Inter-sensor relationships between IKONOS and Landsat-7 ETM+ NDVI data in three ecoregions of Africa

The goal of this research was to establish inter-sensor relationships between IKONOS and Landsat-7 ETM+ data. Dry and wet season images were acquired on the same date or about the same date from IKONOS and ETM+ sensors to enable direct comparison between the two distinctly different data types. The images were from three distinct ecoregions located in African rainforests and savannas that encompass a wide range of land use/land cover classes and ecological units. The IKONOS NDVI had a high degree of correlation with ETM+ NDVI with R ² values between 0.67 and 0.72. Inter-sensor model equations relating IKONOS NDVI with ETM+ NDVI were determined. The characteristics that contribute to the increased sensitivity in dynamic ranges of IKONOS NDVI relative to ETM+ NDVI were attributed to: (1) radiometric resolution that adds more bits per data point (11-bit IKONOS data as opposed to 8-bit ETM+); and (2) spatial resolution that helped in resolving spectral characteristics at micro landscape units. Spectral bandwidths of the two sensors had no effect on the dynamic ranges of NDVIs. Overall, the IKONOS data showed greater sensitivity to landscape units and ecological characteristics when compared with Landsat-7 ETM+ data. Across ecoregions and land use/land cover classes, the IKONOS NDVI dynamic range (?0.07 to 0.71) was considerably greater than the ETM+ NDVI dynamic range (?0.24 to 0.46). IKONOS data explained greater variability (R ²=0.73) in agroforest biomass when compared with ETM+ data (R ²=0.66). The inter-sensor relationships presented in this paper are expected to facilitate better understanding and proper interpretation of terrestrial characteristics studied using multiple sensors over time periods.     

Mapping forest degradation in the Eastern Amazon from SPOT 4 through spectral mixture models

In this paper, we present a methodology to map classes of degraded forest in the Eastern Amazon. Forest degradation field data, available in the literature, and 1-m resolution IKONOS image were linked with fraction images (vegetation, nonphotosynthetic vegetation (NPV), soil and shade) derived from spectral mixture models applied to a Satellite Pour L'observation de la Terre (SPOT) 4 multispectral image. The forest degradation map was produced in two steps. First, we investigated the relationship between ground (i.e., field and IKONOS data) and satellite scales by analyzing statistics and performing visual analyses of the field classes in terms of fraction values. This procedure allowed us to define four classes of forest at the SPOT 4 image scale, which included: intact forest; logged forest (recent and older logged forests in the field); degraded forest (heavily burned, heavily logged and burned forests in the field); and regeneration (old heavily logged and old heavily burned forest in the field). Next, we used a decision tree classifier (DTC) to define a set of rules to separate the forest classes using the fraction images. We classified 35% of the forest area (2097.3 km²) as intact forest. Logged forest accounted for 56% of the forest area and 9% of the forest area was classified as degraded forest. The resultant forest degradation map showed good agreement (86% overall accuracy) with areas of degraded forest visually interpreted from two IKONOS images. In addition, high correlation (R²=0.97) was observed between the total live aboveground biomass of degraded forest classes (defined at the field scale) and the NPV fraction image. The NPV fraction also improved our ability to mapping of old selectively logged forests.     

Classification of Landsat images based on spectral and topographic variables for land-cover change detection in Zagros forests

Detection of land-cover changes through time can be complicated because of sensor-specific differences in spatial and spectral resolutions; classified land-cover changes can be due to either real changes on the ground or a switch in sensors used to collect data. This study focused on two objectives: (1) selecting the best predictor variables for the classification of semi-arid Zagros forests given the characteristics of the study area and available data sets and (2) evaluating the application of the random forest (RF) algorithm as a unified technique for the classification of data sets acquired from different sensors. Three images of the same study area were acquired from the Landsat-5 Thematic Mapper (TM) sensor in 2009, the Landsat-7 Enhanced Thematic Mapper (ETM+) sensor with Scan Line Corrector (SLC) in 1999 and the Landsat-2 Multispectral Scanner (MSS) sensor in 1975. Following image preprocessing, the RF algorithm was applied for variable selection and classification. A test of equivalence was used to compare the overall accuracy of the classified maps from the three sensors. Slope, normalized difference vegetation index (NDVI) and elevation were determined to be the most important predictor variables for all three images. High overall classification accuracies were achieved for all three images (97.90% for MSS, 95.43% for TM and 95.29% for ETM). The ETM- and TM-derived maps had equivalent overall accuracy and even significantly higher overall accuracy was obtained for the MSS-derived map. The post-classification comparison showed an increase in agriculture and a decrease in forest cover. The selected predictor variables were consistent with ecological reality and showed more details on the changes of the land-cover classes across biophysical variables of the study area through time.     

Extracting impervious surfaces from medium spatial resolution multispectral and hyperspectral imagery: a comparison

Remote sensing estimation of impervious surfaces is significant in monitoring urban development and determining the overall environmental health of a watershed, and has therefore recently attracted increasing interest. The main objective of this study was to develop a general approach to estimating and mapping impervious surfaces by using medium spatial resolution satellite imagery. We have applied spectral mixture analysis (SMA) to Earth Observing 1 (EO‐1) Advanced Land Imager (ALI) (multispectral) and Hyperion (hyperspectral) imagery in Marion County, Indiana, USA, to calculate the fraction images of vegetation, soil, high albedo and low albedo. The effectiveness of the two images was compared according to three criteria: (1) high‐quality fraction images for the urban landscape, (2) relatively low error, and (3) the distinction among typical land use and land cover (LULC) types in the study area. The fraction images were further used to estimate and map impervious surfaces. The accuracy of the estimated impervious surface was checked against Digital Orthophoto Quarter Quadrangle (DOQQ) images. The results indicate that both ALI and Hyperion sensors were effective in deriving the fraction images with SMA and in computing impervious surfaces. The SMA results for both ALI and Hyperion images using four endmembers were excellent, with a mean root mean square error (RMSE) less than 0.04 in both cases. The ALI‐derived impervious surface image yielded an RMSE of 15.3%, and the Hyperion‐derived impervious surface image yielded an RMSE of 17.5%. However, the Hyperion image was more powerful in discerning low‐albedo surface materials, which has been a major obstacle for impervious surface estimation with medium resolution multispectral images. A sensitivity analysis of the mapping of impervious surfaces using different scenarios of Hyperion band combinations suggests that the improvement in mapping accuracy in general and the better ability in discriminating low‐albedo surfaces came mainly from additional bands in the mid‐infrared region.     

Heterogeneous forest classification by creating mixed vegetation classes using EO-1 Hyperion

Hyperspectral satellite data is an efficient tool in vegetation mapping; however, previous studies indicate that classifying heterogeneous forests might be difficult. In this study, we propose a mapping method for a heterogeneous forest using the data of the Earth Observing-1 (EO-1) Hyperion supplemented by field survey. We introduced a band reduction method to raise classification accuracy of the Support Vector Machine classification algorithm and compared the results to the one reduced by principal component analysis (PCA), stepwise discriminant analysis (SDA), and the original data set. We also used a modified version of the Vegetation–Impervious–Soil model to create mixed vegetation classes consisting of the commonly mixing species in the area and classified them using Decision Tree classification method. We managed to achieve 84.28% approximately using our band reduction method which is 2.36% increase compared to PCA (81.92%), 1.43% compared to the SDA (82.85%), and 7.61% compared to the original data set (76.67%). Introducing the mixed vegetation classes raised the overall accuracy even higher (85.79%).     

Application of spectral mixture analysis to Amazonian land-use and land-cover classification

Abundant vegetation species and associated complex forest stand structures in moist tropical regions often create difficulties in accurately classifying land-use and land-cover (LULC) features. This paper examines the value of spectral mixture analysis (SMA) using Landsat Thematic Mapper (TM) data for improving LULC classification accuracy in a moist tropical area in Rondônia, Brazil. Different routines, such as constrained and unconstrained least-squares solutions, different numbers of endmembers, and minimum noise fraction transformation, were examined while implementing the SMA approach. A maximum likelihood classifier was also used to classify fraction images into seven LULC classes: mature forest, intermediate secondary succession, initial secondary succession, pasture, agricultural land, water, and bare land. The results of this study indicate that reducing correlation between image bands and using four endmembers improve classification accuracy. The overall classification accuracy was 86.6% for the seven LULC classes using the best SMA processing routine, which represents very good results for such a complex environment. The overall classification accuracy using a maximum likelihood approach was 81.4%. Another finding is that use of constrained or unconstrained solutions for unmixing the atmospherically corrected or raw Landsat TM images does not have significant influence on LULC classification performances when image endmembers are used in a SMA approach.     

Hyperspectral remote sensing for estimating aboveground biomass and for exploring species richness patterns of grassland habitats

Dry grassland sites are amongst the most species-rich habitats of central Europe and it is necessary to design effective management schemes for monitoring of their biomass production. This study explored the potential of hyperspectral remote sensing for mapping aboveground biomass in grassland habitats along a dry-mesic gradient, independent of a specific type or phenological period. Statistical models were developed between biomass samples and spectral reflectance collected with a field spectroradiometer, and it was further investigated to what degree the calibrated biomass models could be scaled to Hyperion data. Furthermore, biomass prediction was used as a surrogate for productivity for grassland habitats and the relationship between biomass and plant species richness was explored. Grassland samples were collected at four time steps during the growing season to capture normally occurring variation due to canopy growth stage and management factors. The relationships were investigated between biomass and (1) existing broad- and narrowband vegetation indices, (2) narrowband normalized difference vegetation index (NDVI) type indices, and (3) multiple linear regression (MLR) with individual spectral bands. Best models were obtained from the MLR and narrowband NDVI-type indices. Spectral regions related to plant water content were identified as the best estimators of biomass. Models calibrated with narrowband NDVI indices were best for up-scaling the field-developed models to the Hyperion scene. Furthermore, promising results were obtained from linking biomass estimations from the Hyperion scene with plant species richness of grassland habitats. Overall, it is concluded that ratio-based NDVI-type indices are less prone to scaling errors and thus offer higher potential for mapping grassland biomass using hyperspectral data from space-borne sensors.     

Evaluation of hyperspectral data for pasture estimate in the Brazilian Amazon using field and imaging spectrometers

We used two hyperspectral sensors at two different scales to test their potential to estimate biophysical properties of grazed pastures in Rondônia in the Brazilian Amazon. Using a field spectrometer, ten remotely sensed measurements (i.e., two vegetation indices, four fractions of spectral mixture analysis, and four spectral absorption features) were generated for two grass species, Brachiaria brizantha and Brachiaria decumbens. These measures were compared to above ground biomass, live and senesced biomass, and grass canopy water content. The sample size was 69 samples for field grass biophysical data and grass canopy reflectance. Water absorption measures between 1100 and 1250 nm had the highest correlations with above ground biomass, live biomass and canopy water content, while ligno-cellulose absorption measures between 2045 and 2218 nm were the best for estimating senesced biomass. These results suggest possible improvements on estimating grass measures using spectral absorption features derived from hyperspectral sensors. However, relationships were highly influenced by grass species architecture. B. decumbens, a more homogeneous, low growing species, had higher correlations between remotely sensed measures and biomass than B. brizantha, a more heterogeneous, vertically oriented species. The potential of using the Earth Observing-1 Hyperion data for pasture characterization was assessed and validated using field spectrometer and CCD camera data. Hyperion-derived NPV fraction provided better estimates of grass surface fraction compared to fractions generated from convolved ETM+/Landsat 7 data and minimized the problem of spectral ambiguity between NPV and Soil. The results suggest possible improvement of the quality of land-cover maps compared to maps made using multispectral sensors for the Amazon region.     

Classification of remotely sensed imagery using stochastic gradient boosting as a refinement of classification tree analysis

Classification tree analysis (CTA) provides an effective suite of algorithms for classifying remotely sensed data, but it has the limitations of (1) not searching for optimal tree structures and (2) being adversely affected by outliers, inaccurate training data, and unbalanced data sets. Stochastic gradient boosting (SGB) is a refinement of standard CTA that attempts to minimize these limitations by (1) using classification errors to iteratively refine the trees using a random sample of the training data and (2) combining the multiple trees iteratively developed to classify the data. We compared traditional CTA results to SGB for three remote sensing based data sets, an IKONOS image from the Sierra Nevada Mountains of California, a Probe-1 hyperspectral image from the Virginia City mining district of Montana, and a series of Landsat ETM+ images from the Greater Yellowstone Ecosystem (GYE). SGB improved the overall accuracy of the IKONOS classification from 84% to 95% and the Probe-1 classification from 83% to 93%. The worst performing classes using CTA exhibited the largest increases in class accuracy using SGB. A slight decrease in overall classification accuracy resulted from the SGB analysis of the Landsat data.     

Analysis and optimization of NDVI definitions and areal fraction models in remote sensing of vegetation

Variations in the definition of the Normalized Difference Vegetation Index (NDVI) and inconsistencies in vegetation areal fraction models prejudice the understanding of long‐term variability and change in land cover. We analysed the consequences of using NDVI definitions based on the digital number (DN), spectral radiance and spectral reflectance for six active and high spatial resolution multi‐ and hyperspectral satellite sensors (ALI, ASTER, ETM+, HRVIR, Hyperion and IKONOS) and optimized the NDVI definitions, and then examined the performance of three vegetation areal fraction models: the linear reflectance, linear NDVI and quadratic NDVI models. The examination was performed for three plots chosen from two biomass zones: a short and small leaf area index (LAI) creosote shrub zone, and a tall and large‐LAI piñon‐juniper zone. The results show that: (1) the difference in NDVI values among the NDVI definitions is sensor dependent and always significant; spectral reflectance should be used in NDVI calculations, and using radiance or DN values in calculating the NDVI should be avoided; (2) in deriving vegetation areal coverage, the linear reflectance model outperforms the other two models in the shrub biomass zone; and (3) the linear NDVI model outperforms the other two models in the piñon‐juniper biomass zone. These observations are consistent with the fact that the non‐linear effect is less important in shrubland than in piñon‐juniper woodland and that the linear NDVI model is more capable of capturing non‐linearity in the spectral analysis.     

Comparing techniques for vegetation classification using multi- and hyperspectral images and ancillary environmental data

This paper evaluates the predictive power of innovative and more conventional statistical classification techniques. We use Landsat 7 Enhanced Thematic Mapper Plus (ETM+), Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and airborne imaging spectrometer (HyMap) images to classify Mediterranean vegetation types, with and without inclusion of ancillary data (geology, soil classes and digital elevation model derivatives). When the number of classes is low, both conventional and innovative techniques perform well. For larger numbers of classes the innovative techniques of random forests and support vector machines outperform the other techniques. Compared to conventional techniques, classification trees, random forests and support vector machines proved to be better suited for the incorporation of continuous and categorical ancillary data: overall accuracies and accuracies for individual classes improve significantly when many, difficult to separate, classes are taken into account. Therefore, these techniques are definitely worth including in common image analysis software packages.     

Calculating environmental moisture for per-field discrimination of rice crops

The accuracies of rice classifications determined from density slices of broadband moisture indices were compared to results from a standard supervised technique using six reflective Enhanced Thematic Mapper plus (ETM+) bands. Index-based methods resulted in higher accuracies early in the growing season when background moisture differences were at a maximum. Analysis of depth of ETM+ band 5 resulted in the highest accuracy over the growing season (97.74%). This was more accurate than the highest supervised classification accuracy (95.81%), demonstrating the usefulness of spectral feature selection of moisture for classifying rice.     

EO-1 Hyperion数据的预处理、特征提取和岩性填图研究

EO-1 Hyperion传感器是第一个可以获取可见光与近红外以及短波红外波长范围光谱信息的星载高光谱传感器。本文以美国最早的金矿采矿区之一,加利福尼亚州东南巧克力山的Rainbow金矿区作为研究案例,探讨了Hyperion数据的预处理方法,专题信息提取与填图,评估了Hyperion高光谱数据在识别与金矿有关的岩性类型的应用价值。结果表明,本文所提出的Hyperion数据预处理方法是有效的,MNF方法能有效用于Hyperion数据维数的降低和数据冗余的去除以及分类特征的提取。最大似然分类器能够有效地从Hyperion高光谱数据中提取与金矿相关的重要岩体信息,所得到的岩性单元与地质图上对应的岩性分布具有很好的一致性。岩体分类的总精度为86%。该研究表明,Hyperion高光谱数据能够很好识别有细微光谱差别的岩性,因而在地质学研究与找矿领域有着良好的应用前景。     

Mapping health levels of Robinia pseudoacacia forests in the Yellow River delta,China, using IKONOS and Landsat 8 OLI imagery

The largest artificial Robinia pseudoacacia forests in the Yellow River delta of China have been infected by dieback diseases. Over the past several decades, this has caused a large amount of mortality of Robinia pseudoacacia forests in this area. Timely and accurate information on the health levels of the forests is crucial to improving local ecological and economic conditions. Remote sensing has been demonstrated to be a useful tool to map forest diseases over a large area. In this study, IKONOS and Landsat 8 Operational Land Imager (OLI) sensor data were collected for comparing their capability of accurately mapping health levels of the artificial forests. There were three health levels (i.e. healthy, medium dieback, and severe dieback) based on explicit tree crown symptoms. After the IKONOS and OLI images were preprocessed, both spatial and spectral features were extracted from the IKONOS and OLI imagery, and a maximum likelihood classification method was used to identify and map health levels of Robinia pseudoacacia forests. The experimental results indicate that the IKONOS sensor has greater potential for identifying and mapping forest health levels. Furthermore, texture features, especially texture variance, derived from the IKONOS panchromatic band, contributed greatly to the accuracy of classification results, achieving an overall accuracy (OA) of 96% for the IKONOS sensor and an OA of 88% for the OLI 2, which used both OLI spectral and IKONOS spatial features, compared with an OA of 74% for the OLI sensor alone. Our results indicate that the texture features extracted from high resolution imagery can improve the classification accuracy of health levels of planted forests with a regular spatial pattern. Our experimental results also demonstrate that classification of an image with a spatial resolution similar to, or finer than, tree crown diameter outperforms that of relatively coarse resolution imagery for differentiating living tree crowns and understorey dense green grass.     

Linear mixture model classification of burned forests in the Eastern Amazon

A methodology is described for detecting and classifying burned forests in Amazonia. Linear mixture models using three image endmembers (vegetation, soil, shade) were used to separate forest from non-forest. Forested areas were unmixed using vegetation, non-photosynthetic vegetation (NPV) and shade endmembers and reclassified as unburned, recently burned and older burned forests. The NPV fraction provided the greatest separability of the forest classes and has potential for subclassification of burned areas into damage classes. For 184 km2 of burned forest, a conservative estimate of 9% (22 metric tons ha-1) of living biomass was lost due to forest fires between 1991-1993.     

Spatial and temporal patterns of gap dominance by low-canopy lianas detected using EO-1 Hyperion and Landsat Thematic Mapper

Woody lianas are critical to tropical forest dynamics because of their strong influence on forest regeneration, disturbance ecology, and biodiversity. Recent studies synthesizing plot data from the tropics indicate that lianas are increasing in both abundance and importance in tropical forests. Moreover, lianas exhibit competitive advantages over trees in elevated CO₂ environments and under strong seasonal droughts, suggesting that lianas may be poised to increase not only in abundance but also in spatial distribution in response to changing climate. We used a combination of high-resolution color-infrared videography and hyperspectral imagery from EO-1 Hyperion to map low-lying lianas in Noel Kempff Mercado National Park (NKMNP) in the Bolivian Amazon. Evergreen liana forests comprise as much as 14% of the NKMNP landscape, and low-stature liana patches occupy 1.5% of these forests. We used change vector analysis (CVA) of dry season Landsat TM and ETM+ imagery from 1986 and 2000 to determine changes in liana-dominated patches over time and to assess whether those patches were regenerating to canopy forest. The spatial distribution of liana patches showed that patches were spatially aggregated and were preferentially located in proximity to waterways. The CVA results showed that most of the dense liana patches increased in brightness and greenness and decreased in wetness over the 14 years of the change analysis, while non-liana forest patches changed less and in more random directions. Persistent liana patches increased in area by an average of 59% over the time period. In comparison, large burned areas appeared to recover completely to canopy forest in the same time period. This suggests that the dense liana patches of NKMNP represent an alternative successional pathway characterized not by tree regeneration but rather by a stalled state of low-canopy liana dominance. This research supports hypotheses that liana forests can be a persistent rather than transitional component of tropical forests, and may remain so due to competitive advantages that lianas enjoy under changing climatic conditions.     

Use of a dark object concept and support vector machines to automate forest cover change analysis

An automated method was developed for mapping forest cover change using satellite remote sensing data sets. This multi-temporal classification method consists of a training data automation (TDA) procedure and uses the advanced support vector machines (SVM) algorithm. The TDA procedure automatically generates training data using input satellite images and existing land cover products. The derived high quality training data allow the SVM to produce reliable forest cover change products. This approach was tested in 19 study areas selected from major forest biomes across the globe. In each area a forest cover change map was produced using a pair of Landsat images acquired around 1990 and 2000. High resolution IKONOS images and independently developed reference data sets were available for evaluating the derived change products in 7 of those areas. The overall accuracy values were over 90% for 5 areas, and were 89.4% and 89.6% for the remaining two areas. The user's and producer's accuracies of the forest loss class were over 80% for all 7 study areas, demonstrating that this method is especially effective for mapping major disturbances with low commission errors. IKONOS images were also available in the remaining 12 study areas but they were either located in non-forest areas or in forest areas that did not experience forest cover change between 1990 and 2000. For those areas the IKONOS images were used to assist visual interpretation of the Landsat images in assessing the derived change products. This visual assessment revealed that for most of those areas the derived change products likely were as reliable as those in the 7 areas where accuracy assessment was conducted. The results also suggest that images acquired during leaf-off seasons should not be used in forest cover change analysis in areas where deciduous forests exist. Being highly automatic and with demonstrated capability to produce reliable change products, the TDA-SVM method should be especially useful for quantifying forest cover change over large areas.     

Studies of land-cover, land-use, and biophysical properties of vegetation in the Large Scale Biosphere Atmosphere experiment in Amazônia

We summarize early research on land-cover, land-use, and biophysical properties of vegetation from the Large Scale Biosphere Atmosphere (LBA) experiment in Amazônia. LBA is an international research program developed to evaluate regional function and to determine how land-use and climate modify biological, chemical and physical processes there. Remote sensing has played a fundamental role in LBA in research planning, land-cover mapping and in long-term monitoring of changes in land-cover and land-use at multiple scales. This special issue includes 12 papers that cover a range in spatial scales from regional mapping to local scales that cover only a portion of a Landsat scene. Several themes dominate, including land-cover mapping with an emphasis on wetlands and second-growth forest, evaluation of pasture sustainability and forest degradation and the impact of land-cover change on stream chemistry. New techniques introduced include automated Monte Carlo unmixing (AutoMCU) and several new approaches for mapping land-cover. A diversity of sensors are utilized, including ETM+, IKONOS, SPOT-4, Airborne P-band synthetic aperture radar (SAR), and L-band SAR. Census data are fused with an existing land-cover map to generate spatially explicit estimates of land-use from historical data. Several papers include important, new field measures of species composition, forest structure and biomass in mature forest and secondary succession.     

Discriminating native and plantation forests in a Landsat time-series for land use policy design

The Warren River Catchment of south-western Australia is an area of high biodiversity threatened by the loss of native vegetation and dryland salinity. Over the last 20 years, it has been the target of a series of policies that encourage conversion of agricultural land to plantation forest. Remote sensing has a key role in measuring trends in the area of plantation forest observed across the landscape and hence the effectiveness of policy initiatives. Despite its importance to land use policy, accurate data on historical land use and land cover (LULC) dynamics of two spectrally similar but ecologically distinct forest types – such as native forest and plantation forest – are not readily available for south-western Australia, largely due to prohibitive data delivery costs. However, we argue that regular low-cost monitoring of long-term change in the spatial distribution of plantation forest through remote sensing is a critical input into environmental policy for the catchment. To this end, a 35-year time-series of Landsat imagery was acquired, and three different classifiers were tested (Support Vector Machines – SVM; Random Forests – RF; and Classification and Regression Trees – CART) on spectral and textural indices applied to four spectral bands. The six major LULC classes considered were agriculture, water, native forest, sand dunes, plantation forest and harvested native forest. In classifying the imagery the SVM and RF outperformed the CART across all classes. However, the SVM classifier gave a slightly higher F-score for most individual classes than the RF. Eucalypt dominated plantation forest reaching full canopy cover was subject to the highest rates of misclassification inasmuch as it shares spectral properties with the Eucalypt dominant native forest. When applied to Landsat time-series imagery, SVM classifier combined with four bands held in common between the four Landsat sensors, and derived textures metrics are valuable in classifying plantation and native forest, particularly where these have a similar species composition. The differences in prediction accuracy when including additional Landsat bands were not statistically significant, as demonstrated by the McNemar test. Thus, we achieved a trade-off in reducing processing time without significantly impacting on classification accuracy (≥86%). The relatively high accuracy of the proposed method enables the effects of past policy initiatives to be observed, and hence the efficient design of environmental and conservation policy in the future.     

Capabilities of remote sensors to classify coral, algae, and sand as pure and mixed spectra

We investigate the abilities of seven remote sensors to classify coral, algae, and carbonate sand based on 10,632 reflectance spectra measured in situ on reefs around the world. Discriminant and classification analyses demonstrate that full-resolution (1 nm) spectra provide very good spectral separation of the bottom-types. We assess the spectral capabilities of the sensors by applying to the in situ spectra the spectral responses of two airborne hyperspectral sensors (AAHIS and AVIRIS), three satellite broadband multispectral sensors (Ikonos, Landsat-ETM+ and SPOT-HRV), and two hypothetical satellite narrowband multispectral sensors (Proto and CRESPO). Classification analyses of the simulated sensor-specific spectra produce overall classification accuracy rates of 98%, 98%, 93%, 91%, 64%, 58%, and 50% for AAHIS, AVIRIS, Proto, CRESPO, Ikonos, Landsat-ETM+, and SPOT-HRV, respectively. Analyses of linearly mixed sensor-specific spectra reveal that the hyperspectral and narrowband multispectral sensors have the ability to discriminate between coral and algae across many levels of mixing, while the broadband multispectral sensors do not. Applying the results of the general mixing analyses to a specific spatial organization of coral, algae, and sand indicates that the hyperspectral sensors accurately estimate areal cover of the bottom-types regardless of pixel resolution. The narrowband multispectral sensors overestimate coral cover by 11-15%, while the broadband sensors underestimate algae cover by 7-29% and overestimate coral cover by 24-103%. We conclude that currently available satellite sensors are inadequate for assessment of global coral reef status, but that it is both necessary and possible to design a sensor system suited to the task.