Hydrological models for climate-based assessments at the watershed scale: A critical review of existing hydrologic and water quality models.

Global changes, including climate and land use changes, can result in significant impact to water resources. Planning for these changes requires making projections, even in the face of considerable uncertainties, to make informed management and policy decisions. A number of climate change scenarios and projections at global and regional levels are available that can be used to predict the likely range of outcomes. However, there is a need to translate these projections into potential implications for hydrology and water quality. Since there are dozens of hydrologic models, there is a need to evaluate them critically and to develop guidance regarding selecting the appropriate model for a given objective. We conducted a review of 21 different models commonly used for modeling hydrology (8), water quality (6) or both (7) at the watershed scale. Six of the models are strictly water quality models that depend on a separate model or observed data for hydrology. Seven additional models are useful for estimating hydrology and water quality simultaneously. The models were then evaluated based on ten different criteria, including functionality, scope, ability to model extreme events, data requirements, availability, and technical support, among others. The models were ranked Low, Medium or High in each of the criteria. The results indicate that three hydrologic models, MIKE-SHE, HEC-HMS, and MODHMS, as well as two full hydrology and water quality models, SWAT and WARMF, stand out in terms of functionality, availability, applicability to a wide range of watersheds and scales, ease of implementation, and availability of support. Modelers should carefully select the best model for their application, in part guided by the criteria discussed herein.

Assessing the Risk of Engineered Nanomaterials in the Environment: Development and Application of the nanoFate Model.

We developed a dynamic multimedia fate and transport model (nanoFate) to predict the time-dependent accumulation of metallic engineered nanomaterials (ENMs) across environmental media. nanoFate considers a wider range of processes and environmental subcompartments than most previous models and considers ENM releases to compartments (e.g., urban, agriculture) in a manner that reflects their different patterns of use and disposal. As an example, we simulated ten years of release of nano CeO2, CuO, TiO2, and ZnO in the San Francisco Bay area. Results show that even soluble metal oxide ENMs may accumulate as nanoparticles in the environment in sufficient concentrations to exceed the minimum toxic threshold in freshwater and some soils, though this is more likely with high-production ENMs such as TiO2 and ZnO. Fluctuations in weather and release scenario may lead to circumstances where predicted ENM concentrations approach acute toxic concentrations. The fate of these ENMs is to mostly remain either aggregated or dissolved in agricultural lands receiving biosolids and in freshwater or marine sediments. Comparison to previous studies indicates the importance of some key model aspects including climatic and temporal variations, how ENMs may be released into the environment, and the effect of compartment composition on predicted concentrations.

Impacts of sea level rise and climate change on coastal plant species in the central California coast.

Local increases in sea level caused by global climate change pose a significant threat to the persistence of many coastal plant species through exacerbating inundation, flooding, and erosion. In addition to sea level rise (SLR), climate changes in the form of air temperature and precipitation regimes will also alter habitats of coastal plant species. Although numerous studies have analyzed the effect of climate change on future habitats through species distribution models (SDMs), none have incorporated the threat of exposure to SLR. We developed a model that quantified the effect of both SLR and climate change on habitat for 88 rare coastal plant species in San Luis Obispo, Santa Barbara, and Ventura Counties, California, USA (an area of 23,948 km(2)). Our SLR model projects that by the year 2100, 60 of the 88 species will be threatened by SLR. We found that the probability of being threatened by SLR strongly correlates with a species' area, elevation, and distance from the coast, and that 10 species could lose their entire current habitat in the study region. We modeled the habitat suitability of these 10 species under future climate using a species distribution model (SDM). Our SDM projects that 4 of the 10 species will lose all suitable current habitats in the region as a result of climate change. While SLR accounts for up to 9.2 km(2) loss in habitat, climate change accounts for habitat suitability changes ranging from a loss of 1,439 km(2) for one species to a gain of 9,795 km(2) for another species. For three species, SLR is projected to reduce future suitable area by as much as 28% of total area. This suggests that while SLR poses a higher risk, climate changes in precipitation and air temperature represents a lesser known but potentially larger risk and a small cumulative effect from both.

Species sensitivity distributions for engineered nanomaterials.

Engineered nanomaterials (ENMs) are a relatively new strain of materials for which little is understood about their impacts. A species sensitivity distribution (SSDs) is a cumulative probability distribution of a chemical's toxicity measurements obtained from single-species bioassays of various species that can be used to estimate the ecotoxicological impacts of a chemical. The recent increase in the availability of acute toxicity data for ENMs enabled the construction of 10 ENM-specific SSDs, with which we analyzed (1) the range of toxic concentrations, (2) whether ENMs cause greater hazard to an ecosystem than the ionic or bulk form, and (3) the key parameters that affect variability in toxicity. The resulting estimates for hazardous concentrations at which 5% of species will be harmed ranged from <1 ug/L for PVP-coated n-Ag to >3.5 mg/L for CNTs. The results indicated that size, formulation, and the presence of a coating can alter toxicity, and thereby corresponding SSDs. Few statistical differences were observed between SSDs of an ENM and its ionic counterpart. However, we did find a significant correlation between the solubility of ENMs and corresponding SSD. Uncertainty in SSD values can be reduced through greater consideration of ENM characteristics and physiochemical transformations in the environment.

Toxicity of nano-zero valent iron to freshwater and marine organisms.

We tested whether three commercial forms (uncoated, organic coating, and iron oxide coating) of nano zero-valent iron (nZVI) are toxic to freshwater and marine organisms, specifically three species of marine phytoplankton, one species of freshwater phytoplankton, and a freshwater zooplankton species (Daphnia magna), because these organisms may be exposed downstream of where nZVI is applied to remediate polluted soil. The aggregation and reactivity of the three types of nZVI varied considerably, which was reflected in their toxicity. Since levels of Fe(2+) and Fe(3+) increase as the nZVI react, we also evaluated their toxicity independently. All four phytoplankton species displayed decreasing population growth rates, and Daphnia magna showed increasing mortality, in response to increasing levels of nZVI, and to a lesser degree with increasing Fe(2+) and Fe(3+). All forms of nZVI aggregated in soil and water, especially in the presence of a high concentration of calcium ions in groundwater, thus reducing their transports through the environment. However, uncoated nZVI aggregated extremely rapidly, thus vastly reducing the probability of environmental transport and potential for toxicity. This information can be used to design a risk management strategy to arrest the transport of injected nZVI beyond the intended remediation area, by injecting inert calcium salts as a barrier to transport.