China is confronting the challenge of accelerated lake eutrophication, where Lake Dianchi is considered as the most serious one. Eutrophication control for Lake Dianchi began in the mid-1980s. However, decision makers have been puzzled by the lack of visible water quality response to past efforts given the tremendous investment. Therefore, decision makers desperately need a scientifically sound way to quantitatively evaluate the response of lake water quality to proposed management measures and engineering works. We used a water quality modeling based scenario analysis approach to quantitatively evaluate the eutrophication responses of Lake Dianchi to an under-construction water diversion project. The primary analytic framework was built on a three-dimensional hydrodynamic, nutrient fate and transport, as well as algae dynamics model, which has previously been calibrated and validated using historical data. We designed 16 scenarios to analyze the water quality effects of three driving forces, including watershed nutrient loading, variations in diverted inflow water, and lake water level. A two-step statistical analysis consisting of an orthogonal test analysis and linear regression was then conducted to distinguish the contributions of various driving forces to lake water quality. The analysis results show that (a) the different ways of managing the diversion projects would result in different water quality response in Lake Dianchi, though the differences do not appear to be significant; (b) the maximum reduction in annual average and peak Chl-a concentration from the various ways of diversion project operation are respectively 11% and 5%; (c) a combined 66% watershed load reduction and water diversion can eliminate the lake hypoxia volume percentage from the existing 6.82% to 3.00%; and (d) the water diversion will decrease the occurrence of algal blooms, and the effect of algae reduction can be enhanced if diverted water are seasonally allocated such that wet season has more flows. (C) 2013 Elsevier B.V. All rights reserved.
The Fast Evolution of Vehicle Emissions from Roadway (FEVER) study was undertaken to strategically measure pollutant gradients perpendicular to a major highway north of Toronto, Canada. A case study period was analyzed when there was an average perpendicular wind direction. Two independent, fast response measurements were used to infer rapid organic aerosol (OA) growth on a spatial scale from 34 m to 285 m at the same time as a decrease was observed in the mixing ratio of primary emitted species, such as CO2 and NOx. An integrated organic gas and particle sampler also showed that near the highway, the aerosol had a larger semivolatile fraction than lower volatile fraction, but over a relatively short distance downwind of the highway, the aerosol transformed to being more low volatile with the change being driven by both evaporation of semivolatile and production of lower volatile organic aerosol. A new 1-D column Lagrangian atmospheric chemistry model was developed to help interpret the measured increase in the OA/CO2 curve from 34 m to 285 m downwind of highway, where the refers to background-corrected concentrations. The model was sensitive to the assumptions for semivolatile organic compounds (SVOCs). Different combinations of SVOC emissions and background mixing ratios were able to yield similar model curves and reproduce the observations. Future measurements of total gas-phase SVOC in equilibrium with aerosol both upwind and downwind of the highway would be helpful to constrain the model.
Extraction and transmission of compact descriptors are of great importance for next-generation mobile visual search applications. Existing visual descriptor techniques mainly compress visual features into compact codes of fixed bit rate, which is not adaptive to the bandwidth fluctuation in wireless environment. In this letter, we propose a Rate-adaptive Compact Fisher Codes (RCFC) to produce a bit rate scalable image signature. In particular, RCFC supports fast matching of descriptors based on Hamming distance; meanwhile, low memory footprint is offered. Extensive evaluation over benchmark databases shows that RCFC significantly outperforms the state-of-the-art and provides a promising descriptor scalability in terms of bit rates versus desired search performance.