Climate modelling for agroforestry species selection in Yunnan Province,China

Changing climate is likely to impact on both tree species and agroforestry systems in a variety of ways. A multi-model ensemble approach based on ecological niche modelling was used to understand the impact of climate on distribution of agroforestry trees in Yunnan Province of China. Future changes in distribution of 10 agroforestry tree species were projected using an ensemble of climate projections derived from the results of 19 Earth System Models provided by the Coupled Model Inter-comparison Project-Phase 5. Our model explained suitable habitat, and identified potential locations for mixed agroforestry using selected species. The model suggested west and southwest Yunnan as important location for tea and alder-based agroforestry, while southern parts of Yunnan are better suited for tea and hog plum, and northern parts could support walnut-based agroforestry options. Agroforestry is an important adaptation option for climate change, which could benefiting farmers and enhancing environmental conservation and restoration of the landscape.     

全球环境变化视角下的土地覆盖分类系统研究综述

全球环境变化需要精确和最新的区域到全球尺度的土地覆盖数据集以支撑生态系统评估、生物多样性保护、气候变化研究和环境建模。然而,在土地覆盖分类数据集建立过程中,建立科学标准的分类系统至关重要,它影响着数据产品的集成与共享,数据的应用领域与范围。通过对区域尺度、全球尺度及可扩展的FAO土地覆盖分类系统进行评述,指出：① 目前国内外没有普遍认可并广泛应用的标准土地覆盖分类系统,这种分类系统的非标准化影响了数据产品的应用以及对土地覆盖变化的监测;② 复合特征是土地覆盖的固有属性,有效地表达与特征量化复合类型是需要不断努力去解决的问题;③ 我国迫切需要建立一套标准的土地覆盖分类系统,一方面能够与国际土地覆盖产品接轨,另一方面充分体现我国的自然环境特征。     

openModeller: a generic approach to species’ potential distribution modelling

Species’ potential distribution modelling is the process of building a representation of the fundamental ecological requirements for a species and extrapolating these requirements into a geographical region. The importance of being able to predict the distribution of species is currently highlighted by issues like global climate change, public health problems caused by disease vectors, anthropogenic impacts that can lead to massive species extinction, among other challenges. There are several computational approaches that can be used to generate potential distribution models, each achieving optimal results under different conditions. However, the existing software packages available for this purpose typically implement a single algorithm, and each software package presents a new learning curve to the user. Whenever new software is developed for species’ potential distribution modelling, significant duplication of effort results because many feature requirements are shared between the different packages. Additionally, data preparation and comparison between algorithms becomes difficult when using separate software applications, since each application has different data input and output capabilities. This paper describes a generic approach for building a single computing framework capable of handling different data formats and multiple algorithms that can be used in potential distribution modelling. The ideas described in this paper have been implemented in a free and open source software package called openModeller. The main concepts of species’ potential distribution modelling are also explained and an example use case illustrates potential distribution maps generated by the framework.     

Nonlinear transformation of term frequencies for term weighting in text categorization

In automatic text categorization, the influence of features on the decision is set by the term weights which are conventionally computed as the product of term frequency and collection frequency factors. The raw form of term frequencies or their logarithmic forms are generally used as the term frequency factor whereas the leading collection frequency factors take into account the document frequency of each term. In this study, it is firstly shown that the best-fitting form of the term frequency factor depends on the distribution of term frequency values in the dataset under concern. Taking this observation into account, a novel collection frequency factor is proposed which considers term frequencies. Five datasets are firstly tested to show that the distribution of term frequency values is task dependent. The proposed method is then proven to provide better F₁ scores compared to two recent approaches on majority of the datasets considered. It is confirmed that the use of term frequencies in the collection frequency factor is beneficial on tasks which does not involve highly repeated terms. It is also shown that the best F₁ scores are achieved on majority of the datasets when smaller number of features are considered.     

中国长时间序列地表冻融状态数据集

地表冻结和融化状态的转换表征着陆地表层过程的休眠和活跃,亦可作为水文循环和生态系统活动的“开关”。本文主要介绍根据双指标冻融状态分类算法和决策树算法制备的中国区域长时间序列地表冻融数据集(分别为1978~2015年与1987~2009年),该数据集可用于分析中国区域地表冻融循环的开始/结束日期、冻结期、冻结天数、冻结范围等指标的时空分布和变化趋势,可为中国区域的冰冻圈和水文相关研究,以及气候变化分析提供重要的数据支持。
     

Predicting the environment from social media: A collective classification approach

We propose a method which uses Flickr tags to predict a wide variety of environmental features, such as climate data, land cover categories, species occurrence, and human assessments of scenicness. The role of Flickr tags in our method is two-fold. First, we show that Flickr tags capture information which is highly complementary to what is found in traditional structured environmental datasets. By combining Flickr tags with traditional datasets, we can thus make more accurate predictions than is possible using either Flickr tags or traditional datasets alone. Second, we propose a collective prediction model which crucially relies on Flickr tags to define a neighbourhood structure. The use of a collective prediction formulation is motivated by the fact that most environmental features are strongly spatially autocorrelated. While this suggests that geographic distance should play a key role in determining neighbourhoods, we show that considerable gains can be made by additionally taking Flickr tags and traditional data into consideration.     

Standardized ecological indicators to assess aquatic food webs: The ECOIND software plug-in for Ecopath with Ecosim models

Ecological indicators are useful tools to analyse and communicate historical changes in ecosystems and plausible future scenarios while evaluating environmental status. Here we introduce a new plug-in to the Ecopath with Ecosim (EwE) food web modelling approach, which is widely used to quantitatively describe aquatic ecosystems. The plug-in (ECOIND) calculates standardized ecological indicators. We describe the primary functionality of ECOIND and provide an example of its application in both static and temporal-spatial dynamic modelling, while we highlight several related features including a new taxonomy input database (species traits) and the ability to analyse input uncertainty on output results. ECOIND adds new capabilities to the widely used EwE food web modelling approach and enables broadening its applications into biodiversity and conservation-based frameworks to contribute to integrated ecosystem analyses.     

Modelling the long term water yield impact of wildfire and other forest disturbance in Eucalypt forests

Disturbance of forested catchments by fire, logging, or other natural or human induced events that alter the evapotranspiration regime may be a substantial threat to domestic, environmental and industrial water supplies. This paper describes the physically-based modelling of the long term changes in water yield from two wildfire affected catchments in north-eastern Victoria, Australia, and of fire and climate change scenarios in Melbourne's principal water supply catchment. The effect of scale, data availability and quality, and of forest species parameterisation are explored. The modelling demonstrates the importance of precipitation inputs, with Nash and Sutcliffe Coefficients of Efficiency of predicted versus observed monthly flows increasing from 0.5 to 0.8 with a higher density of rainfall stations, and where forest types are well parameterised. Total predicted flow volumes for the calibrations were within 1% of the observed for the Mitta Mitta River catchment and <4% for the Thomson River, but almost ?10% for the less well parameterised Tambo River. Despite the issues of data availability simulations demonstrated the potential for significant impacts to water supply in SE Australia from wildfire and climate change. For example, for the catchments modelled the moderate climate change impact on water yield was more pronounced than the worst fire scenario. Both modelled cases resulted in long term water yield declines exceeding 20%, with the climate change impact nearing 30%. A simulation using observed data for the first four post-fire years at the Mitta Mitta River catchment showed Macaque was able to accurately predict total flow.     

Dynamic models for bird population—A parameter-varying partial differential equation identification approach

In order to study the global decline of biodiversity, accurate models of animal population dynamics are required. In this paper, this challenging problem of biodiversity decline analysis is tackled by modelling the dynamics of bird populations. More specifically, a new data-driven modelling of bird population dynamics is suggested which resorts to a parameter-varying partial differential equation (PDE) model, the Galerkin method and the proper orthogonal decomposition. The parameter-varying formulation allows us to introduce in models prior information such as temperatures or landscape patterns. Hence, such models can be used to study the impact on biodiversity of the global warming or the agricultural intensification, for instance. In order to deal with the specific conditions of ecological applications, a specific attention is paid to the initialization as well as the implementation of an iterative identification procedure based on 3D partial moments and a Levenberg–Marquardt algorithm. These tools are tested on the data-driven modelling of the population of European Stonechat Saxicola torquatus, a European common bird species, by using data from the national French Breeding Bird Survey and the CORINE Land Cover. The perspectives are to model specific community of birds in order to evaluate the effect of global changes in population trends and to develop tools to help decision makers take into account biodiversity goals into public policies.     

Modelling with knowledge: A review of emerging semantic approaches to environmental modelling

Models, and to a lesser extent datasets, embody sophisticated statements of environmental knowledge. Yet, the knowledge they incorporate is rarely self-contained enough for them to be understood and used – by humans or machines – without the modeller's mediation. This severely limits the options in reusing environmental models and connecting them to datasets or other models. The notion of “declarative modelling” has been suggested as a remedy to help design, communicate, share and integrate models. Yet, not all these objectives have been achieved by declarative modelling in its current implementations.Semantically aware environmental modelling is a way of designing, implementing and deploying environmental datasets and models based on the independent, standardized formalization of the underlying environmental science. It can be seen as the result of merging the rationale of declarative modelling with modern knowledge representation theory, through the mediation of the integrative vision of a Semantic Web. In this paper, we review the present and preview the future of semantic modelling in environmental science: from the mediation approach, where formal knowledge is the key to automatic integration of datasets, models and analytical pipelines, to the knowledge-driven approach, where the knowledge is the key not only to integration, but also to overcoming scale and paradigm differences and to novel potentials for model design and automated knowledge discovery.     

Untangling drivers of species distributions: Global sensitivity and uncertainty analyses of MaxEnt

Untangling drivers of systems and uncertainty for species distribution models (SDMs) is important to provide reliable predictions that are useful for conservation campaigns. This is particularly true for species whose habitat is threatened by climate change that enhances the uncertainty in future species distributions. Global sensitivity and uncertainty analyses (GSUA) is a robust method to globally investigate the uncertainty of SDMs and the importance of species distributions' drivers in space and time.Here we apply GSUA to MaxEnt that is one of the popular presence-only SDMs. We consider the Snowy Plover (Charadrius alexandrinus nivosus) (SP) in Florida that is a shorebird whose habitat is affected by sea level rise due to climate change. The importance of intrinsic and exogenous input factors to the uncertainty of the species distribution is evaluated for MaxEnt. GSUA is applied for three projections of the habitat (2006, 2060, and 2100) according to the A1B sea level rise scenario. The large land cover variation determines a moderate decrease in habitat suitability in 2060 and 2100 prospecting a low risk of decline for the SP. The regularization parameter for the environmental features, the uncertainty into the classification of salt-marsh, transitional marsh, and ocean beach, and the maximum number of iterations for the model training are in this order the most important input factors for the average habitat suitability. These results are related to the SP but, in general MaxEnt appears as a very non-linear model where uncertainty mostly derives from the interactions among input factors.The uncertainty of the output is a species-specific variable. Thus, GSUA need be performed for each case considering local exogenous input factors of the model. GSUA allows quantitative informed species-management decisions by providing scenarios with controlled uncertainty and confidence over factors' importance that can be used by resource managers.     

Modelling local distribution of an Arctic dwarf shrub indicates an important role for remote sensing of snow cover

Despite the intensive research effort directed at predicting the effects of climate change on plants in the Arctic, the impact of environmental change on species' distributions remains difficult to quantify. Predictive habitat distribution models provide a tool to predict the geographical distribution of a species based on the ecological gradients that determine it, and to estimate how the distribution of a species might respond to environmental change. Here, we present a model of the distribution of the dwarf shrub Dryas octopetala L. around the fjord Kongsfjorden, Svalbard. The model was built from field observations, an Advanced Space-borne Thermal Emission and Reflection Radiometer (ASTER) image, a GIS database containing environmental data at a spatial resolution of 20 m, and relied on generalized linear models (GLMs). We used a logistic GLM to predict the occurrence of the species and a Gaussian GLM to predict its abundance at the sites where it occurred. Temperature and topographical exposure and inclination of a site appeared to promote both the occurrence and the abundance of D. octopetala. The occurrence of the species was additionally negatively influenced by snow and water cover and topographical exposure towards the north, whereas the abundance of the species appeared lower on calciferous substrates. Validation of the model using independent data and the resulting distribution map showed that they successfully recover the distribution of D. octopetala in the study area (κ = 0.46, AUC = 0.81 for the logistic GLM [n = 200], r² = 0.29 for the Gaussian GLM [n = 36]). The results further highlight that models predicting the local distribution of plant species in an Arctic environment would greatly benefit from data on the distribution and duration of snow cover. Furthermore, such data are necessary to make quantitative estimates for the impact of changes in temperature and winter precipitation on the distribution of plants in the Arctic.     

The uncertain impact of climate change on forest ecosystems – How qualitative modelling can guide future research for quantitative model development

Climate change could significantly alter forest productivity and climax states. Hence modelling productivity under climate change will need to account for many alternative ecosystem states. We apply qualitative modelling to identify the most likely ecosystem representations for a well-researched Tasmanian forest. Its main ecosystem is a tiered forest with rainforest, wet sclerophyll and myrtaceae components. Interactions between these components are uncertain, especially under additional pressures from climate change. Qualitative modelling is a structured method to analyse these uncertainties. We identify the most appropriate models and research efforts for model development. Further, we identify research needs for interactions between root pathogens and forest components, with research on some impacts of system components on fire being ruled out. The qualitative modelling approach applied here was useful in identifying research priorities for modelling complex ecosystems, even under uncertain system understanding or deficiencies in quantitative data.     

Identifying potential areas of understorey coffee in Ethiopia’s highlands using predictive modelling

Coffee production is one of the main economic activities in Ethiopia, representing about 40% of the country’s economy. Coffee is particularly important in the Ethiopian highlands, where appropriate climate allows higher productivity and quality. The Ethiopian highlands also host an outstanding biodiversity, being considered one of the world’s most important biodiversity hotspots. In this context, conciliating agricultural practices with biodiversity conservation is a priority task for researchers and other stakeholders. However, identifying and mapping understorey coffee plantations in Ethiopian highlands is particularly challenging due to the presence of scattered exotic trees and the characteristics of understorey cultivation. In this research, we mapped potential areas of understorey coffee using predictive modelling and evaluated how projected changes in climate would affect the suitability of coffee production in the study area. Landscape maps, which were mapped using remote-sensing data based on object-based image analysis, remotely sensed spectral vegetation indices, and climatic variables were used to delineate probability maps showing the most likely location of understorey coffee. Normalized difference vegetation index and maximum temperature and precipitation were considered the best predictors for explaining the spatial distribution of understorey coffee. The accuracy of the probability map was validated based on existing understorey coffee areas mapped during field surveys. In addition, we show that potential changes in temperature and precipitation by 2050 are likely to shift suitable areas of understorey coffee to higher altitudes, affecting the landscape changes dynamics in the region.     

An overview of 21 global and 43 regional land-cover mapping products

Land-cover (LC) products, especially at the regional and global scales, comprise essential data for a wide range of environmental studies affecting biodiversity, climate, and human health. This review builds on previous compartmentalized efforts by summarizing 23 global and 41 regional LC products. Characteristics related to spatial resolution, overall accuracy, time of data acquisition, sensor used, classification scheme and method, support for LC change detection, download location, and key corresponding references are provided. Operational limitations and uncertainties are discussed, mostly as a result of different original modelling outcomes. Upcoming products are presented and future prospects towards increasing usability of different LC products are offered. Despite the common realization of product usage by non-experts, the remote-sensing community has not fully addressed the challenge. Algorithmic development for the effective representation of inherent product limitations to facilitate proper usage by non-experts is necessary. Further emphasis should be placed on international coordination and harmonization initiatives for compatible LC product generation. We expect the applicability of current and future LC products to increase, especially as our environmental understanding increases through multi-temporal studies.     

Random forests to evaluate biotic interactions in fish distribution models

Previous research indicated that high predictive performance in species distribution modelling can be obtained by combining both biotic and abiotic habitat variables. However, models developed for fish often only address physical habitat characteristics, thus omitting potentially important biotic factors. Therefore, we assessed the impact of biotic variables on fish habitat preferences in four selected stretches of the upper Cabriel River (E Spain). The occurrence of Squalius pyrenaicus and Luciobarbus guiraonis was related to environmental variables describing biotic interactions (inferred by relationships among fish abundances) and channel hydro-morphological characteristics. Random Forests (RF) models were trained and then validated using independent datasets. To build RF models, the conditional variable importance was used together with the model improvement ratio technique. The procedure showed effectiveness in identifying a parsimonious set of not correlated variables, which minimize noise and improve model performance in both training and validation phases. Water depth, channel width, fine substrate and water-surface gradient were selected as most important habitat variables for both fish. Results showed clear habitat overlapping between fish species and suggest that competition is not a strong factor in the study area.     

Agricultural production systems modelling and software: Current status and future prospects

During the past decade, the application of agricultural production systems modelling has rapidly expanded while there has been less emphasis on model improvement. Cropping systems modelling has become agricultural modelling, incorporating new capabilities enabling analyses in the domains of greenhouse gas emissions, soil carbon changes, ecosystem services, environmental performance, food security, pests and disease losses, livestock and pasture production, and climate change mitigation and adaptation. New science has been added to the models to support this broadening application domain, and new consortia of modellers have been formed that span the multiple disciplines.There has not, however, been a significant and sustained focus on software platforms to increase efficiency in agricultural production systems research in the interaction between the software industry and the agricultural modelling community. This paper describes the changing agricultural modelling landscape since 2002, largely from a software perspective, and makes a case for a focussed effort on the software implementations of the major models.     

An approximate representation of hypercliques

A hyperclique is an itemset containing items that are strongly correlated with each other, based on a user-specified threshold. Hypercliques (HC) have been successfully used in a number of applications, for example clustering and noise removal. Even though the HC collection has been shown to respond well to datasets with skewed support distribution and low support threshold, it may still grow very large for dense datasets and lower h-confidence threshold. Recently, we proposed a condensed representation of HC, the Non-Derivable Hypercliques (NDHC). NDHC was shown to have substantial advantages over HC, especially for dense datasets and lower h-confidence values. In this paper, we propose an approximation of the NDHC collection, called Approximate Non-Derivable Hypercliques (ANDHC). This collection is a subset of NDHC and is generated based on a user-input error. We present an efficient algorithm to mine all ANDHC sets, and an algorithm to approximately derive the HC collection from the ANDHC collection. Through experiments with several datasets and parameter combinations, the ANDHC collection is shown to have significant collection size reduction from NDHC even for small error values. It is also shown that based on ANDHC we can generate a good approximation of the entire HC collection. Finally, we experiment with applying our proposed collections to enhance clustering with encouraging results.     

丹顶鹤繁殖地气候数据特征的聚类分析

以1981—1990年我国东北地区繁殖的候鸟野生丹顶鹤为例, 调查其繁殖地分布区域内的月平均气温、月平均最高温度、月平均最低温度、月平均降水和月平均净辐射五个气候因子数据的基础上, 采用分区间统计及基于归一化距离的分层聚类和聚类融合等理论和方法进行相关气候数据的特征分析, 获得丹顶鹤在东北繁殖地的最适宜栖息时间为5~9月份。这一结果与东北地区丹顶鹤的生物学和生态学特征相吻合。与以往文献中用于物种分布预测的气候因子提取方法不同的是, 该方法完全依赖于预测物种在调查分布区内的气候数据, 通过数据挖掘与数据处理而获得, 而不是通过预测物种的生物学与生态学特征及其在分布区内的气候因子相关性分析得到。这一研究方法具有一般性, 将为未来气候变化对候鸟在东北地区栖息地的影响的研究提供基础。