Using virtual environments to improve the realism of choice experiments: A case study about coastal erosion management

Choice experiment surveys are commonly used to assess the general public׳s willingness to pay for different levels of environmental quality. However, respondents need to understand what they are valuing or they will make potentially wrong assumptions based on different experiences and frames of reference. Three-dimensional computer generated models or Virtual Environments (VE) have so far seen little use in economics research, probably due to the complexity and cost of developing and delivering them to study participants. The few studies that have used them find that VE are superior to static image presentations in helping people evaluate complex data. For this study we developed virtual environments for a choice experiment about coastal erosion management using free, easy-to-use software and Google Earth© satellite imagery and presented these to respondents as video tours. Our results indicate that the VE treatment reduced choice error, reduced left-right bias and improved respondent engagement and retention when compared with static images. There were also differences in WTP between the two groups.     

Habitat mapping using machine learning-extended kernel-based reclassification of an Ikonos satellite image

The spatial resolution of satellite imagery suitable for earth resources studies has improved from 80 m (Landsat-MSS, launched in 1972) to 0.6 m (QuickBird, launched in 2001). The conventional pixel-based methods developed for medium resolution satellite images are not suitable for classification of very high spatial resolution images, because the spectral responses of particular habitat classes are much more variable. On the other hand, in the original Barnsley–Barr kernel-based reclassification algorithm not only the spectral information of a pixel but also the textural information in the vicinity of the pixel is used when the pixel labeling decision is made. The first step of the kernel reclassification algorithm is to perform an initial classification of the original image. In the second step, the adjacency-event matrices are computed for each pixel according to co-occurrence frequencies of the initial classes in the kernel window. The degree of matching between an adjacency-event matrix corresponding to specific pixel and the set of class-specific template matrices produced during training is the criterion for pixel re-labeling. We extend the original kernel-based reclassification algorithm with a decision tree-based reclassification, simultaneously taking into account the class-specific similarity images, which are a side-product of the original algorithm. The advantage of decision tree-extended approach over the original approach seems to be the ability of the former to consider more input information, thus increasing the Kappa classification accuracy for an Ikonos image of our study area from 0.56 to 0.60, using a nomenclature containing 10 habitat classes.     

Deriving state-and-transition models from an image series of grassland pattern dynamics

We present how state-and-transition models (STMs) may be derived from image data, providing a graphical means of understanding how ecological dynamics are driven by complex interactions among ecosystem events. A temporal sequence of imagery of fine scale vegetation patterning was acquired from close range photogrammetry (CRP) of 1 m quadrats, in a long term monitoring project of Themeda triandra (Forsskal) grasslands in north western Australia. A principal components scaling of image metrics calculated on the imagery defined the state space of the STM, and thereby characterised the different patterns found in the imagery. Using the state space, we were able to relate key events (i.e. fire and rainfall) to both the image data and aboveground biomass, and identified distinct ecological ‘phases’ and ‘transitions’ of the system. The methodology objectively constructs a STM from imagery and, in principle, may be applied to any temporal sequence of imagery captured in any event-driven system. Our approach, by integrating image data, addresses the labour constraint limiting the extensive use of STMs in managing vegetation change in arid and semiarid rangelands.     

Spatial error propagation when computing linear combinations of spectral bands: The case of vegetation indices

The collection of accurate and timely information on land use, crops, forest and vegetation are increasingly based on remote sensing spectral measurements produced by satellites. The most recent spacecrafts like the Earth Observing 1 (EO-1) produce a rich source of information being endowed with hyperspectral sensors that can provide up to 200 or more channels. In many instances such a multivariate signal has to be reduced to just one single value per pixel representing a particular characteristic of land. Linear combinations of bands are the general form of many indices. Since each individual image used to construct indices contains errors, when combined they produce a propagation of errors, a process that can distort a final output map. In this paper we measure the extent of error propagation when building linear vegetation indices. We consider three types of indices: the difference vegetation index (DVI), selected Kauth-Thomas indices (SBI, GVI and WET), and principal components, using benchmarking examples taken from the remote sensing literature. The main implication emerging from these examples is that the SBI and the first principal component are the indices more prone to error propagation. The formalization presented here allows a user to derive measures of error propagation in cases where technical characteristics of a sensor and physical characteristics of a landscape are known. These results can help a user to choose between alternative vegetation indices, and to associate a measure of reliability with such indices.     

Capacity building in tropical coastal resource monitoring in developing countries: A re-appreciation of the oldest remote sensing method

Long-term decadal retrospection in spatio-temporal imagery analyses can only be carried out using aerial photographs, which are still the most detailed remotely sensed data available. Visual interpretation of such imagery is most efficient and inexpensive in the light of ecosystem monitoring research in developing countries, which are often unable to cope with the development or the cost of acquisition of commercial space-borne imaging (e.g. IKONOS, Quickbird). In this light, the present paper explicitly analyses the methodological use of image attributes of air-borne imagery from mangrove forests, and investigates the consistency and constraints of mangrove image attributes in visually interpreted air-borne imagery. Six image attributes are analysed, and their application is illustrated using various mangrove sites in Kenya and Sri Lanka. Comparison of identification keys reveals that minor attributes such as 'ecological position' are informative, and that image attributes for a particular species or genus are apparently less plastic and more widely applicable than formerly assumed. Emphasis on compulsory fieldwork is made and constraints related to reflection and interference, amongst others, are discussed.     

Multisensor monitoring of plume dynamics in the northwestern Mediterranean Sea

We used data from various space-borne sensors to monitor the marine ecosystem in the northwestern Mediterranean Sea, at the Costa Dorada, between the City of Barcelona and the estuary of the river Ebro. The aim of this study was to demonstrate that the combination of different remote sensing data (acquired at different electromagnetic frequencies) allows for an improved monitoring system, in particular for a better monitoring of the marine ecosystem and, hence, a better coastal zone management. We present remote sensing data acquired by the Synthetic Aperture Radar (SAR) and the Along-Track Scanning Radiometer (ATSR) aboard the Second European Remote Sensing Satellite (ERS-2), and by the Sea-viewing Wide Field-of-view Sensor (Sea WiFS) on the SeaStar satellite. By combining the different data we are able to overcome specific drawbacks of the single sensors, like an insufficient temporal coverage, or a strong dependence on weather and daylight conditions. Within the study area two main features have been selected as examples, which are well visible on many of the analysed images. The first one exhibits a higher load of chlorophyll-a and surface-active compounds and a lower sea surface temperature (SST), which is likely to be caused by the plume of the river Llobregat, southwest of Barcelona. It can clearly be seen from the imagery how the river plume is driven along the coast by the local currents. The second feature can be related to cooling water being released from a nuclear power plant and causing turbulence in the water body, which in turn gives rise to signatures visible on the ERS-SAR imagery.     

Landsat Thematic Mapper: Detection of Shifts in Community Composition of Coral Reefs

Abstract: We assembled a time series of 20 Landsat thematic mapper images from 1982 to 1996 for Key Largo, Florida, to ascertain whether satellite imagery can detect temporal changes in coral reef communities. Selected reef and control areas were examined for changes in brightness, spectral reflectance, band ratios, spatial texture, and temporal texture (  pixel-to-pixel change over time). We compared the data to known changes in the reef ecosystem of Carysfort Reef and terrestrial sample sites. Changes in image brightness and spectral-band ratios were suggestive of shifts from coral- to algal-dominated community structure, but the trends were not statistically significant. The spatial heterogeneity of the reef community decreased in the early 1980s at scales consistent with known ecological changes to the coral community on Carysfort Reef. An analysis of pixel-scale variation through time, termed temporal texture, revealed that the shallow reef areas are the most variable in regions of the reef that have experienced significant ecological decline. Thus, the process of reef degradation, which alters both the spatial patterning and variability of pixel brightness, can be identified in unclassified thematic mapper images.     

The spatial pattern of nitrogen cycling in the Adirondack Park, New York

Maps of canopy nitrogen obtained through analysis of high-resolution, hyperspectral, remotely sensed images now offer a powerful means to make landscape-scale to regional-scale estimates of forest N cycling and net primary production (NPP). Moreover, recent research has suggested that the spatial variability within maps of canopy N may be driven by environmental gradients in such features as historic forest disturbance, temperature, species composition, moisture, geology, and atmospheric N deposition. Using the wide variation in these six features found within the diverse forest ecosystems of the 2.5 million ha Adirondack Park, New York, USA, we examined linkages among environmental gradients and three measures of N cycling collected during the 2003 growing season: (1) field survey of canopy N, (2) field survey of soil C:N, and (3) canopy N measured through analysis of two 185 x 7.5 km Hyperion hyperspectral images. These three measures of N cycling strongly related to forest type but related poorly to all other environmental gradients. Further analysis revealed that the spatial pattern in N cycling appears to have distinct inter- and intraspecific components of variability. The interspecific component, or the proportional contribution of species functional traits to canopy biomass, explained 93% of spatial variability within the field canopy N survey and 37% of variability within the soil C:N survey. Residual analysis revealed that N deposition accounted for an additional 2% of variability in soil C:N, and N deposition and historical forest disturbance accounted for an additional 2.8% of variability in canopy N. Given our finding that 95.8% of the variability in the field canopy N survey could be attributed to variation in the physical environment, our research suggests that remotely sensed maps of canopy N may be useful not only to assess the spatial variability in N cycling and NPP, but also to unravel the relative importance of their multiple controlling factors.     

Application of Remote Sensing in Detection of Ocean Oil Pollution

It is well known that the increase of surface tension due to the presence of oil slicks causes the surface wave motion is depressed or even disappeared, thus making the surface electromagnetically smoother. Therefore, the radar backscattered energy is correspondingly decreased. These damping effects is now well understood and such effects enables the oil slicks to be discernible from the radar image. in this paper, we are concerned with the digital technique that effectively delineates the oil slicks pattern from the SAR image. the detected pattern allows us to estimate the coverage of oil spillage. the technique is based on the fact that the oil slicks make a gray value surface in the image which is a concave area with a certain size. in order to identify correctly these oil slicks and suppress the speckle noise and other natural phenomena, an image pyramid with multi-resolution layers is generated sequentially from the original image. Then a top-down approach, which applies both first and second order derivative operators, the Difference of Gaussian (DoG) and the Laplace of Gaussian (LoG), to the image pyramid, is used to detect oil patches. Two ERS-1 SAR images acquired on August 5, 1994 and September 5, 1994, around Taiwan, were used for testing. Results indicate that the proposed method depresses the speckle noise and other sea noise signals, and enhances the oil slicks pattern clearly. the robustness to the empirical parameters introduced in this scheme was also demonstrated.     

Hyperspectral and LiDAR remote sensing of fire fuels in Hawaii Volcanoes National Park.

Alien invasive grasses threaten to transform Hawaiian ecosystems through the alteration of ecosystem dynamics, especially the creation or intensification of a fire cycle. Across sub-montane ecosystems of Hawaii Volcanoes National Park on Hawaii Island, we quantified fine fuels and fire spread potential of invasive grasses using a combination of airborne hyperspectral and light detection and ranging (LiDAR) measurements. Across a gradient from forest to savanna to shrubland, automated mixture analysis of hyperspectral data provided spatially explicit fractional cover estimates of photosynthetic vegetation, non-photosynthetic vegetation, and bare substrate and shade. Small-footprint LiDAR provided measurements of vegetation height along this gradient of ecosystems. Through the fusion of hyperspectral and LiDAR data, a new fire fuel index (FFI) was developed to model the three-dimensional volume of grass fuels. Regionally, savanna ecosystems had the highest volumes of fire fuels, averaging 20% across the ecosystem and frequently filling all of the three-dimensional space represented by each image pixel. The forest and shrubland ecosystems had lower FFI values, averaging 4.4% and 8.4%, respectively. The results indicate that the fusion of hyperspectral and LiDAR remote sensing can provide unique information on the three-dimensional properties of ecosystems, their flammability, and the potential for fire spread.     

Multi-scale marine biodiversity patterns inferred efficiently from habitat image processing

Cost-effective proxies of biodiversity and species abundance, applicable across a range of spatial scales, are needed for setting conservation priorities and planning action. We outline a rapid, efficient, and low-cost measure of spectral signal from digital habitat images that, being an effective proxy for habitat complexity, correlates with species diversity and requires little image processing or interpretation. We validated this method for coral reefs of the Great Barrier Reef (GBR), Australia, across a range of spatial scales (1 m to 10 km), using digital photographs of benthic communities at the transect scale and high-resolution Landsat satellite images at the reef scale. We calculated an index of image-derived spatial heterogeneity, the mean information gain (MIG), for each scale and related it to univariate (species richness and total abundance summed across species) and multivariate (species abundance matrix) measures of fish community structure, using two techniques that account for the hierarchical structure of the data: hierarchical (mixed-effect) linear models and distance-based partial redundancy analysis. Over the length and breadth of the GBR, MIG alone explained up to 29% of deviance in fish species richness, 33% in total fish abundance, and 25% in fish community structure at multiple scales, thus demonstrating the possibility of easily and rapidly exploiting spatial information contained in digital images to complement existing methods for inferring diversity and abundance patterns among fish communities. Thus, the spectral signal of unprocessed remotely sensed images provides an efficient and low-cost way to optimize the design of surveys used in conservation planning. In data-sparse situations, this simple approach also offers a viable method for rapid assessment of potential local biodiversity, particularly where there is little local capacity in terms of skills or resources for mounting in-depth biodiversity surveys.     

A Bayesian approach to multi-source forest area estimation

In efforts such as land use change monitoring, carbon budgeting, and forecasting ecological conditions and timber supply, there is increasing demand for regional and national data layers depicting forest cover. These data layers must permit small area estimates of forest area and, most importantly, provide associated error estimates. This paper presents a model-based approach for coupling mid-resolution satellite imagery with plot-based forest inventory data to produce estimates of probability of forest and associated error at the pixel-level. The proposed Bayesian hierarchical model provides access to each pixel’s posterior predictive distribution allowing for a highly flexible analysis of pixel and multi-pixel areas of interest. The paper presents a trial using multiple dates of Landsat imagery and USDA Forest Service Forest Inventory and Analysis plot data. The results describe the spatial dependence structure within the trial site, provide pixel and multi-pixel summaries of probability of forest land use, and explore discretization schemes of the posterior predictive distributions to forest and non-forest classes. Model prediction results of a holdout set analysis suggest the proposed model provides high classification accuracy, 88%, for the trial site.

Application of Remote Sensing in Detection of Ocean Oil Pollution

It is well known that the increase of surface tension due to the presence of oil slicks causes the surface wave motion is depressed or even disappeared, thus making the surface electromagnetically smoother. Therefore, the radar backscattered energy is correspondingly decreased. These damping effects is now well understood and such effects enables the oil slicks to be discernible from the radar image. in this paper, we are concerned with the digital technique that effectively delineates the oil slicks pattern from the SAR image. the detected pattern allows us to estimate the coverage of oil spillage. the technique is based on the fact that the oil slicks make a gray value surface in the image which is a concave area with a certain size. in order to identify correctly these oil slicks and suppress the speckle noise and other natural phenomena, an image pyramid with multi-resolution layers is generated sequentially from the original image. Then a top-down approach, which applies both first and second order derivative operators, the Difference of Gaussian (DoG) and the Laplace of Gaussian (LoG), to the image pyramid, is used to detect oil patches. Two ERS-1 SAR images acquired on August 5, 1994 and September 5, 1994, around Taiwan, were used for testing. Results indicate that the proposed method depresses the speckle noise and other sea noise signals, and enhances the oil slicks pattern clearly. the robustness to the empirical parameters introduced in this scheme was also demonstrated.     

Conservation Applications of Astronaut Photographs of Earth: Tidal-Flat Loss ( Japan), Elephant Effects on Vegetation (Botswana), and Seagrass and Mangrove Monitoring (Australia)

Abstract: National Aeronautics and Space Administration ( NASA) photographs taken by astronauts from low Earth orbit can provide information relevant to conservation biology. This data source is now more accessible because of improvements in digitizing technology, Internet file transfer, and availability of image processing software. We present three examples of conservation-related projects that benefited from the use of astronaut photographs. First, NASA scientists requested that astronauts photograph the area of the controversial Isahaya Bay reclamation project in Japan. Japanese researchers used photographs from before and after the reclamation as a tool for communication with the public about the effects of tidal-flat loss. The newly acquired images and the availability of high-resolution digital images from NASA archives provided timely public information on the observed changes. Second, we digitally classified and analyzed a Space Shuttle photograph of Chobe National Park in Botswana to identify the locations of woodlands affected by elephants. Field validation later confirmed that areas identified on the image showed evidence of elephant damage. Third, we used a summary map from intensive field surveys of seagrasses in Shoalwater Bay, Australia, as reference data for a supervised classification of a digitized photograph taken from orbit. The classification distinguished seagrasses, sediments, and mangroves with accuracy approximating that in studies using other satellite remote-sensing data. Astronaut photographs are in the public domain, and the database of nearly 400,000 photographs from the late 1960s to the present is available at a single searchable location on the Internet (  http://eol.jsc.nasa.gov/sseop ). These photographs can be used by conservation biologists as a source of general information about the landscape and for quantitative mapping.     

3种人工草地不同植被覆盖度实地测量方法比较

植被覆盖度是植被垂直投影面积占统计区域面积百分比;理想覆盖度实地测量方法耗时短,工具简单,结果准确,受人为因素影响小.以结缕草(Zoysia japonica)、白三叶(Trifolium repens)和雀稗(Paspalum thunbergii)人工草地为研究对象,以照相法测量值为参考,比较样线法、目估法、样针法、点框架法、网格法1×1、网格法5×5、网格点法1×1和网格点法5×5等8种方法测量精度.结果表明:样线法误差小于5%,耗时短,工具简单.目估法误差大于样线法,适用于植被高度低且覆盖度较低或较高样方.样针法误差大于样线法和目估法,误差受植被高度和茎叶硬度影响,适用于植被高度低且硬度较小样方.点框架法、网格法和网格点法误差均较大,一般大于10%.网格法误差受覆盖度和小网格边长影响,误差峰值出现在覆盖度接近65%的样方,误差随着小网格边长增加而增加.8种方法测量值与照相法测量值均呈显著正相关(P <0.05),因此对于类似人工草地,这些方法测量值可通过本文提供的拟合公式转化为真实值.随着样方数量增加,8种方法误差均呈现出先降低后稳定趋势.本研究推荐样线法作为类似人工草地覆盖度实地测量方法;未来应根据覆盖度、植被高度和硬度等特征选择适合的测量方法.(图4表2参46)     

Activity patterns in the terebellid polychaete <Emphasis Type="Italic">Eupolymnia nebulosa</Emphasis> assessed using a new image analysis system

The activity and surface sediment displacement by the terebellid polychaete Eupolymnia nebulosa were monitored, using new image acquisition and image analysis procedures. We used a video sensor mounted on a motorized table, to acquire adjacent images of the whole studied aquarium within about 2 s. These images were then grouped within a composite image acquired every 15 s. Consecutive composite images were compared to infer activity and surface sediment displacement. This procedure proved efficient for E. nebulosa as indicated by (1) the continuity of the tentacles within composite images, and (2) the direct comparison of images and detected activities. There were important temporal changes in the relative importance of the three main types of activity: feeding, tube-building and (both partial and total) emergence from the tube, accounting, respectively, for 75, 15 and 10% of the entire experiment duration. Activity intensity (cm² min^-1) was assessed through the surface of pixels with grey levels differing by more than 20 (on a 0–255 scale) between two consecutive images. Feeding was associated with low activity intensity, whereas tube-building and emergence from the tubes were associated with high activity intensities. Surface sediment displacement mostly resulted from tentacle activity both during feeding and tube-building and was almost zero when worms emerged from their tube. We used our experimental set up to assess spatial changes in activity intensity and particle displacements along the tentacles. Most of the activity occurred within the first 4–6 cm from the tube aperture. Particle loss during their transfer along the tentacles preferentially affected larger particles as postulated by the model of particle selection and transport in tentaculate deposit-feeders. Moreover, the speed of the particles along the tentacles correlated negatively with their size. This contributed to increase in the residence time of larger particles on the tentacles and thus, to further enhance their probability of being lost. The size distributions of particles during feeding and tube-building did not differ significantly although visual observations confirmed the preferential use of larger particles for tube-building. This suggests the existence of a sorting step occurring elsewhere than on the tentacles as already demonstrated for spionid polychaetes.     

Feasibility of using high‐resolution satellite imagery to assess vertebrate wildlife populations

Although remote sensing has been used for >40 years to learn about Earth, use of very high‐resolution satellite imagery (VHR) (<1‐m resolution) has become more widespread over the past decade for studying wildlife. As image resolution increases, there is a need to understand the capabilities and limitations of this exciting new path in wildlife research. We reviewed studies that used VHR to examine remote populations of wildlife. We then determined characteristics of the landscape and the life history of species that made the studies amenable to use of satellite imagery and developed a list of criteria necessary for appropriate use of VHR in wildlife research. From 14 representative articles, we determined 3 primary criteria that must be met for a system and species to be appropriately studied with VHR: open landscape, target organism's color contrasts with the landscape, and target organism is of detectable size. Habitat association, temporal exclusivity, coloniality, landscape differentiation, and ground truthing increase the utility of VHR for wildlife research. There is an immediate need for VHR imagery in conservation research, particularly in remote areas of developing countries, where research can be difficult. For wildlife researchers interested in but unfamiliar with remote sensing resources and tools, understanding capabilities and current limitations of VHR imagery is critical to its use as a conservation and wildlife research tool.     

基于决策树的辽宁省北部沙漠化信息提取研究

以沙漠化问题较突出的辽宁省北部地区为例,选取2007年Landsat 5 TM遥感影像作为基本数据源,通过对影像中耕地、林地、草地、水域等常见地物及典型沙漠化土地进行光谱特征分析和波段间的相互运算,将修改型土壤调整植被指数（MSAVI）、归一化差异水体指数（NDWI）和遥感图像缨帽变换后的土壤亮度指数（SBI）、绿度植被指数（GVI）及湿度指数（WVI）等特征变量融入决策树分类模型后进行分层分离,从而实现对沙漠化信息的高精度提取。结果显示,决策树分类法可排除提取地物时的干扰信息,是保证沙漠化土地信息快速自动提取的方法之一。