Search for: All records

As the prospect of average global warming exceeding 1.5°C becomes increasingly likely, interest in supplementing mitigation and adaptation with solar geoengineering (SG) responses will almost certainly rise. For example stratospheric aerosol injection to cool the planet could offset some of the warming for a given accumulation of atmospheric greenhouse gases ( 1 ). However, the physical and social science literature on SG remains modest compared with mitigation and adaptation. We outline three research themes for advancing policy-relevant social science related to SG: (i) SG costs, benefits, risks, and uncertainty; (ii) the political economy of SG deployment; and (iii) SG’s role in a climate strategy portfolio. 

Forest insects and pathogens have significant impacts on U.S. forests, annually affecting an area nearly three times that of wildfires and timber harvesting combined. However, coupled with these direct effects of forest insects and pathogens are the indirect impacts through influencing forest management practices, such as harvesting. In an earlier study, we surveyed private woodland owners in the northeastern U.S. and 84% of respondents indicated they intended to harvest in at least one of the presented insect invasion scenarios. This harvest response to insects represents a potentially significant shift in the timing, extent, and species selection of harvesting. Here we used the results from the landowner survey, regional forest inventory data, and characteristics of the emerald ash borer (Species: Agrilus planipennis Fairmaire, 1888) invasion to examine the potential for a rapidly spreading invasive insect to alter harvest regimes and affect regional forest conditions. Our analysis suggests that 25% of the woodland parcels in the Connecticut River Watershed in New England may intend to harvest in response to emerald ash borer. If the emerald ash borer continues to spread at its current rate within the region, and therefore the associated management response occurs in the next decade, this could result in an increase in harvest frequencies, from 2.6% year−1 (historically) to 3.7% year−1 through to approximately 2030. If harvest intensities remain at levels found in remeasured Forest Inventory and Analysis plots, this insect-initiated harvesting would result in the removal of 12%–13% of the total aboveground biomass. Eighty-one percent of the removed biomass would be from species other than ash, creating a forest disturbance that is over twice the magnitude than that created by emerald ash borer alone, with the most valuable co-occurring species most vulnerable to biomass loss. 

Recent advances in deep learning for neural networks with large numbers of parameters have been enabled by automatic differentiation, an algorithmic technique for calculating gradients of measures of model fit with respect to model parameters. Estimation of high‐dimensional parameter sets is an important problem within the hydrological sciences. Here, we demonstrate the effectiveness of gradient‐based estimation techniques for high‐dimensional inverse estimation problems using a conceptual rainfall‐runoff model. In particular, we compare the effectiveness of Hamiltonian Monte Carlo and automatic differentiation variational inference against two nongradient‐dependent methods, random walk Metropolis and differential evolution Metropolis. We show that the former two techniques exhibit superior performance for inverse estimation of daily rainfall values and are much more computationally efficient on larger data sets in an experiment with synthetic data. We also present a case study evaluating the effectiveness of automatic differentiation variational inference for inverse estimation over 25 years of daily precipitation conditional on streamflow observations at three catchments and show that it is scalable to very high dimensional parameter spaces. The presented results highlight the power of combining hydrological process‐based models with optimization techniques from deep learning for high‐dimensional estimation problems.

 

Invasive forest insects can induce tree mortality in two ways: (a) by directly harming trees; or (b) by influencing forest owners to pre‐emptively harvest threatened trees. This study investigates forest owners’ intentions to harvest trees threatened by invasive insects.

Our first objective is to identify and characterize agent functional types (AFTs) of family forest owners in the northeastern United States using a set of contingent behaviour questions contained in a mail survey. We establish AFTs as a form of dimension reduction, effectively casting landowners into a typology in which each type (AFT) has distinct probabilities of tree harvesting in response to forest insects. Our analysis identifies three functional types of landowners: ‘Cutters’ (46% of respondents; high intent to harvest trees impacted by invasive forest insects), ‘Responsive Cutters’ (42% of respondents; intent sensitive to insect impact severity), and ‘Non‐cutters’ (12% of respondents; low intent to cut).

Our second objective is to model AFT membership to predict the distribution of AFTs across the landscape. Predictors are chosen from a set of survey, geographic and demographic features. Our best AFT‐prediction model has three predictor variables: parcel size (hectares of forest), geographical region, and town‐level forested fraction. Application of the model provides a high‐resolution probability distribution of AFTs across the landscape.

By coupling human and insect behaviour, our results allow for holistic assessments of how invasive forest insects disturb forests, inclusive of the management response to these pests.

A freePlain Language Summarycan be found within the Supporting Information of this article.