Geographic variation in projected US forest aboveground carbon responses to climate change and atmospheric deposition.

Forest composition and ecosystem services are sensitive to anthropogenic pressures like climate change and atmospheric deposition of nitrogen (N) and sulfur (S). Here we extend recent forest projections for the current cohort of trees in the contiguous US, characterizing potential changes in aboveground tree carbon at the county level in response to varying mean annual temperature, precipitation, and N and S deposition. We found that relative to a scenario with N and S deposition reduction and no climate change, greater climate change led generally to decreasing aboveground carbon (mean -7.5% under RCP4.5, -16% under RCP8.5). Keeping climate constant, reduced N deposition tended to lessen aboveground carbon (mean -7%), whereas reduced S deposition tended to increase aboveground carbon (+3%) by 2100. Through mid-century (2050), deposition was more important for predicting carbon responses except under the extreme climate scenarios (RCP8.5); but, by 2100, climate drivers generally outweighed deposition. While more than 70% of counties showed reductions in aboveground carbon relative to the reference scenario, these were not evenly distributed across the US. Counties in the Northwest and Northern Great Plains, and the northern parts of New England and the Midwest, primarily showed positive responses, while counties in the Southeast showed negative responses. Counties with greater initial biomass showed less negative responses to climate change while those which exhibited the greatest change in composition (>15%) had a 95% chance of losing carbon relative to a no-climate change scenario. This analysis highlights that declines in forest growth and survival due to increases in mean temperature and reductions in atmospheric N deposition are likely to outweigh positive impacts of reduced S deposition and potential increases in precipitation. These effects vary at the regional and county level, however, so forest managers must consider local rather than national dynamics to maximize forest carbon sinks in the future.

Climate change could negate U.S. forest ecosystem service benefits gained through reductions in nitrogen and sulfur deposition.

Climate change and atmospheric deposition of nitrogen (N) and sulfur (S) impact the health and productivity of forests. Here, we explored the potential impacts of these environmental stressors on ecosystem services provided by future forests in the contiguous U.S. We found that all stand-level services benefitted (+ 2.6 to 8.1%) from reductions in N+S deposition, largely attributable to positive responses to reduced S that offset the net negative effects of lower N levels. Sawtimber responded positively (+ 0.5 to 0.6%) to some climate change, but negatively (- 2.4 to - 3.8%) to the most extreme scenarios. Aboveground carbon (C) sequestration and forest diversity were negatively impacted by all modelled changes in climate. Notably, the most extreme climate scenario eliminated gains in all three services achieved through reduced deposition. As individual tree species responded differently to climate change and atmospheric deposition, associated services unique to each species increased or decreased under future scenarios. Our results suggest that climate change should be considered when evaluating the benefits of N and S air pollution policies on the services provided by U.S. forests.

Assessing Community Vulnerability to Extreme Events in the Presence of Contaminated Sites and Waste Management Facilities: An Indicator Approach.

Communities across the world are facing extreme events, such as excessive heat, droughts, floods, and wildfires. In the presence of contaminated sites and waste management facilities, communities must consider the impacts of potential releases from these sites due to such events. Impacts of extreme events on sites and consequently on surrounding, often disadvantaged communities result from complex interactions between natural, physical, and social factors. A conceptual framework was developed to identify and provide a shared understanding of key vulnerabilities and pathways that transcend disciplines. A transparent and replicable method was developed to create mappable indicators that represent contaminated sites, waste facilities, contaminant transport via air and water, and population sensitivities. This method can be applied as a screening step to assist states and local communities in prioritizing targeted strategies and resources and determining where in-depth assessments are needed. These indicators can facilitate communication with a broad audience more easily than complex modeling approaches or aggregated indices. Case study results demonstrate the importance of considering indicators in conjunction with each other. The indicator method was developed together with U.S.-based partners, but can be adapted for other countries seeking to understand the potential impacts of extreme events on contaminated sites and communities.

Building resilience to extreme weather events in Phoenix: Considering contaminated sites and disadvantaged communities.

The interplay of contaminated sites, climate change, and disadvantaged communities are a growing concern worldwide. Worsening extreme events may result in accidental contaminant releases from sites and waste facilities that may impact nearby communities. If such communities are already suffering from environmental, economic, health, or social burdens, they may face disproportionate impacts. Equitable resilience planning to address effects of extreme events requires information on where the impacts may be, when they may occur, and who might be impacted. Because resources are often scarce for these communities, conducting detailed modeling may be cost-prohibitive. By considering indicators for four sources of vulnerability (changing extreme heat conditions, contaminated sites, contaminant transport via wind, and population sensitivities) in one holistic framework, we provide a scientifically robust approach that can assist planners with prioritizing resources and actions. These indicators can serve as screening measures to identify communities that may be impacted most and isolate the reasons for these impacts. Through a transdisciplinary case study conducted in Maricopa County (Arizona, USA), we demonstrate how the framework and geospatial indicators can be applied to inform plans for preparedness, response, and recovery from the effects of extreme heat on contaminated sites and nearby populations. The indicators employed in this demonstration can be applied to other locations with contaminated sites to build community resilience to future climate impacts.

Future climate change effects on US forest composition may offset benefits of reduced atmospheric deposition of N and S.

Climate change and atmospheric deposition of nitrogen (N) and sulfur (S) are important drivers of forest demography. Here we apply previously derived growth and survival responses for 94 tree species, representing >90% of the contiguous US forest basal area, to project how changes in mean annual temperature, precipitation, and N and S deposition from 20 different future scenarios may affect forest composition to 2100. We find that under the low climate change scenario (RCP 4.5), reductions in aboveground tree biomass from higher temperatures are roughly offset by increases in aboveground tree biomass from reductions in N and S deposition. However, under the higher climate change scenario (RCP 8.5) the decreases from climate change overwhelm increases from reductions in N and S deposition. These broad trends underlie wide variation among species. We found averaged across temperature scenarios the relative abundance of 60 species were projected to decrease more than 5% and 20 species were projected to increase more than 5%; and reductions of N and S deposition led to a decrease for 13 species and an increase for 40 species. This suggests large shifts in the composition of US forests in the future. Negative climate effects were mostly from elevated temperature and were not offset by scenarios with wetter conditions. We found that by 2100 an estimated 1 billion trees under the RCP 4.5 scenario and 20 billion trees under the RCP 8.5 scenario may be pushed outside the temperature record upon which these relationships were derived. These results may not fully capture future changes in forest composition as several other factors were not included. Overall efforts to reduce atmospheric deposition of N and S will likely be insufficient to overcome climate change impacts on forest demography across much of the United States unless we adhere to the low climate change scenario.

What is Threatening Forests in Protected Areas? A Global Assessment of Deforestation in Protected Areas, 2001-2018.

The protection of forests is crucial to providing important ecosystem services, such as supplying clean air and water, safeguarding critical habitats for biodiversity, and reducing global greenhouse gas emissions. Despite this importance, global forest loss has steadily increased in recent decades. Protected Areas (PAs) currently account for almost 15% of Earth's terrestrial surface and protect 5% of global tree cover and were developed as a principal approach to limit the impact of anthropogenic activities on natural, intact ecosystems and habitats. We assess global trends in forest loss inside and outside of PAs, and land cover following this forest loss, using a global map of tree cover loss and global maps of land cover. While forests in PAs experience loss at lower rates than non-protected forests, we find that the temporal trend of forest loss in PAs is markedly similar to that of all forest loss globally. We find that forest loss in PAs is most commonly-and increasingly-followed by shrubland, a broad category that could represent re-growing forest, agricultural fallows, or pasture lands in some regional contexts. Anthropogenic forest loss for agriculture is common in some regions, particularly in the global tropics, while wildfires, pests, and storm blowdown are a significant and consistent cause of forest loss in more northern latitudes, such as the United States, Canada, and Russia. Our study describes a process for screening tree cover loss and agriculture expansion taking place within PAs, and identification of priority targets for further site-specific assessments of threats to PAs. We illustrate an approach for more detailed assessment of forest loss in four case study PAs in Brazil, Indonesia, Democratic Republic of Congo, and the United States.

Residential scene classification for gridded population sampling in developing countries using deep convolutional neural networks on satellite imagery.

BACKGROUND: Conducting surveys in low- and middle-income countries is often challenging because many areas lack a complete sampling frame, have outdated census information, or have limited data available for designing and selecting a representative sample. Geosampling is a probability-based, gridded population sampling method that addresses some of these issues by using geographic information system (GIS) tools to create logistically manageable area units for sampling. GIS grid cells are overlaid to partition a country's existing administrative boundaries into area units that vary in size from 50 m × 50 m to 150 m × 150 m. To avoid sending interviewers to unoccupied areas, researchers manually classify grid cells as "residential" or "nonresidential" through visual inspection of aerial images. "Nonresidential" units are then excluded from sampling and data collection. This process of manually classifying sampling units has drawbacks since it is labor intensive, prone to human error, and creates the need for simplifying assumptions during calculation of design-based sampling weights. In this paper, we discuss the development of a deep learning classification model to predict whether aerial images are residential or nonresidential, thus reducing manual labor and eliminating the need for simplifying assumptions. RESULTS: On our test sets, the model performs comparable to a human-level baseline in both Nigeria (94.5% accuracy) and Guatemala (96.4% accuracy), and outperforms baseline machine learning models trained on crowdsourced or remote-sensed geospatial features. Additionally, our findings suggest that this approach can work well in new areas with relatively modest amounts of training data. CONCLUSIONS: Gridded population sampling methods like geosampling are becoming increasingly popular in countries with outdated or inaccurate census data because of their timeliness, flexibility, and cost. Using deep learning models directly on satellite images, we provide a novel method for sample frame construction that identifies residential gridded aerial units. In cases where manual classification of satellite images is used to (1) correct for errors in gridded population data sets or (2) classify grids where population estimates are unavailable, this methodology can help reduce annotation burden with comparable quality to human analysts.

Atmospheric deposition and exceedances of critical loads from 1800-2025 for the conterminous United States.

Atmospheric deposition of nitrogen (N) and sulfur (S) has increased dramatically over pre-industrial levels, with many potential impacts on terrestrial and aquatic ecosystems. Quantitative thresholds, termed "critical loads" (CLs), have been developed to estimate the deposition rate above which damage is thought to occur. However, there remains no comprehensive comparison of when, where, and over what time periods individual CLs have been exceeded. We addressed this knowledge gap by combining several published data sources for historical and contemporary deposition, and overlaying these on six CL types from the National Critical Loads Database (NCLDv2.5; terrestrial acidification, aquatic acidification, lichen, nitrate leaching, plant community composition, and forest-tree health) to examine exceedances from 1800 to 2011. We expressed CLs as the minimum, 10th, and 50th percentiles within 12-km grid cells. Minimum CLs were relatively uniform across the country (200-400 eq·ha-1 ·yr-1 ), and have been exceeded for decades beginning in the early 20th century. The area exceeding minimum CLs peaked in the 1970s and 1980s, exposing 300,000 to 3 million km2 (depending on the CL type) to harmful levels of deposition, with a total area exceeded of 5.8 million km2 (~70% of the conterminous United States). Since then, deposition levels have dropped, especially for S, with modest reductions in exceedance by 2011 for all CL types, totaling 5.2 million km2 in exceedance. The 10th and 50th percentile CLs followed similar trends, but were not consistently available at the 12-km grid scale. We also examined near-term future deposition and exceedances in 2025 under current air quality regulations, and under various scenarios of climate change and additional nitrogen management controls. Current regulations were projected to reduce exceedances of any CL from 5.2 million km2 in 2011 to 4.8 million km2 in 2025. None of the additional N management or climate scenarios significantly affected areal exceedances, although exceedance severity declined. In total, it is clear that many CLs have been exceeded for decades, and are likely to remain so in the short term under current policies. Additionally, we suggest many areas for improvement to enhance our understanding of deposition and its effects to support informed decision making.

Analysis of the Costs and Economic Feasibility of Requiring Postharvest Processing for Raw Oysters.

Because of concerns about Vibrio vulnificus, the U.S. Food and Drug Administration is considering requirements for postharvest processing (PHP) of oysters harvested from the Gulf of Mexico during warm-weather months and intended for raw consumption. As described in the paper, feasible PHP methods for warm-weather-harvested oysters include cool pasteurization, high hydrostatic pressure, and low-dose gamma-irradiation. We estimate that the costs of applying PHP are approximately 5 to 6 cents per half-shell oyster intended for raw consumption. However, most oyster processors have insufficient volumes to cost-effectively install PHP equipment. To assist these smaller operations, central PHP facilities operated by a 3rd party would be needed. A geographic information system analysis that minimized volume-weighted travel distances from each Gulf oyster operation identified 6 optimal PHP facility locations in the Gulf region. Even with the establishment of central PHP facilities, some oyster operations will become unprofitable and be at risk for closure.

Synthesized Population Databases: A Geospatial Database of US Poultry Farms.

The pervasive and potentially severe economic, social, and public health consequences of infectious disease in farmed animals require that plans be in place for a rapid response. Increasingly, agent-based models are being used to analyze the spread of animal-borne infectious disease outbreaks and derive policy alternatives to control future outbreaks. Although the locations, types, and sizes of animal farms are essential model inputs, no public domain nationwide geospatial database of actual farm locations and characteristics currently exists in the United States. This report describes a novel method to develop a synthetic dataset that replicates the spatial distribution of poultry farms, as well as the type and number of birds raised on them. It combines county-aggregated poultry farm counts, land use/land cover, transportation, business, and topographic data to generate locations in the conterminous United States where poultry farms are likely to be found. Simulation approaches used to evaluate the accuracy of this method when compared to that of a random placement alternative found this method to be superior. The results suggest the viability of adapting this method to simulate other livestock farms of interest to infectious disease researchers.

The role of subway travel in an influenza epidemic: a New York City simulation.

The interactions of people using public transportation in large metropolitan areas may help spread an influenza epidemic. An agent-based model computer simulation of New York City's (NYC's) five boroughs was developed that incorporated subway ridership into a Susceptible-Exposed-Infected-Recovered disease model framework. The model contains a total of 7,847,465 virtual people. Each person resides in one of the five boroughs of NYC and has a set of socio-demographic characteristics and daily behaviors that include age, sex, employment status, income, occupation, and household location and membership. The model simulates the interactions of subway riders with their workplaces, schools, households, and community activities. It was calibrated using historical data from the 1957-1958 influenza pandemics and from NYC travel surveys. The surveys were necessary to enable inclusion of subway riders into the model. The model results estimate that if influenza did occur in NYC with the characteristics of the 1957-1958 pandemic, 4% of transmissions would occur on the subway. This suggests that interventions targeted at subway riders would be relatively ineffective in containing the epidemic. A number of hypothetical examples demonstrate this feature. This information could prove useful to public health officials planning responses to epidemics.

Protecting health care workers: a pandemic simulation based on Allegheny County.

BACKGROUND AND OBJECTIVES: The Advisory Committee on Immunization Practices has identified health care workers (HCWs) as a priority group to receive influenza vaccine. Although the importance of HCW to the health care system is well understood, the potential role of HCW in transmission during an epidemic has not been clearly established. METHODS: Using a standard SIR (Susceptible-Infected-Recovered) framework similar to previously developed pandemic models, we developed an agent-based model (ABM) of Allegheny County, PA, that incorporates the key health care system features to simulate the spread of an influenza epidemic and its effect on hospital-based HCWs. FINDINGS: Our simulation runs found the secondary attack rate among unprotected HCWs to be approximately 60% higher (54.3%) as that of all adults (34.1%), which would result in substantial absenteeism and additional risk to HCW families. Understanding how a pandemic may affect HCWs, who must be available to treat infected patients as well as patients with other medical conditions, is crucial to policy makers' and hospital administrators' preparedness planning.

Attribute Assignment to a Synthetic Population in Support of Agent-Based Disease Modeling.

Communicable-disease transmission models are useful for the testing of prevention and intervention strategies. Agent-based models (ABMs) represent a new and important class of the many types of disease transmission models in use. Agent-based disease models benefit from their ability to assign disease transmission probabilities based on characteristics shared by individual agents. These shared characteristics allow ABMs to apply transmission probabilities when agents come together in geographic space. Modeling these types of social interactions requires data, and the results of the model largely depend on the quality of these input data. We initially generated a synthetic population for the United States, in support of the Models of Infectious Disease Agent Study. Subsequently, we created shared characteristics to use in ABMs. The specific goals for this task were to assign the appropriately aged populations to schools, workplaces, and public transit. Each goal presented its own challenges and problems; therefore, we used different techniques to create each type of shared characteristic. These shared characteristics have allowed disease models to more realistically predict the spread of disease, both spatially and temporally.

Synthesized Population Databases: A US Geospatial Database for Agent-Based Models.

Agent-based models simulate large-scale social systems. They assign behaviors and activities to "agents" (individuals) within the population being modeled and then allow the agents to interact with the environment and each other in complex simulations. Agent-based models are frequently used to simulate infectious disease outbreaks, among other uses.RTI used and extended an iterative proportional fitting method to generate a synthesized, geospatially explicit, human agent database that represents the US population in the 50 states and the District of Columbia in the year 2000. Each agent is assigned to a household; other agents make up the household occupants.For this database, RTI developed the methods for generating synthesized households and personsassigning agents to schools and workplaces so that complex interactions among agents as they go about their daily activities can be taken into accountgenerating synthesized human agents who occupy group quarters (military bases, college dormitories, prisons, nursing homes).In this report, we describe both the methods used to generate the synthesized population database and the final data structure and data content of the database. This information will provide researchers with the information they need to use the database in developing agent-based models.Portions of the synthesized agent database are available to any user upon request. RTI will extract a portion (a county, region, or state) of the database for users who wish to use this database in their own agent-based models.

Strategies for mitigating an influenza pandemic.

Development of strategies for mitigating the severity of a new influenza pandemic is now a top global public health priority. Influenza prevention and containment strategies can be considered under the broad categories of antiviral, vaccine and non-pharmaceutical (case isolation, household quarantine, school or workplace closure, restrictions on travel) measures. Mathematical models are powerful tools for exploring this complex landscape of intervention strategies and quantifying the potential costs and benefits of different options. Here we use a large-scale epidemic simulation to examine intervention options should initial containment of a novel influenza outbreak fail, using Great Britain and the United States as examples. We find that border restrictions and/or internal travel restrictions are unlikely to delay spread by more than 2-3 weeks unless more than 99% effective. School closure during the peak of a pandemic can reduce peak attack rates by up to 40%, but has little impact on overall attack rates, whereas case isolation or household quarantine could have a significant impact, if feasible. Treatment of clinical cases can reduce transmission, but only if antivirals are given within a day of symptoms starting. Given enough drugs for 50% of the population, household-based prophylaxis coupled with reactive school closure could reduce clinical attack rates by 40-50%. More widespread prophylaxis would be even more logistically challenging but might reduce attack rates by over 75%. Vaccine stockpiled in advance of a pandemic could significantly reduce attack rates even if of low efficacy. Estimates of policy effectiveness will change if the characteristics of a future pandemic strain differ substantially from those seen in past pandemics.