Near-term forecasts of stream temperature using deep learning and data assimilation in support of management decisions

Deep learning (DL) models are increasingly used to make accurate hindcasts of management-relevant variables, but they are less commonly used in forecasting applications. Data assimilation (DA) can be used for forecasts to leverage real-time observations, where the difference between model predictions and observations today is used to adjust the model to make better predictions tomorrow. In this use case, we developed a process-guided DL and DA approach to make 7-day probabilistic forecasts of daily maximum water temperature in the Delaware River Basin in support of water management decisions. Our modeling system produced forecasts of daily maximum water temperature with an average root mean squared error (RMSE) from 1.1 to 1.4°C for 1-day-ahead and 1.4 to 1.9°C for 7-day-ahead forecasts across all sites. The DA algorithm marginally improved forecast performance when compared with forecasts produced using the process-guided DL model alone (0%–14% lower RMSE with the DA algorithm). Across all sites and lead times, 65%–82% of observations were within 90% forecast confidence intervals, which allowed managers to anticipate probability of exceedances of ecologically relevant thresholds and aid in decisions about releasing reservoir water downstream. The flexibility of DL models shows promise for forecasting other important environmental variables and aid in decision-making.     

Boosting large-scale river connectivity restoration by planning for the presence of unrecorded barriers

Conservation decisions are invariably made with incomplete data on species’ distributions, habitats, and threats, but frameworks for allocating conservation investments rarely account for missing data. We examined how explicit consideration of missing data can boost return on investment in ecosystem restoration, focusing on the challenge of restoring aquatic ecosystem connectivity by removing dams and road crossings from rivers. A novel way of integrating the presence of unmapped barriers into a barrier optimization model was developed and applied to the U.S. state of Maine to maximize expected habitat gain for migratory fish. Failing to account for unmapped barriers during prioritization led to nearly 50% lower habitat gain than was anticipated using a conventional barrier optimization approach. Explicitly acknowledging that data are incomplete during project selection, however, boosted expected habitat gains by 20–273% on average, depending on the true number of unmapped barriers. Importantly, these gains occurred without additional data. Simply acknowledging that some barriers were unmapped, regardless of their precise number and location, improved conservation outcomes. Given incomplete data on ecosystems worldwide, our results demonstrate the value of accounting for data shortcomings during project selection.     

手机定位人口数据在应急救援中的应用

随着智能手机的普及,基于手机定位的人口数据在突发性灾害应急救援领域发挥着日益重要的作用。以近几年地震灾区为主要研究对象,选择灾区中有代表性的旅游景区、县城及乡镇,获取这些区域手机定位的人口数据,同时采用传统方法估算区域内人口数量,并进行差异性对比。研究结果表明:差异性最大的是旅游景区、其次是县城和乡镇;在人口年度变化上,旅游景区的人口在7、8月份达到高峰,乡镇人口高值区出现在2月份,县城人口年度差异无明显规律。     

基于改进的非线性GM(1,1)模型的职业病预测研究

为了更准确地预测职业病，在传统GM(1,1)模型的基础上，提出改进的非线性GM(1,1)模型，在传统的GM(1,1)模型中引入弱化算子，将紧邻均值与原始数据之间的线性假设改为非线性假设，提高曲线的拟合度。以2005—2014年的全国职业病例数为研究对象，进行数据拟合和预测分析，其中以2014年职业病例数作为验证数据，并利用后验差比值和小误差概率2个参数，检验该改进模型的预测精度。由应用实例的分析结果可知:在职业病发病趋势的预测方面，改进的非线性GM(1,1)模型的预测精度提高到一级，曲线拟合度较高，预测得到2015年的职业病例数为34 900例。     

Factors influencing property selection for conservation revolving funds

Finding sustainable ways to increase the amount of private land protected for biodiversity is challenging for many conservation organizations. In some countries, organizations use revolving‐fund programs, whereby land is purchased and then sold to conservation‐minded owners under condition they enter into a conservation covenant or easement. The sale proceeds are used to purchase, protect, and sell additional properties, incrementally increasing the amount of protected private land. Because the effectiveness of this approach relies on selecting appropriate properties, we explored factors currently considered by practitioners and how these are integrated into decision making. We conducted exploratory, semistructured interviews with managers from each of the 5 major revolving funds in Australia. Responses indicated although conservation factors are important, financial and social factors are also highly influential. A major determinant was whether the property could be resold within a reasonable period at a price that replenishes the fund. To facilitate resale, often selected properties include the potential for the construction of a dwelling. Practitioners face with clear trade‐offs between conservation, financial, amenity, and other factors in selecting properties and 3 main challenges: recovering the costs of acquisition, protection, and resale; reselling the property; and meeting conservation goals. Our findings suggest the complexity of these decisions may constrain revolving‐fund effectiveness. Drawing from participant responses, we identified potential strategies to mitigate these risks, such as providing adequate recreational space without jeopardizing ecological assets. We suggest managers could benefit from a shared‐learning and adaptive approach to property selection given the commonalities between programs. Understanding how practitioners deal with complex decisions in the implementation of revolving funds helps identify future research to improve the performance of this conservation tool.     

The need for spatially explicit quantification of benefits in invasive‐species management

Worldwide, invasive species are a leading driver of environmental change across terrestrial, marine, and freshwater environments and cost billions of dollars annually in ecological damages and economic losses. Resources limit invasive‐species control, and planning processes are needed to identify cost‐effective solutions. Thus, studies are increasingly considering spatially variable natural and socioeconomic assets (e.g., species persistence, recreational fishing) when planning the allocation of actions for invasive‐species management. There is a need to improve understanding of how such assets are considered in invasive‐species management. We reviewed over 1600 studies focused on management of invasive species, including flora and fauna. Eighty‐four of these studies were included in our final analysis because they focused on the prioritization of actions for invasive species management. Forty‐five percent (n = 38) of these studies were based on spatial optimization methods, and 35% (n = 13) accounted for spatially variable assets. Across all 84 optimization studies considered, 27% (n = 23) explicitly accounted for spatially variable assets. Based on our findings, we further explored the potential costs and benefits to invasive species management when spatially variable assets are explicitly considered or not. To include spatially variable assets in decision‐making processes that guide invasive‐species management there is a need to quantify environmental responses to invasive species and to enhance understanding of potential impacts of invasive species on different natural or socioeconomic assets. We suggest these gaps could be filled by systematic reviews, quantifying invasive species impacts on native species at different periods, and broadening sources and enhancing sharing of knowledge.     

Progress and challenges in consolidating the management of Amazonian protected areas and indigenous territories

Effective management refers to the ability of a protected area or indigenous territory to meet its objectives, particularly as they relate to the protection of biodiversity and forest cover. Effective management is achieved through a process of consolidation, which among other things requires legally protecting sites, integrating sites into land‐use planning, developing and implementing management and resource‐use plans, and securing long‐term funding to pay for recurrent costs. Effectively managing all protected areas and indigenous territories in the Amazon may be needed to avoid a deforestation tipping point beyond which regional climatic feedbacks and global climate change interact to catalyze irreversible drying and savannization of large areas. At present, protected areas and indigenous territories cover 45.5% (3.55 million km²) of the Amazon, most of the 60–70% forest cover required to maintain hydrologic and climatic function. Three independent evaluations of a long‐term large‐scale philanthropic initiative in the Amazon yielded insights into the challenges and advances toward achieving effective management of protected areas and indigenous territories. Over the life of the initiative, management of sites has improved considerably, particularly with respect to management planning and capacity building, but few sites are effectively managed and many lack sufficient long‐term financing, adequate governance, support of nongovernmental organizations, and the means to withstand economic pressures. The time and money required to complete consolidation is still poorly understood, but it is clear that philanthropic funding is critical so long as essential funding needs are not met by governments and other sources, which could be on the order of decades. Despite challenges, it is encouraging that legal protection has expanded greatly and management of sites is improving steadily. Management of protected areas in other developing countries could be informed by improvements that have occurred in Amazonian countries.     

Patterns and biases of climate change threats in the IUCN Red List

International Union for Conservation of Nature (IUCN) Red List assessments rely on published data and expert inputs, and biases can be introduced where underlying definitions and concepts are ambiguous. Consideration of climate change threat is no exception, and recently numerous approaches to assessing the threat of climate change to species have been developed. We explored IUCN Red List assessments of amphibians and birds to determine whether species listed as threatened by climate change display distinct patterns in terms of habitat occupied and additional nonclimatic threats faced. We compared IUCN Red List data with a published data set of species’ biological and ecological traits believed to infer high vulnerability to climate change and determined whether distributions of climate change‐threatened species on the IUCN Red List concur with those of climate change‐threatened species identified with the trait‐based approach and whether species possessing these traits are more likely to have climate change listed as a threat on the IUCN Red List. Species in some ecosystems (e.g., grassland, shrubland) and subject to particular threats (e.g., invasive species) were more likely to have climate change as a listed threat. Geographical patterns of climate change‐threatened amphibians and birds on the IUCN Red List were incongruent with patterns of global species richness and patterns identified using trait‐based approaches. Certain traits were linked to increases or decreases in the likelihood of a species being threatened by climate change. Broad temperature tolerance of a species was consistently related to an increased likelihood of climate change threat, indicating counterintuitive relationships in IUCN assessments. To improve the robustness of species assessments of the vulnerability or extinction risk associated with climate change, we suggest IUCN adopt a more cohesive approach whereby specific traits highlighted by our results are considered in Red List assessments. To achieve this and to strengthen the climate change‐vulnerability assessments approach, it is necessary to identify and implement logical avenues for further research into traits that make species vulnerable to climate change (including population‐level threats).     

Estimating abundance without recaptures of marked pallid sturgeon in the Mississippi River

Abundance estimates are essential for assessing the viability of populations and the risks posed by alternative management actions. An effort to estimate abundance via a repeated mark‐recapture experiment may fail to recapture marked individuals. We devised a method for obtaining lower bounds on abundance in the absence of recaptures for both panmictic and spatially structured populations. The method assumes few enough recaptures were expected to be missed by random chance. The upper Bayesian credible limit on expected recaptures allows probabilistic statements about the minimum number of individuals present in the population. We applied this method to data from a 12‐year survey of pallid sturgeon (Scaphirhynchus albus) in the lower and middle Mississippi River (U.S.A.). None of the 241 individuals marked was recaptured in the survey. After accounting for survival and movement, our model‐averaged estimate of the total abundance of pallid sturgeon ≥3 years old in the study area had a 1%, 5%, or 25% chance of being <4,600, 7,000, or 15,000, respectively. When we assumed fish were distributed in proportion to survey catch per unit effort, the farthest downstream reach in the survey hosted at least 4.5–15 fish per river kilometer (rkm), whereas the remainder of the reaches in the lower and middle Mississippi River hosted at least 2.6–8.5 fish/rkm for all model variations examined. The lower Mississippi River had an average density of pallid sturgeon ≥3 years old of at least 3.0–9.8 fish/rkm. The choice of Bayesian prior was the largest source of uncertainty we considered but did not alter the order of magnitude of lower bounds. Nil‐recapture estimates of abundance are highly uncertain and require careful communication but can deliver insights from experiments that might otherwise be considered a failure.     

基于GM-SVR的小样本条件下化工设备可靠性预测

为了准确预测化工设备可靠性趋势，针对化工设备失效寿命数据为小样本的情形，基于灰色估计法与支持向量回归机在小样本数据处理中的优势，建立了失效寿命时间服从三参数威布尔分布的化工设备可靠性模型；结合GM（1，1）和SVR对模型进行参数估计，在压缩机可靠性分析中进行了实例应用，对比分析了最小二乘法、灰色估计法和GM-SVR的估计效果。研究结果表明:GM-SVR对威布尔分布参数的估计精度明显优于最小二乘法和灰色估计法，可以有效地应用于化工设备失效数据为小样本时的可靠性预测。