429 publications from this institution
The potential to derive indicators of sustainable resource use from satellite remote sensing is discussed. Particular attention focuses on indicators related to land cover condition and type in tropical forest environments. This includes the mapping of forest cover, estimation of biomass and biodiversity as well as the impacts of extreme events such as drought on the forest. Each of these issues is discussed with emphasis placed on the potential to increase the level of information extraction beyond that derived with conventional approaches in order to more usefully inform sustainable development practices.
No abstract is provided for this article.
No abstract is provided for this article.
Conventional image classification routines are often inappropriate for the mapping of continuous phenomena suchy as heathland vegetation. To allow for the natural fuzziness of such an environment, a fuzzy sets algorithm may be used to model the heathland vegetation more appropriately than a classification. The results of this study show that the fuzzy c-means algorithm can be used to discriminate accurately between the end points of a set of continua, and that class membership functions derived from the analysis are sensitive to the botanical composition of the vegetation canopy. Mapping the fuzzy membership functions will therefore enable a more realistic portrayal of the healthland vegetation than a conventional classification.
No abstract is provided for this article.
The effect of spatial, spectral and noise degradations on the accuracy of two thematic labelling scenarios with hyperspectral data was investigated. Although all of the degradations significantly influenced accuracy, the noise content of the data was consistently noted as a major variable affecting the accuracy of both supervised classification and sub-pixel anomaly detection analyses.
Error in the ground reference data set used in studies of land cover change can be a source of bias in the estimation of land cover change and of change detection accuracy. The magnitude of the bias introduced may be very large even if the ground reference data set is of a high accuracy. Sometimes the bias is of a predictable systematic nature and so may be reduced or even removed. The impacts of ground reference data error on the accuracy of estimates of the extent of change and on change detection accuracy were explored with simulated data. In one scenario illustrated, the producer's accuracy of change detection was estimated to be ∼61% when in reality it was 80%, the substantial underestimation of accuracy arising through the use of a ground reference data set with an accuracy of 90%. In the same scenario, the extent of change was also substantially overestimated at 26%, when in reality a change of only 20% had occurred. Reducing the effect of error in ground reference data will enable more accurate estimation of land cover change and a more realistic appraisal of the quality of remote sensing as a source of data on land cover change.
Malaysia’s shoreline is dynamic and constantly changing. Although the shoreline may be mapped accurately from fine spatial resolution imagery, this is an impractical approach for use over large areas. An alternative approach using coarse spatial resolution satellite sensor imagery, addressed here, is to fit a shoreline boundary at sub-pixel scale. This paper investigates the effects of utilizing relatively coarse spatial resolution satellite sensor imagery to produce accurate shoreline maps. For the purposes of this research a 1m spatial resolution IKONOS satellite sensor image was used to define the actual location of the shoreline. This image was degraded to spatial resolutions of 16 m and 32 m, comparable to that of widely used civilian remote sensors. The coarse spatial resolution images derived were used in the evaluation of four methods for shoreline mapping. The conventional method based on hard classification provided an inaccurate and inappropriate representation of the shoreline. Super-resolution methods based on sub-pixel information derived from a soft classification provided accurate and realistic prediction of the shoreline. The most accurate prediction of the shoreline, with RMSE less than 2.1 m and 5.2 m for all the shorelines at 16 m and 32 m spatial resolutions respectively, were derived from a method based on simulated annealing.
Artificial neural networks are attractive for the classification of remotely sensed data. However, a wide range of factors influence the accuracy with which a data set may be classified. In this paper, the effect of four factors on the accuracy with which agricultural crops may be classified from airborne thematic mapper (ATM) data was investigated. These factors related to the dimensionality of the remotely sensed data, the neural network architecture, and the characteristics of the training and testing sets. A total of 288 classifications were performed and their accuracies evaluated. The artificial neural networks were able to classify the data to high accuracies, with kappa coefficients of up to 0.97 obtained, but the accuracy derived was highly dependent on the factors investigated. A log-linear modelling approach was used to evaluate the simultaneous effect of the factors on classification accuracy. Variations in the dimensionality of the data set, as well as the training and testing set characteristics had a significant effect on classification accuracy. The network architecture, specifically the number of hidden units and layers, did not, however, have a significant effect on classification accuracy in this investigation. This highlights the need to consider a broader set of issues than network architecture when using an artificial neural network for image classification.
Remote sensing provides valuable insights into pressing environmental challenges and is a critical tool for driving solutions. In this Primer, we briefly introduce the important role of remote sensing in forest ecology and management, which includes applications as diverse as mapping the distribution of forest ecosystems and characterizing the three-dimensional structure of forests. We describe six key reasons why remote sensing has become an important data source and introduce the different types of sensors (e.g., multispectral and synthetic aperture radar) and platforms (e.g., unmanned aerial vehicles and satellites) that have been used for mapping a diversity of forest variables. The rapid advancement in remote-sensing technology, techniques, and platforms is likely to result in a greater democratization of remote-sensing data to support forest management and conservation in parts of the world where environmental issues are the most urgent. Remote sensing provides valuable insights into pressing environmental challenges and is a critical tool for driving solutions. In this Primer, we briefly introduce the important role of remote sensing in forest ecology and management, which includes applications as diverse as mapping the distribution of forest ecosystems and characterizing the three-dimensional structure of forests. We describe six key reasons why remote sensing has become an important data source and introduce the different types of sensors (e.g., multispectral and synthetic aperture radar) and platforms (e.g., unmanned aerial vehicles and satellites) that have been used for mapping a diversity of forest variables. The rapid advancement in remote-sensing technology, techniques, and platforms is likely to result in a greater democratization of remote-sensing data to support forest management and conservation in parts of the world where environmental issues are the most urgent.
Remote sensing has considerable potential for the provision of information on the distribution of habitats that may be used to inform a variety of activities such as those required through the European Union's Habitats Directive. Such programmes are often resource-limited with a need for innovative methods that optimise resource use. This paper explores two approaches to resource savings when mapping habitats from remotely sensed imagery. The first approach realises that in an area of study often interest is focused on a specific habitat with the remaining land cover classes in the region of no importance. In such circumstances conventional statistical supervised classification analyses may be inefficient and yield a map of sub-optimal accuracy. The second approach seeks to further reduce the training requirements of a supervised classification. For this support vector machine (SVM) based approaches to classification are explored to map coastal saltmarsh habitats in North Norfolk, UK from a Landsat Enhanced Thematic Mapper (ETM+) image. A series of classifications using SVM based approaches and the Maximum Likelihood classifier (MLC) were undertaken. Classification accuracies were significantly higher using the SVM based approaches (e.g., 92.0% overall accuracy) than the MLC (64.8% overall accuracy). The SVM based classifications were demonstrated to be attractive for mapping a priority habitat in that the focus is on the habitat of interest to be mapped throughout the classification process resulting in a reduced need for training data. Moreover, it was shown that this can be further optimised through the use of intelligent training. This approach, based on the use of the support vector data description (SVDD), saved resource requirements even further in that training data were required only for the class of interest and yet still obtained high classification accuracies (95.2% overall accuracy) . The wider adoption of SVM based classification of remotely sensed imagery is advocated for use in conservation activities.
The signal-to-noise ratio (SNR) of remotely sensed imagery has been estimated directly using a variety of image-based methods such as the Homogeneous Area (HA) and Geostatistical (GS) methods. However, previous research has shown that such estimates may be dependent on land cover type. We examine this dependence on land cover type using Compact Airborne Spectrographic Imager (CASI) imagery of an agricultural region in Falmouth, Cornwall. The SNR was estimated using the GS method for six different land covers and a range of wavelengths. Large differences in the SNR existed between land cover types. It follows that single estimates of SNR (e.g. for one land cover) should not be associated with an image (as a whole). It is recommended that either (i) each statistic is reported per land cover type per wavelength or (ii) that an image of local statistics is reported per wavelength. The regression of noise on signal can be used to separate independent noise (intercept) from signal-dependent noise (slope). Variation in the noise and SNR estimates can be used to (i) allow more accurate prediction of the SNR and (ii) provide information on uncertainty.
The role of citizens in mapping has evolved considerably over the last decade.This chapter outlines the background to citizen sensing in mapping and sets the
Valuations of ecosystem services often use data on land cover class areal extent. Area estimates from land cover maps may be biased by misclassification error resulting in flawed assessments and inaccurate valuations. Adjustment for misclassification error is possible for maps subjected to a rigorous validation programme including an accuracy assessment. Unfortunately, validation is rare and/or poorly undertaken as often not regarded as a high priority. The benefit of map validation and hence its value is indicated with two maps. The International Geosphere Biosphere Programme's DISCover map was used to estimate wetland value globally. The latter changed from US$ 1.92trillionyr−1 to US$ 2.79trillionyr−1 when adjusted for misclassification bias. For the conterminous USA, ecosystem services value based on six land cover classes from the National Land Cover Database (2006) changed from US$ 1118billionyr−1 to US$ 600billionyr−1 after adjustment for misclassification bias. The effect of error-adjustment on the valuations indicates the value of map validation to rigorous evidence-based science and policy work in relation to aspects of natural capital. The benefit arising from validation was orders of magnitude larger than mapping costs and it is argued that validation should be a high priority in mapping programmes and inform valuations.
Validation data are often used to evaluate the performance of a trained neural network and used in the selection of a network deemed optimal for the task at-hand. Optimality is commonly assessed with a measure, such as overall classification accuracy. The latter is often calculated directly from a confusion matrix showing the counts of cases in the validation set with particular labelling properties. The sample design used to form the validation set can, however, influence the estimated magnitude of the accuracy. Commonly, the validation set is formed with a stratified sample to give balanced classes, but also via random sampling, which reflects class abundance. It is suggested that if the ultimate aim is to accurately classify a dataset in which the classes do vary in abundance, a validation set formed via random, rather than stratified, sampling is preferred. This is illustrated with the classification of simulated and remotely-sensed datasets. With both datasets, statistically significant differences in the accuracy with which the data could be classified arose from the use of validation sets formed via random and stratified sampling (z = 2.7 and 1.9 for the simulated and real datasets respectively, for both p < 0.05%). The accuracy of the classifications that used a stratified sample in validation were smaller, a result of cases of an abundant class being commissioned into a rarer class. Simple means to address the issue are suggested.
Post-classification comparison techniques are frequently used in studies of change detection. If the classifications used were derived with the use of conventional 'hard' classification techniques, change detection is constrained to the identification of complete changes in class label. This may be inappropriate in circumstances when the land cover conversion is operating at a scale finer than the spatial resolution of the sensor which acquired the imagery and for the detection of land cover modifications. By basing the change detection on the comparison of fuzzy classifications it should be possible to identify partial changes, including both land cover conversion and modification. Fuzzy classifications of Advanced Very High Resolution Radiometer (AVHRR) data were used to identify changes in the apparent position of the forest-savanna transition in West Africa. A comparison of the classifications revealed the variations in the nature and magnitude of land cover change. It was apparent that the migration of the transitional area could be characterized and showed similarity to a simple hypothesized model of land cover change. The comparison of fuzzy classifications was able to provide a richer information base on class membership and its dynamics than that obtainable through the comparison of conventional 'hard' classifications.
