Publications

Understanding the costs and the spatial distribution of health and employment outcomes of low-carbon electricity pathways is critical to enable an equitable transition. We integrate an electricity system planning model (GridPath), a health impact model (InMAP), and a multiregional input–output model to quantify China’s provincial-level impacts of electricity system decarbonization on costs, health outcomes, employment, and labor compensation. We find that even without specific CO2 constraints, declining renewable energy and storage costs enable a 26% decline in CO2 emissions in 2040 compared to 2020 under the Reference scenario. Compared to the Reference scenario, pursuing 2 °C and 1.5 °C compatible carbon emission targets (85% and 99% decrease in 2040 CO2 emissions relative to 2020 levels, respectively) reduces air pollution-related premature deaths from electricity generation over 2020–2040 by 51% and 63%, but substantially increases annual average costs per unit of electricity demand in 2040 (21% and 39%, respectively). While the 2 °C pathway leads to a 3% increase in electricity sector-related net labor compensation, the 1.5 °C pathway results in a 19% increase in labor compensation driven by greater renewable energy deployment. Although disparities in health impacts across provinces narrow as fossil fuels phase out, disparities in labor compensation widen with wealthier East Coast provinces gaining the most in labor compensation because of materials and equipment manufacturing, and offshore wind deployment.

In response to growing scrutiny surrounding commodity-driven deforestation, companies have introduced zero-deforestation commitments (ZDCs) with ambitious environmental and social targets. However, such initiatives may not effectively reduce deforestation if they are not aligned with the spatial extent of remaining forests at risk. They may also fail to avert socio-economic risks if ZDCs do not consider smallholder farmers’ needs. We assess the spatial and functional fit of ZDCs by mapping commodity-driven deforestation and socio-economic risks, and comparing them to the spatial coverage and implementation of ZDCs in the Indonesian palm oil sector. Our study finds that companies’ ZDCs often underperform in four areas: traceability, compliance support for high-risk palm oil mills, transparency, and smallholder inclusion. In 2020, only one-third of companies sourcing from their own mills, and just 6% of those sourcing from external suppliers, achieved full traceability to plantations. Comparing the reach of ZDCs adopted by downstream buyers with those adopted by mill owners located further upstream, we find that high-quality ZDCs from buyers covered 62% of forests at risk, while mill owners’ ZDCs only covered 23% of forests at risk within the mill supply base. In Kalimantan and Papua, the current and future deforestation frontiers, the forests most at risk of conversion were predominantly covered by weak ZDCs lacking in policy comprehensiveness and implementation. Additionally, we find that only 46% of independent smallholder oil palm plots are in mill supply sheds whose owners offer programs and support for independent smallholders, indicating that smallholder inclusion is a significant challenge for ZDC companies. These results highlight the lack of spatial and functional alignment between supply chain policies and their local context as a significant gap in ZDC implementation and a challenge that the EU Deforestation Regulation will face.

As the climate changes, efficient climate policy requires a better understanding of how individuals adapt. Despite extensive research on various climate adaptation frictions, including financial and technological constraints, models of adaptive decision-making assume that agents have perfect information and accurate beliefs about climate. Combining rural household data in Bangladesh with a meteorological measure of dryness, this paper studies the role of individual drought beliefs and their accuracy in irrigation decisions as a key adaptive margin. In a theoretical model, I introduce a behavioral friction to document how heterogeneous beliefs differentially influence responsiveness to the same meteorological signal in dryness. The empirical analysis reveals an asymmetric response to dry shocks in irrigation conditional on the accuracy of prior beliefs. A counterfactual analysis shows lower technology adoption levels and higher monetary losses when beliefs are inaccurate.

Efficiently managing agricultural irrigation is vital for food security today and into the future under climate change. Yet, evaluating agriculture’s hydrological impacts and strategies to reduce them remains challenging due to a lack of field-scale data on crop water consumption. Here, we develop a method to fill this gap using remote sensing and machine learning, and leverage it to assess water saving strategies in California’s Central Valley. We find that switching to lower water intensity crops can reduce consumption by up to 93%, but this requires adopting uncommon crop types. Northern counties have substantially lower irrigation efficiencies than southern counties, suggesting another potential source of water savings. Other practices that do not alter land cover can save up to 11% of water consumption. These results reveal diverse approaches for achieving sustainable water use, emphasizing the potential of sub-field scale crop water consumption maps to guide water management in California and beyond.

The application of quasiexperimental impact evaluation to remotely sensed measures of deforestation has yielded important evidence detailing the effectiveness of conservation policies. However, researchers have paid insufficient attention to the binary and nonrepeatable structure of most deforestation datasets. Using analytical proofs and simulations, we demonstrate that many commonly employed econometric approaches are biased when applied to binary and nonrepeatable outcomes. The significance, magnitude and even direction of estimated effects from many studies are likely incorrect, threatening to undermine the evidence base that underpins conservation policy adoption and design. To address these concerns, we provide guidance and new strategies for the design of panel econometric models that yield more reliable estimates of the impacts of forest conservation policies.

Forest fire frequency, extent, and severity have rapidly increased in recent decades across the western United States (US) due to climate change and suppression-oriented wildfire management. Fuels reduction treatments are an increasingly popular management tool, as evidenced by California's plan to treat 1 million acres annually by 2050. However, the aggregate efficacy of fuels treatments in dry forests at regional and multi-decadal scales is unknown. We develop a novel fuels treatment module within a coupled dynamic vegetation and fire model to study the effects of dead biomass removal from forests in the Sierra Nevada region of California. We ask how annual treatment extent, stand-level treatment intensiveness, and spatial treatment placement alter fire severity and live carbon loss. We find that a ∼30% reduction in stand-replacing fire was achieved under our baseline treatment scenario of 1,000 km² year⁻¹ after a 100-year treatment period. Prioritizing the most fuel-heavy stands based on precise fuel distributions yielded cumulative reductions in pyrogenic stand-replacement of up to 50%. Both removing constraints on treatment location due to remoteness, topography, and management jurisdiction and prioritizing the most fuel-heavy stands yielded the highest stand-replacement rate reduction of ∼90%. Even treatments that succeeded in lowering aggregate fire severity often took multiple decades to yield measurable effects, and avoided live carbon loss remained negligible across scenarios. Our results suggest that strategically placed fuels treatments are a promising tool for controlling forest fire severity at regional, multi-decadal scales, but may be less effective for mitigating live carbon losses.

Marine protected areas (MPAs) are widely used for ocean conservation, yet the relative impacts of various types of MPAs are poorly understood. We estimated impacts on fish biomass from no-take and multiple-use (fished) MPAs, employing a rigorous matched counterfactual design with a global dataset of >14,000 surveys in and around 216 MPAs. Both no-take and multiple-use MPAs generated positive conservation outcomes relative to no protection (58.2% and 12.6% fish biomass increases, respectively), with smaller estimated differences between the two MPA types when controlling for additional confounding factors (8.3% increase). Relative performance depended on context and management: no-take MPAs performed better in areas of high human pressure but similar to multiple-use in remote locations. Multiple-use MPA performance was low in high-pressure areas but improved significantly with better management, producing similar outcomes to no-take MPAs when adequately staffed and appropriate use regulations were applied. For priority conservation areas where no-take restrictions are not possible or ethical, our findings show that a portfolio of well-designed and well-managed multiple-use MPAs represents a viable and potentially equitable pathway to advance local and global conservation.

Clean-cooking transitions have the potential to generate large public health, environmental and societal gains for 2.6 billion people in the Global South. Here we use data from Ghana’s largest household energy survey (n = 7,389) to provide two main insights. First, regression analysis of 13 commonly cited socio-economic and demographic determinants of household fuel use indicates remarkably different relationships with clean-fuel use at different stages of the transition process. We propose a stage-based transition framework that can help inform the rollout of clean-cooking interventions. Second, we identify factors that are associated with the exclusive use of liquefied petroleum gas (LPG) using a statistically powered sample of exclusive LPG users (n = 693). We show that, all else equal, increases in wealth and urbanicity are not—contrary to conventional wisdom—associated with a transition from primary to exclusive LPG use. Whereas further research is needed to determine causality, our findings highlight the potential for more careful measurement, isolating each stage of the clean-cooking transition, to inform new insights and policy opportunities.

Emerging evidence suggests that air pollution may play a role in shaping suicide risk by altering brain function. However, this link is difficult to quantify and has yet to be investigated in China, where 16% of global suicides occur. Here we apply a statistical model that leverages random increases in particulate pollution (PM_2.5) due to meteorological conditions to comprehensive data on suicide rates across Chinese counties. We find that a 1 s.d. (σ) increase in PM_2.5 raises weekly suicide rates by ∼25%. This effect occurs without delay, consistent with neurobiological evidence that PM_2.5 influences emotional regulation and impulsive–aggressive behaviour. Effects are sex and age specific; women over 65 exhibit significantly higher vulnerability. We estimate that PM_2.5 reductions under China’s Air Pollution Action Plan prevented 13,000–79,000 (95% confidence interval) suicides over 2013–2017, accounting for ∼10% of this period’s observed suicide rate decline. Our findings uncover a causal link between particulate pollution and suicide, adding urgency to calls for pollution control policies across the globe.

Offshore renewable energy, particularly wind farms, is rapidly expanding globally and has become an essential component of many coastal nations’ decarbonization plans, including the United States. The addition of these physical structures to the marine space may impact fish production and may preclude fishers from traditional fishing grounds - both of which have the potential to affect fisheries outcomes. Understanding the socioeconomic and sociocultural impacts of implementing offshore wind is crucial to determining appropriate mitigation strategies and to developing data collection, monitoring, and adaptive management strategies. This review synthesizes quantitative and qualitative indicators that have been used to assess the impact of fisheries preclusion and shifts in fished species’ biomass on fishery participants. By providing a description of the indicator, a list of the datasets required to calculate its value, and a list of studies that used the indicator, this review can serve as a guide to those designing monitoring plans to determine socioeconomic and sociocultural offshore wind impacts.

Human–wildlife conflict is an important factor in the modern biodiversity crisis and has negative effects on both humans and wildlife (such as property destruction, injury, or death) that can impede conservation efforts for threatened species. Effectively addressing conflict requires an understanding of where it is likely to occur, particularly as climate change shifts wildlife ranges and human activities globally. Here, we examine how projected shifts in cropland density, human population density, and climatic suitability—three key drivers of human–elephant conflict—will shift conflict pressures for endangered Asian and African elephants to inform conflict management in a changing climate. We find that conflict risk (cropland density and/or human population density moving into the 90th percentile based on current-day values) increases in 2050, with a larger increase under the high-emissions “regional rivalry” SSP3 - RCP 7.0 scenario than the low-emissions “sustainability” SSP1 - RCP 2.6 scenario. We also find a net decrease in climatic suitability for both species along their extended range boundaries, with decreasing suitability most often overlapping increasing conflict risk when both suitability and conflict risk are changing. Our findings suggest that as climate changes, the risk of conflict with Asian and African elephants may shift and increase and managers should proactively mitigate that conflict to preserve these charismatic animals.

Conservation interventions typically focus on protecting public goods, but they often also create private spillover co-benefits. For example, protecting open space may increase the values of adjacent properties and protecting a coral reef may increase fishing opportunities outside. These privately-captured co-benefits can confer substantial value, but are rarely tapped to help promote and expand conservation efforts. One reason, we argue, is that doing so is difficult: While co-beneficiaries are easily convinced of the benefits of the conservation intervention, they are not obliged to pay for it, and so usually free-ride and enjoy these benefits gratis. In this paper, we document and quantify the magnitude of co-benefits in the literature and identify the conditions under which co-benefits could be tapped to offset the cost of conservation for conservationists. In light of these conditions, we propose an approach that involves voluntary compensation for the provision of co-benefits to expand the total amount of resources available for conservation efforts. We show that taking advantage of these co-benefits lowers the cost of implementing conservation actions while being incentive compatible for all parties involved.

Trawling the seafloor can disturb carbon that took millennia to accumulate, but the fate of that carbon and its impact on climate and ecosystems remains unknown. Using satellite-inferred fishing events and carbon cycle models, we find that 55-60% of trawling-induced aqueous CO₂ is released to the atmosphere over 7-9 years. Using recent estimates of bottom trawling’s impact on sedimentary carbon, we found that between 1996-2020 trawling could have released, at the global scale, up to 0.34-0.37 Pg CO₂ yr^-1 to the atmosphere, and locally altered water pH in some semi-enclosed and heavy trawled seas. Our results suggest that the management of bottom-trawling efforts could be an important climate solution.

Crop residue burning (CRB) is a major source of air pollution in many parts of the world, especially Asia. Policymakers, practitioners, and researchers have invested in measuring the extent and impacts of burning and developing interventions to reduce its occurrence. However, any attempt to measure burning, in terms of its extent, impact, or the effectiveness of interventions to reduce it, requires data on where burning occurs. These data are challenging to collect in the field, both in terms of cost and feasibility, because crop-residue fires are short-lived, each covers only a small area, and evidence of burning disappears once fields are tilled. Remote sensing offers a way to observe fields without the complications of on-the-ground monitoring. However, the same features that make CRB hard to observe on the ground also make remote-sensing-based measurements prone to inaccuracies. The extent of crop burning is generally underestimated due to missing observations, while individual plots are often falsely identified as burned due to the local dominance of the practice, a lack of training data on tilled vs. burned plots, and a weak signal-to-noise ratio that makes it difficult to distinguish between the two states. Here, we summarize the current literature on the measurement of CRB and flag five common pitfalls that hinder analyses of CRB with remotely sensed data: inadequate spatial resolution, inadequate temporal resolution, ill-fitted signals, improper comparison groups, and inadequate accuracy assessment. We take advantage of data from ground-based monitoring of CRB in Punjab, India, to calibrate and validate analyses with PlanetScope and Sentinel-2 imagery and illuminate each of these pitfalls. We provide tools to assist others in planning and conducting remote sensing analyses of CRB and stress the need for rigorous validation.

Over the past two decades, sharks have been increasingly recognized among the world’s most threatened wildlife and hence have received heightened scientific and regulatory scrutiny. Yet, the effect of protective regulations on shark fishing mortality has not been evaluated at a global scale. Here we estimate that total fishing mortality increased from at least 76 to 80 million sharks between 2012 and 2019, ~25 million of which were threatened species. Mortality increased by 4% in coastal waters but decreased by 7% in pelagic fisheries, especially across the Atlantic and Western Pacific. By linking fishing mortality data to the global regulatory landscape, we show that widespread legislation designed to prevent shark finning did not reduce mortality but that regional shark fishing or retention bans had some success. These analyses, combined with expert interviews, highlight evidence-based solutions to reverse the continued overexploitation of sharks.

The world’s population increasingly relies on the ocean for food, energy production and global trade^1,2,3, yet human activities at sea are not well quantified^4,5. We combine satellite imagery, vessel GPS data and deep-learning models to map industrial vessel activities and offshore energy infrastructure across the world’s coastal waters from 2017 to 2021. We find that 72–76% of the world’s industrial fishing vessels are not publicly tracked, with much of that fishing taking place around South Asia, Southeast Asia and Africa. We also find that 21–30% of transport and energy vessel activity is missing from public tracking systems. Globally, fishing decreased by 12 ± 1% at the onset of the COVID-19 pandemic in 2020 and had not recovered to pre-pandemic levels by 2021. By contrast, transport and energy vessel activities were relatively unaffected during the same period. Offshore wind is growing rapidly, with most wind turbines confined to small areas of the ocean but surpassing the number of oil structures in 2021. Our map of ocean industrialization reveals changes in some of the most extensive and economically important human activities at sea.

The transition to low-carbon electricity is crucial for meeting global climate goals. However, given the uneven spatial distribution and temporal variability of renewable resources, balancing the supply and demand of electricity will be challenging when relying on close to 100% shares of renewable energy. Here, we use an electricity planning model with hourly supply-demand projections and high-resolution renewable resource maps, to examine whether transcontinental power pools reliably meet the growing global demand for renewable electricity and reduce the system cost. If all suitable sites for renewable energy are available for development, transcontinental trade in electricity reduces the annual system cost of electricity in 2050 by 5–52% across six transcontinental power pools compared to no electricity trade. Under land constraints, if only the global top 10% of suitable renewable energy sites are available, then without international trade, renewables are unable to meet 12% of global demand in 2050. Introducing transcontinental power pools with the same land constraints, however, enables renewables to meet 100% of future electricity demand, while also reducing costs by up to 23% across power pools. Our results highlight the benefits of expanding regional transmission networks in highly decarbonized but land-constrained future electricity systems.

Much of the developed world has adopted substantial, complex agricultural subsidy schemes in an attempt to produce desired rural livelihood and environmental outcomes. Understanding how farmers adjust their production activity in response to farm subsidies is crucial for setting optimal agricultural policy. Whereas standard economic theory suggests that farmers largely adjust production levels in response to prices and marginal subsidy rates, recent work in consumer behavior suggests that average (dis-)incentives may play a relevant role. We use a unique panel covering all farms applying for subsidies in Norway and a flexible deep-learning method to exploit kinks in the subsidy scheme to answer whether farmers respond more to average or marginal subsidies. In contrast to the standard economic theory of production, we find suggestive empirical evidence that farmers respond more to changes in average payments than to changes in marginal payments. We anticipate that our findings on the relevance of average payment levels for farmers' decision making may inspire further theoretical and empirical inquiries into agricultural policy effects. The study also highlights how novel deep-learning tools can be applied for detailed policy analysis and what advantages and challenges come with it. We believe that this approach has substantial potential for analysts and policymakers to evaluate and predict the impacts of policy options.

Climate change is shifting the structure and function of global forests, underscoring the critical need to predict which forests are most vulnerable to a hotter and drier future. We analyzed 6.6 million tree rings from 122 species to assess trees’ sensitivity to water and energy availability. We found that trees growing in wetter portions of their range exhibit the greatest drought sensitivity. To test how these patterns of drought sensitivity influence vulnerability to climate change, we predicted tree growth through 2100. Our results suggest that drought adaptations in arid regions will partially buffer trees against climate change. By contrast, trees growing in the wetter, hotter portions of their climatic range may experience unexpectedly large adverse impacts under climate change.

More system-specific and finer-scale analyses are necessary to better understand how spatio-temporal variability in geophysical forces affects grid planning. Moreover, we need a broader focus on the multi-sectoral implications of decarbonization efforts, including the societal consequences of grid management decisions. Importantly, all these efforts are challenged by the computational requirements of existing power system models, which often limit our ability to characterize uncertainty and scale analyses across larger domains.