Neural Network Input Selection for Hydrological Forecasting Affected by Snowmelt1

Abstract: Snowmelt largely affects runoff in watersheds in Nordic countries. Neural networks (NN) are particularly attractive for streamflow forecasting whereas they rely at least on daily streamflow and precipitation observations. The selection of pertinent model inputs is a major concern in NNs implementation. This study investigates performance of auxiliary NN inputs that allow short‐term streamflow forecasting without resorting to a deterministic snowmelt routine. A case study is presented for the Rivière des Anglais watershed (700 km²) located in Southern Québec, Canada. Streamflow (Q), precipitations (rain R and snow S, or total P), temperature (T) and snow lying (A) observations, combined with climatic and snowmelt proxy data, including snowmelt flow (Q_SM) obtained from a deterministic model, were tested. NN implemented with antecedent Q and R produced the largest gains in performance. Introducing increments of A and T to the NNs further improved the performance. Long‐term averages, seasonal data, and Q_SM failed to improve the networks.     

Do trans-Pacific air masses deliver PBDEs to coastal British Columbia, Canada?

In order to distinguish between ‘local’ and ‘background’ sources of polychlorinated biphenyls (PCBs) and polybrominated diphenyl ethers (PBDEs) in coastal British Columbia (Canada) air, we collected samples from two sites: a remote site on western Vancouver Island, and a near-urban site in the Strait of Georgia. Seasonally-integrated samples of vapor, particulate, and rain were collected continuously during 365 days for analysis of 275 PCB and PBDE congeners. While deposition of the legacy PCBs was similar at both sampling sites, deposition of PBDEs at the remote site amounted to 42% (10.4 mg/ha/year) of that at the near-urban site. Additional research into atmospheric circulation in the NE Pacific Ocean will provide more insight into the transport and fate of priority pollutants in this region, but trans-Pacific delivery of PBDEs to the west coast of North America may underlie in part our observations. For example, approximately 40% of >12,000 ten-day back trajectories calculated for the remote site originated over Asia, compared to only 2% over North America.     

Robustness in life history of the brown seaweed Ascophyllum nodosum (Fucales, Phaeophyceae) across large scales: effects of spatially and temporally induced variability on population growth

Understanding the demography and function of biotope-forming seaweed species is of great importance for the conservation of the target species itself, as well as its associated organisms. The brown seaweed Ascophyllum nodosum is fundamental for the functioning of coastal marine ecosystems in the North Atlantic. In this study, we use a data-based size-classified matrix model to investigate the temporal and spatial variability in demography, and the environment-specific stochastic sensitivity and elasticity, of two A. nodosum populations, one in western Sweden and one on the Isle of Man in the Irish Sea. A significant difference between the two populations was that the Swedish population had comparably low and more variable stochastic population growth rate (λ _s). This pattern was partly explained by the relatively high and varying mortality rates during extreme ice-years in Sweden, and by the lower survival of small individuals during all years. There were also fewer large individuals in Sweden due to lower transitions to the larger size-classes and higher probability of shrinkage. Sensitivities were analogous in the two populations, and showed a high selection pressure for increased individual growth. Elasticities were also similar, with the exception that survival of the smallest individuals (i.e., transition a _1,1), had a higher elasticity on the Isle of Man. Overall, the stochastic growth rate (λ _s) was most sensitive to proportional changes in loop- (i.e., survival within size-class) and, to some extent, growth-transitions in both study areas. These results show that structurally and demographically diverging A. nodosum populations may be similarly sensitive to changes in vital rates. This, in turn, indicates a plastic life history of A. nodosum that may cope with large environmental variability. The results further suggest that environmental change affecting the survival or growth of the larger, reproductive A. nodosum individuals could have severe and regional effects on the abundance and biomass of this species, with potential negative effects on the biodiversity of the associated communities.