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.     

THE DEVELOPMENT OF THE NILE HYDROMETEOROLOGICAL FORECAST SYSTEM1

ABSTRACT: The National Oceanic and Atmospheric Administration is developing a river forecast system for the Nile River in Egypt. The river forecast system operates on scientific work stations using hydrometeorological models and software to predict inflows into the high Aswan Dam and forecast flow hydrographs at selected gaging locations above the dam The Nile Forecasting System (NFS) utilizes satellite imagery from the METEOSAT satellite as the input to the forecast system. Satellite imagery is used to estimate precipitation over the Blue Nile Basin using five different techniques. Observed precipitation data and climatic statistics are used to improve precipitation estimation. Precipitation data for grid locations are input to a distributed water balance model, a hill slope routing model, and a channel routing model. A customized Geographic Information System (GIS) was developed to show political boundaries, rivers, terrain elevation, and gaging network. The GIS was used to develop hydrologic parameters for the basin and is used for multiple display features.     

An erosion-based land classification system for military installations

The universal soil loss equation (USLE) has been integrated with a geographic information system known as the geographical resources analysis support system (GRASS) to create a land classification system for use by military trainers and land managers to minimize the environmental impacts of military training activities. The USLE provides an estimate of current average annual sheet and rill erosion based upon factors representing climate, soil erodibility, topography, cover, and conservation support practices. The erosion estimate is compared to erosion tolerance values to produce an expression of the current erosion status. An index of inherent site erodibility is also achieved through manipulation of the USLE. Based on published soil surveys, satellite imagery, and ground-truth vegetation transects, data layers are created within GRASS for each of the component factors of the USLE. Appropriate mathematical operations are performed with the data layers, and color-coded maps are produced that represent the erosion status and erodibility index for each 50-m × 50-m area of soil surface. These maps aid military trainers and land managers in scheduling appropriate kinds and intensities of military training activities.     

Remote Sensing and GIS Technologies as Tools to Support Sustainable Management of Areas Devastated by Landslides

IKONOS panchromatic images from a single year were used to characterise the effects of an extreme rainstorm event on six mountain catchment areas in Venezuela. Image registrations were accomplished using topographic base maps at 1 : 5000 scale giving a mapping accuracy between 37 and 40 m. It was used an 8-bit channel for correction, rectification, filteration and tone enhancements.Landsliding was discerned in the affected watersheds using morphometric criteria, including the shape of the slope failure and its position, exposure bedrock in the scar and deposition of debris down-slope. This study is restricted to the impact analysis of the distribution of landslide erosion scars and the depositional processes on the valley floor. Remote sensing data were combined into a geographic information system (GIS) with planimetric data, contour lines, hydrology and vegetation types to evaluate the distribution of the scars and their effects on the highly populated areas located on the alluvial fans. Hillslope mass wasting induced mass movements, logging and increased the mud and silt in floodwaters affecting settlement down-slope.     

Prediction of diffuse solar radiation using satellite data

In the project of the solar systems, the values of the solar radiation of a region must be known. The global solar radiation measurements are performed by the Turkish Meteorological Service in Turkey, while the diffuse solar radiation measurements are unavailable. In this study, some new models for predicting the monthly average daily diffuse solar radiation on a horizontal surface for Erzincan, Turkey are developed using satellite data. The evaluation of the models is carried out with the statistical analysis methods of mean bias error (MBE), root mean square error (RMSE) and correlation coefficient (r). The results are proved that the correlation equations obtained in here can be used to forecast diffuse solar radiation reasonably well.     

Applying Satellite Imagery to Triage Assessment of Ecosystem Health

Considerable evidence documents that certain changes in vegetation and soils result in irreversibly degraded rangeland ecosystems. We used Advanced Very High Resolution Radiometer (AVHRR) imagery to develop calibration patterns of change in the Normalized Difference Vegetation Index (NDVI) over the growing season for selected sites for which we had ground data and historical data characterizing these sites as irreversibly degraded. We used the NDVI curves for these training sites to classify and map the irreversibly degraded rangelands in southern New Mexico. We composited images into four year blocks: 1988–1991, 1989–1992, and 1990–1993. The overlap in pixels classified as irreversibly degraded ranged from 42.6% to 84.3% in year block comparisons. Quantitative data on vegetation composition and cover were collected at 13 sites within a small portion of the study area. Wide coverage reconnaissance of boundaries between vegetation types was also conducted for comparisons with year block maps. The year block 1988–1991 provided the most accurate delineation of degraded areas. The rangelands of southern New Mexico experienced above average precipitation from 1990–1993. The above average precipitation resulted in spatially variable productivity of ephemeral weedy plants on the training sites and degraded rangelands which resulted in much smaller areas classified as irreversibly degraded. We selected imagery for a single year, 1989, which was characterized by the absence of spring annual plant production in order to eliminate the confounding effect of reflectance from annual weeds. That image analysis classified more than 20% of the rangelands as irreversibly degraded because areas with shrub-grass mosaic were included in the degraded classification. The single year image included more than double the area classified as irreversibly degraded by the year blocks. AVHRR imagery can be used to make triage assessments of irreversibly degraded rangeland but such assessment requires understanding productivity patterns and variability across the landscapes of the region and careful selection of the years from which imagery is chosen.     

Mapping giant salvinia with satellite imagery and image analysis

QuickBird multispectral satellite imagery was evaluated for distinguishing giant salvinia (Salvinia molesta Mitchell) in a large reservoir in east Texas. The imagery had four bands (blue, green, red, and near-infrared) and contained 11-bit data. Color-infrared (green, red, and near-infrared bands), normal color (blue, green and red bands), and four-band composite (blue, green, red, and near-infrared bands) images were studied. Unsupervised image analysis was used to classify the imagery. Accuracy assessments performed on the classification maps of the three composite images had producer’s and user’s accuracies for giant salvinia ranging from 87.8 to 93.5%. Color-infrared, normal color, and four-band satellite imagery were excellent for distinguishing giant salvinia in a complex field habitat.     

环境一号卫星CCD数据在太湖蓝藻水华遥感监测中的应用

利用环境一号卫星(HJ-1)CCD数据,对太湖水华进行遥感监测,并比对同时相的EOS/MODIS卫星遥感数据。结果表明,HJ-1星CCD数据具有优于EOS/MODIS数据的蓝藻水华识别能力,并有良好抗云层干扰能力,适合用于太湖蓝藻水华应急监测。     

环境一号卫星环境应用系统工程及其关键技术研究进展

简介长达11年的国家重点项目“环境一号卫星环境应用系统工程及其关键技术研究”的核心内容及其重要研究成果,对环境一号卫星系统需求分析、设计理论、研制发射和运行应用全过程进行了完整的研究与实践,为环境一号卫星系统的建设、应用和业务运行奠定了重要基础。     

成都大气污染物在焚烧秸秆时的溯源初步探究

在空气污染指数(API)的数据基础上,结合卫星遥感、气团后向轨迹及气象数据,通过初步分析秸秆焚烧对本地及区域范围输送对城市空气质量影响,研究不同空气污染类型(本地型、本地区域相结合型、输送型)特征,探索重污染条件的大气污染物溯源方法。