide evidence for the effectiveness of voting when agents are diverse.

The mid-Pliocene is a valuable time interval for investigating equilibrium climate at current atmospheric CO2 concentrations because atmospheric CO2 concentrations are thought to have been comparable to the current day and yet the climate and distribution of ecosystems were quite different. One intriguing, but not fully understood, feature of the early to mid-Pliocene climate is the amplified Arctic temperature response and its impact on Arctic ecosystems. Only the most recent models appear to correctly estimate the degree of warming in the Pliocene Arctic and validation of the currently proposed feedbacks is limited by scarce terrestrial records of climate and environment. Here we reconstruct the summer temperature and fire regime from a subfossil fen-peat deposit on west–central Ellesmere Island, Canada, that has been chronologically constrained using cosmogenic nuclide burial dating to 3.9+1.5/-0.5 Ma. The estimate for average mean summer temperature is 15.4±0.8 ∘C using specific bacterial membrane lipids, i.e., branched glycerol dialkyl glycerol tetraethers. This is above the proposed threshold that predicts a substantial increase in wildfire in the modern high latitudes. Macro-charcoal was present in all samples from this Pliocene section with notably higher charcoal concentration in the upper part of the sequence. This change in charcoal was synchronous with a change in vegetation that included an increase in abundance of fire-promoting Pinus and Picea. Paleo-vegetation reconstructions are consistent with warm summer temperatures, relatively low summer precipitation and an incidence of fire comparable to fire-adapted boreal forests of North America and central Siberia. To our knowledge, this site provides the northernmost evidence of fire during the Pliocene. It suggests that ecosystem productivity was greater than in the present day, providing fuel for wildfires, and that the climate was conducive to the ignition of fire during this period. The results reveal that interactions between paleo-vegetation and paleoclimate were mediated by fire in the High Arctic during the Pliocene, even though CO2 concentrations were similar to modern values.

Mesoscale eddies play a major role in controlling ocean biogeochemistry. By impacting nutrient availability and water column ventilation, they are of critical importance for oceanic primary production. In the eastern tropical South Pacific Ocean off Peru, where a large and persistent oxygen-deficient zone is present, mesoscale processes have been reported to occur frequently. However, investigations into their biological activity are mostly based on model simulations, and direct measurements of carbon and dinitrogen (N2) fixation are scarce. We examined an open-ocean cyclonic eddy and two anticyclonic mode water eddies: a coastal one and an open-ocean one in the waters off Peru along a section at 16°S in austral summer 2012. Molecular data and bioassay incubations point towards a difference between the active diazotrophic communities present in the cyclonic eddy and the anticyclonic mode water eddies. In the cyclonic eddy, highest rates of N2 fixation were measured in surface waters but no N2 fixation signal was detected at intermediate water depths. In contrast, both anticyclonic mode water eddies showed pronounced maxima in N2 fixation below the euphotic zone as evidenced by rate measurements and geochemical data. N2 fixation and carbon (C) fixation were higher in the young coastal mode water eddy compared to the older offshore mode water eddy. A co-occurrence between N2 fixation and biogenic N2, an indicator for N loss, indicated a link between N loss and N2 fixation in the mode water eddies, which was not observed for the cyclonic eddy. The comparison of two consecutive surveys of the coastal mode water eddy in November 2012 and December 2012 also revealed a reduction in N2 and C fixation at intermediate depths along with a reduction in chlorophyll by half, mirroring an aging effect in this eddy. Our data indicate an important role for anticyclonic mode water eddies in stimulating N2 fixation and thus supplying N offshore.

We have examined several approaches for estimating the surface concentration of particulate organic carbon, POC, from optical measurements of spectral remote-sensing reflectance, Rrs(Lambda), using field data collected in tropical and subtropical waters of the eastern South Pacific and eastern Atlantic Oceans. These approaches include a direct empirical relationship between POC and the blue-to-green band ratio of reflectance, Rrs(Lambda B)/Rrs(555), and two-step algorithms that consist of relationships linking reflectance to an inherent optical property IOP (beam attenuation or backscattering coefficient) and POC to the IOP. We considered two-step empirical algorithms that exclusively include pairs of empirical relationships and two-step hybrid algorithms that consist of semianalytical models and empirical relationships. The surface POC in our data set ranges from about 10 mg/m**3 within the South Pacific Subtropical Gyre to 270 mg/m**3 in the Chilean upwelling area, and ancillary data suggest a considerable variation in the characteristics of particulate assemblages in the investigated waters. The POC algorithm based on the direct relationship between POC and Rrs(Lambda B)/Rrs(555) promises reasonably good performance in the vast areas of the open ocean covering different provinces from hyperoligotrophic and oligotrophic waters within subtropical gyres to eutrophic coastal upwelling regimes characteristic of eastern ocean boundaries. The best error statistics were found for power function fits to the data of POC vs. Rrs(443)/Rrs(555) and POC vs. Rrs(490)/Rrs(555). For our data set that includes over 50 data pairs, these relationships are characterized by the mean normalized bias of about 2% and the normalized root mean square error of about 20%. We recommend that these algorithms be implemented for routine processing of ocean color satellite data to produce maps of surface POC with the status of an evaluation data product for continued work on algorithm development and refinements. The two-step algorithms also deserve further attention because they can utilize various models for estimating IOPs from reflectance, offer advantages for developing an understanding of bio-optical variability underlying the algorithms, and provide flexibility for regional or seasonal parameterizations of the algorithms.

Imperfect mixing is a concern in industrial processes, everyday processes (mixing paint, bread machines), and in understanding salt water-fresh water mixing in ecosystems. The effects of imperfect mixing become evident in the unstirred ferroin-catalyzed Belousov-Zhabotinsky reaction, the prototype for chemical pattern formation. Over time, waves of oxidation (high ferriin concentration, blue) propagate into a background of low ferriin concentration (red); their structure reflects in part the history of mixing in the reaction vessel. However, it may be difficult to separate mixing effects from reaction effects. We describe a simpler model system for visualizing density-driven pattern formation in an essentially unmixed chemical system: the reaction of pale yellow Fe3+ with colorless SCN− to form the blood-red Fe(SCN)2+ complex ion in aqueous solution. Careful addition of one drop of Fe(NO3)3 to KSCN yields striped patterns after several minutes. The patterns appear reminiscent of Rayleigh-Taylor instabilities and convection rolls, arguing that pattern formation is caused by density-driven mixing.

We study the Kleinberg problem of navigation in small-world networks when the underlying lattice is stretched along a preferred direction. Extensive simulations confirm that maximally efficient navigation is attained when the length r of long-range links is taken from the distribution P(r)∼r−α, when the exponent α is equal to 2, the dimension of the underlying lattice, regardless of the amount of anisotropy, but only in the limit of infinite lattice size, L→∞. For finite size lattices we find an optimal α(L) that depends strongly on L. The convergence to α=2 as L→∞ shows interesting power-law dependence on the anisotropy strength.

In a series of three papers in 2013, researchers at the Cornell Node of the NSF Census Research Network (http://ncrn.cornell.edu) investigated and proposed solutions for two fundamental yet distinct issues in the curation of quantitative social science data: confidentiality and provenance. We argued that the W3C PROV model, a foundation for semantically-rich, interoperable, and web-compatible provenance metatdata, is especially important in a web environment in which data from distributed sources and of varying integrity can be combined and derived. In this paper we combine and expand upon these two separate threads—confidentiality and provenance—and experiment with the use of PROV and DDI in documenting the complex provenance chain between the highly confidential environment of the U.S. Census Bureau and restricted and public versions of internal census demographic files. In particular, our presentation will report on our effort to: 1) test PROV’s ability to describe meaningful relationships between confidential, restricted and public data at the variable level; 2) develop a user interface for researchers attempting to understand the relationships between distinct versions of confidential, restricted, and public census files. Longer term our work should produce a useful metadata resource for users of public and restricted American Community Survey data.

We present the first synthesis of global peatland extent through the last glacial cycle (130 ka) based on >975 detailed stratigraphic descriptions from exposures, soil pits, and sediment cores. Buried peats are defined as organic-rich sediments overlain by mineral sediments. Also included are deposits rich in wetland macrofossils indicated a local peatland environment. The dataset includes location (lat/long), chronologic information (when available), a description of the buried peat sediment, overlying and underlying sediments, whether geochemical information is available, and the original references.