Showing papers in "Photogrammetric Engineering and Remote Sensing in 1999"

Probabilistic prediction models for landslide hazard mapping

Abstract: A joint conditional probability model is proposed to represent a measure of a future landslide hazard, and five estimation procedures for the model are presented. The distribution of past landslides was divided into two groups with respect to a fixed time. A training set consisting of the earlier landslides and the geographical information system-based multi-layer spatial data in the study area was used to construct the prediction maps. The predictions were then cross-validated by comparing them with the remaining later landslides. When the database falls short of providing sufficient support for the prediction, the model allows the introduction of the expert's knowledge to modify the observed frequencies of the landslides with respect to the spatial data. The additional information should improve the prediction results. A case study from the Rio Chincina region in Colombia was used to illustrate the methodologies.

The IGBP-DIS global 1-km land-cover data set DIS-Cover : A project overview

Abstract: Since 1992 the International Geosphere Biosphere Programme's (IGBP) Data and Information System (DIS) has been working towards the completion of a validated global landcover data set, DISCover. This 1-km resolution data set consists of 17 cover classes identified on the basis of the science requirements of the IGBP's core projects. DISCover has been created from over 4.4 Terabytes of data from the Advanced Very High Resolution Radiometer collected from 23 receiving stations. These data were processed and assembled into a coherent set of monthly Normalized Difference Vegetation Index composites (April 1992 to April 1993) and classified using unsupervised techniques with post-classification refinement. The first global hind-cover classification was completed in July 1997. The IGBP-DIS Land Cover Working Group, in turn, convened a Validation Working Group to provide and implement a validation method to provide statistical statements concerning the accuracy of the global land-cover product and to allow the estimation of the error variance in areal totals of classes globally and within regions. The validation workshop was completed in September 1998 and the analysis by March 1999. This paper describes the history of the DISCover version 1.0 implementation.

An analysis of the IGBP global land-cover characterization process

Abstract: The International Geosphere Biosphere Programme (rcnr) has called for the development of improved global Land-cover data for use in increasingly sophisticated global environmental models. To meet this need, the staff of the U.S. Geological Survey and the University of Nebruska-Lincoln developed and applied a global land-cover characterization methodology using 1992-1993 1-km resolution Advanced Very High Resolution Radiometer fnvunn) and other spatial data. The methodology, based on unsupervised classification with extensive postclassification refinement, yielded a multi-)ayer database consisting of eight hand-cover data sets, descriptive attributes, and source data. An independent IGBP accurucy assessment reports a global accurccy of zs.s percent, and continental results vary from 63 percent to 83 percent. Althou gh data qu al ity, metho dolo gy, interprcter p erf ormanc e, and logistics affected the results, significant problems were associated with the relationship between AvHnR data and finescale, spectrally similar land-cover patterns in complex natural or disturbed landscapes. lntroductioninitial impetus to develop a 1-km global land-cover characteristics database grew from calls for improved land-cover data from numerous scientific organizations (i.e., National Academy of Sciences, 1990; National Aeronautics and Space Administration, 1994). There was a significant need to develop validated, contemporary, and spatially and thematically detailed global land-cover data for scientific inquiries associated with global change research, assessments of sustainable development, and operational functions such as weather forecasting. The available global land-cover data were determined to be inadequate for the coming generation of climate models (Sellers, 1903), carbon cycle assessments (S. Brown, ef o1., 1993), ecological models (Schimel et a1.,1'991), and conservation studies (Davis ef 01., 1990). The catalyst for the development of an improved global land-cover product was the International Geosphere Biosphere Programme Data and Information System (Icnr-nrs). Through user requirements forums that canv-assed the needs oftcnp core science projects (Rasool, 1992), the need and rationale for a new 1-km global land-cover characteristics database were defined (IGBR 1992). The IGBPIIS Land Cover Working Group subsequently devoted more than five years of technical planning and oversigh!to the defini tion, speCification, and completion of a new global land-cover database. or DISCover, that was based on 1992-1993 Advanced Very High Resolution Radiometer (.qvHRR) data acquired by the National Oceanic and Atmospheric Administration-11 polar orbiting satellite.

Thematic validation of high-resolution Global Land-Cover Data sets

Abstract: This paper describes a procedure to validate the thematic accuracy of the International Geosphere-Biosphere Programme, Data and Information System (IGBP-DIS) DISCover (Version 1.0) 1-Kilometer Global Land-Cover Data Set. Issues of data set sampling design, image geometry and registration and core sample interpretation procedures are addressed. Landsat Thematic Mapper and SPOT satellite image data were used to verify 379 primary core samples selected from DISCover 1.0 using a stratified random sampling procedure. The goal was to verify a minimum of 25 samples per DISCover class, this was accomplished for 13 of the 15 verified classes. Three regional Expert Image Interpreters independently verified each sample, and a majority decision rule was used to determine sample accuracy. For the 15 DISCover classes validated. the average class accuracy was 59.4 percent with accuracies for the 15 verified DISCover classes ranging between 40.0 percent and 100 percent. The overall area-weighted accuracy of the data set was determined to be 66.9 percent. When only samples which had a majority interpretation for errors as well as for correct classification were considered, the average class accuracy of the data set was calculated to be 73.5 percent.

Automatic road extraction based on multi-scale, grouping, and context

Abstract: An approach for the automatic extraction of roads from digital aerial imagery is proposed. It makes use of several versions of the same aerial image with different resolutions. Roads are modeled as a network of intersections and links between these intersections, and are found by a grouping process. The context of roads is hierarchically structured into a global and a local level. The automatic segmentation of the aerial image into different global contexts, i.e., rural, forest, and urban area, is used to focus the extraction to the most promising regions. For the actual extraction of the roads, edges are extracted in the original high resolution image (0.2 to 0.5 m) and lines are extracted in an image of reduced resolution. Using both resolution levels and explicit knowledge about roads, hypotheses for road segments are generated. They are grouped iteratively into larger segments. In addition to the grouping algorithms, knowledge about the local context, e.g., shadows cast by a tree onto a road segment, is used to bridge gaps. To construct the road network, finally intersections are extracted. Examples and results of an evaluation based on manually plotted reference data are given, indicating the potential of the approach.

Remote Sensing and GIs at Farm Property Level: Demography and Deforestation in the Brazilian Amazon

Abstract: Analysis of remotely sensed data at the level of individual farm properties provides additional insights to those derived from a landscape approach. Property-level analysis was carried out by overlaying a property boundary grid in a GIS. Data were derived from aerial photographs for 1970 and 1978 and Landsat Thematic Mapper images for 1985, 1988, and 1991. The study area contains approximately 3,800 properties, but this paper is based on a subset of 398 properties in the Brazilian Amazon. Analysis at the property level found patterns of land-cover classes that reflect differences in farming strategies of households. Data analysis at the household level was useful in explaining apparent mature forest to advanced secondaly succession degradation in three years, not readily apparent from landscape analysis of remotely sensed data. The change was due to property-specific selective logging and the spread of fire from pastures into the adjacent forest.

Application of multi-temporal Landsat 5 TM imagery for wetland identification

Abstract: Multi-temporal Landsat 5 Thematic Mapper (TM) imagery was evaluated for the identification and monitoring of potential jurisdictional wetlands located in the states of Maryland and Delaware. A wetland map prepared from single-date TM imagery was compared to a hybrid map developed using two dates of imagery. The basic approach was to identify landcover vegetation types using spring leaf-on imagery, and identify the location and extent of the seasonally saturated soil conditions and areas exhibiting wetland hydrology using spring leaf-off imagery. The accuracy of the wetland maps produced from both single- and multiple-date TM imagery were assessed using reference data derived from aerial photographic interpretations and field observation data. Subsequent to the merging of wetland forest and shrub categories, the overall accuracy of the wetland map produced from two dates of imagery was 88 percent compared to the 69 percent result from single-date imagery. A Kappa Test Z statistic of 5.8 indicated a significant increase in accuracy was achieved using multiple-date TM images. Wetland maps developed from multi-temporal Landsat TM imagery may potentially provide a valuable tool to supplement existing National Wetland Inventory maps for identifying the location and extent of wetlands in northern temperate regions of the United States.

The Continuum of Classification Fuzziness in Thematic Mapping

Abstract: Thematic mapping from remotely sensed data is generally achieved through the application of a supervised image classification. Although this is one of the most common applications of remote sensing, the maps derived are often of insufficient accuracy for operational use. Consequently, considemble research effort has been directed at increasing the accuracy of thematic mapping, particularly through the development of classification techniques that make fuller use of the information content of remotely sensed data. One major set of problems limiting the accuracy of thematic maps derived from remotely sensed data relates to conceptual issues associated with the use of classification techniques as the tool for mapping. It is shown that the conventional "hard" classifications may be less appropriate than fuzzy classifications and that a continuum of classification fuzziness can be defined. The potential for classification at any point along this continuum, from completely crisp to fully fuzzy, is discussed and may provide a framework for realizing more fully the potential of remote sensing as a source of thematic map data. The implications of the continuum on spatial data 1 standards and reporting are also briefly discussed.

Scale-Dependent Relationships between Population and Environment in Northeastern Thailand

Abstract: with social processes to examine the nature of patterns Social and biophysical data were collected, integrated, and Vatial and/or scales for landscapes having a analyzed to examine scale-dependent relationships between pronounced "social" imprint. Scale dependent studies conselected population and environmental variables for a study ducted in environments have typically lacked the site in northeast Thailand. Data sets were generated through human data at fine spatial ~cales (e%., the use of remote sensing to characterize land-use/land-cover Or and a sufficiently large Moreand plant biomass variation across the Nang Rong district; over~ researchers have been primarily biophysical ~cholars ~1s to derive slope angle, and soil moisture poten- unaccustomed to working with social science data. In additia]; social survey data at the village level to categorize dem- tion, the remote sensing and GIs research communities have ographic variables; and a population distribution model to only begun focusing attention on the interplay betransform demographic data collected at discrete village lo- tween social, biophysical, and geographic factors as drivers cations to spatially continuous surfaces stratified by agri- of land-uselland-cover change (LULCC). cultural land uses. Statistical analysis employed multiple Spatial scale is inherently involved in recognizing sparegression to populatjon density in relation to tial patterns on the landscape and in estimating the relationand biophysical variables, and canonical analysis to relate ships between landscape and environmental and population variables to environmental variables across a social processes deriving those patterns (Bian and Walsh, range of spatial scales extending from 30 to 1050 m. Find- 1993; Allen and Walsh, 1996). Specific biotic, environmental, ings indicate the importance of spatial scale in the study of social, and historical processes function at various ranges of population and the environment. Regression models reflect Vatial The ranges and among Processes. the scale dependence of the selected variables through plots Patterns the landscape be discernible at cerof slope coefficients and RZ values across nine scale steps. tain spatial scales and ranges of spatial scales (~eenteme~er The variation in relationships among environment and popu- and lga7)9 and landscape appear homogeneous at lation variables, evidenced through factor loadings associ- Some scales but heterogeneous at others (Nellis and Briggs, ated with canonical correlation, suggest that relationships 1989). Based the assumption that the spatial Patterns of are not generalizeable across the sampled spatial scales. landscapes are formed by environmental and social processes whose effects vary with spatial scale, it is reasonable to hy

Remotely Sensed Change Detection Based on Artificial Neural Networks

Abstract: A new method for remotely sensed change detection based on artificial neural networks is presented. The algorithm for an automated land-cover change-detection system was developed and implemented based on the current neural network techniques for multispectral image classification. The suitability of application of neural networks in change detection and its related network design considerations unique to change detection were first investigated. A neural-network-based change-detection system using the backpropagation training algorithm was then developed. The trained four-layered neural network was able to provide complete categorical information about the nature of changes and detect land-cover changes with an overall accuracy of 95.6 percent for a four-class (i.e., 16 change classes) classification scheme. Using the same training data, a maximum-likelihood supervised classification produced an accuracy of 86.5 percent. The experimental results using multitemporal Landsat Thematic Mapper imagery of Wilmington, North Carolina are provided. Findings of this study demonstrated the potential and advantages of using neural network in multitemporal change analysis.

Change analysis in the United Arab Emirates: An investigation of techniques

Abstract: Much of the landscape of the United Arab Emirates has been transformed over the past 15 years by massive afforestation, beautification, and agricultural programs. The "greening" of the United Arab Emirates has had environmental consequences, however, including degraded groundwater quality and possible damage to natural regional ecosystems. Personnel from the Ground- Water Research project, a joint effort between the National Drilling Company of the Abu Dhabi Emirate and the U.S. Geological Survey, were interested in studying landscape change in the Abu Dhabi Emirate using Landsat thematic mapper (TM) data. The EROs Data Center in Sioux Falls, South Dakota was asked to investigate land-cover change techniques that (1) provided locational, quantitative, and qualitative information on landcover change within the Abu Dhabi Emirate; and (2) could be easily implemented by project personnel who were relatively inexperienced in remote sensing. A number of products were created with 1987 and 1996 Landsat TM data using change-detection techniques, including univariate image differencing, an "enhanced" image differencing, vegetation index differencing, post-classification differencing, and changevector analysis. The different techniques provided products that varied in levels of adequacy according to the specific application and the ease of implementation and interpretation. Specific quantitative values of change were most accurately and easily provided by the enhanced image-differencing technique, while the change-vector analysis excelled at providing rich qualitative detail about the nature of a change.

Comparing Effects of Aggregation Methods on Statistical and Spatial Properties of Simulated Spatial Data

Abstract: Spatial data aggregation is widely practiced for "scaling-up" environmental analyses and modeling from local to regional or global scales. Despite acknowledgments of the general effects of aggregation, there is a lack of systematic comparison between aggregation methods. The study evaluated three methods - averaging, central-pixel resampling, and median using simulated images. Both the averaging and median methods can retain the mean and median values, respectively, but alter significantly the standard deviation. The central-pixel method alters both statistics. The statistical changes can be modified by the presence of spatial autocorrelation for a11 three methods. Spatially, the averaging method can reveal underlying spatial patterns at scales within the spatial autocorrelation ranges. The median method produces almost identical results because of the similarities between the averaged and median values of the simulated data. To a limited extent, the central-pixel method retains contrast and spatial patterns of the original images. At scales coarser than the autocorrelation range, the averaged and median images become homogeneous and do not difler significantly between these scales. The central-pixel method can induce severe spatially biased errors at coarse scales. Understanding these trends can help select appropriate aggregation methods and aggregation levels for particular -- applications.

Multi-Scale Fractal Analysis of Image Texture and Pattern

Abstract: Fractals embody important ideas of self-similarity, in which the spatial behavior or appearance of a system is largely independent of scale. Self-similarity is defined as a property of curves or surfaces where each part is indistinguishable from the whole, or where the form of the curve or surface is invariant with respect to scale. An ideal fractal (or monofractal) curve or surface has a constant dimension over all scales, although it may not be an integer value. This is in contrast to Euclidean or topological dimensions, where discrete one, two, and three dimensions describe curves, planes, and volumes. Theoretically, if the digital numbers of a remotely sensed image resemble an ideal fractal surface, then due to the self-similarity property, the fractal dimension of the image will not vary with scale and resolution. However, most geographical phenomena are not strictly self-similar at all scales, but they can often be modeled by a stochastic fractal in which the scaling and self-similarity properties of the fractal have inexact patterns that can be described by statistics. Stochastic fractal sets relax the monofractal self-similarity assumption and measure many scales and resolutions in order to represent the varying form of a phenomenon as a function of local variables across space. In image interpretation, pattern is defined as the overall spatial form of related features, and the repetition of certain forms is a characteristic pattern found in many cultural objects and some natural features. Texture is the visual impression of coarseness or smoothness caused by the variability or uniformity of image tone or color. A potential use of fractals concerns the analysis of image texture. In these situations it is commonly observed that the degree of roughness or inexactness in an image or surface is a function of scale and not of experimental technique. The fractal dimension of remote sensing data could yield quantitative insight on the spatial complexity and information content contained within these data. A software package known as the Image Characterization and Modeling System (ICAMS) was used to explore how fractal dimension is related to surface texture and pattern. The ICAMS software was verified using simulated images of ideal fractal surfaces with specified dimensions. The fractal dimension for areas of homogeneous land cover in the vicinity of Huntsville, Alabama was measured to investigate the relationship between texture and resolution for different land covers.

Evaluation of NDVI for monitoring live moisture in three vegetation types of the Western U.S.

Abstract: been estimated through empirically derived algorithms or The Normalized Dimence Vegetation Index (~~vr) is evalu- Was measured using intensive field methodology. ideally, inated for monitoring seasonal a Burgan et al., 1997). The Advanced Very High Resoluthan the conifer canopy was examined. Seasonal changes in tion sensor On the TIROS-N series vegetation moisture for all sites were statistically significant polar-orbiting weather satellites provides daily observations (p<0.05). Time-series profiles of the NDVI were functionally of the Earth's mrface at a nominal spatial resolution (at narelated to changes in vegetation moisture only for the grass dir) square lcrn. it was and forest understory vegetation. N~ significant relationships with natural resource observation objectives, several indices were observed for either the shrub or coniferous forest can- derived the AVHRR Vectral data provide meaningfu1 opy vegetation. This field experiment will improve our inter- measures of vegetation condition (Loveland and ~hlen, pretation of seasonal NDVI data with respect to fire potential. lgg3). The lIifference Index (NDvl)*

Objects with fuzzy spatial extent

Abstract: The determination of the spatial extent of geo-objects is generally approached through their boundaries or, more precisely, through the positions of their boundary points The analysis of the geometric uncertainty of the objects is therefore often based on accuracy models for the coordinates of these points In many survey disciplines objects are mapped, however, that are not crisp well defined In that case, the geometric uncertainty is not only a matter of coordinate accuracy, but also a problem of object definition and thematic vagueness The spatial uncertainty of such objects cannot be handled by a geometric approach alone, such as the epsilon band method This paper investigates the reasons for the fuzzy spatial extent of objects and proposes an approach to map the spatial extent of objects and their uncertainties when objects are extracted from field observation data The relationship of uncertainties between thematic aspects and geometric aspects is investigated A practical example of a coastal geomorphology study is discussed to illustrate the approach

Air photointerpretation and satellite imagery analysis techniques for mapping cattail coverage in a northern Everglades impoundment

Abstract: Color-infrared aerial photography taken in 1991 and 1995, and SPOT satellite imagery taken in 1991, were utilized to create cattail coverage maps for Water Conservation Area 2A (WCA2A), an impounded portion of the remnant Everglades. Cattail stands were delineated and classified using conventional air photointerpretation and digital image processing techniques, respectively. Four interacting confounding factors (i.e., water depth/color, impacts from fire, periphyton species composition, and growth morphology within a single species) are implicated as possible elements that complicated vegetation classification. Photointerpretation techniques showed an increasing trend in cattail encroachment from 421.6 hectares of monotypic cattail in 1991 to 1646.3 hectares in 1995. A 1991 SPOT classified image appears to have overestimated cattail coverage due to the interacting confounding mechanisms. Overall accuracies for 1995 air photointerpreted map and 1991 SPOT classified image were 95.2 and 83.4 percent, respectively.

Deforestation in North-Central Yucatan 1985-1995) : Mapping secondary succession of forest and agricultural land use in Sotuta using the cosine of the angle concept

Abstract: A new spectral pattern matching approach that utilizes the spectral angle (the cosine of the angle) concept was used for mapping deforestation and successional stages of forest regrowth in Sotuta in the state of Yucatan, Mexico. By calculating spectral angles between finely defined spectral clusters and known reference signatures, and assigning each spectral cluster to one of the reference classes based on the minimum spectral angle rule, we were able to map forest regrowth stages and agricultural land-use classes. Our research shows that, by adapting a spectral pattern matching approach demonstrated in this paper, spectral clusters can be assigned into information classes precisely and objectively, and the inconsistency involved in visual interpretations can be avoided. The conceptual difference between the spectral distance and spectral angle in feature space is also reviewed. In the study area, the rate of deforestation is high and agricultural land use is intensifying increasingly. The limited amount of land granted to ejidos and rapid population growth seem to be major causes of deforestation in the study area.

The global land-cover characteristics database : The users' perspective

Abstract: A unique global land-cover characteristics database developed by the U.S. Geological Survey has been available to users since mid-1997. Access to the data is through the Internet under the EROS (Earth Resources Observation Systems) Data Center's home page (http://edcwww.cr.usgs.gov/landdaac/glcc/ glcc.html). Since the release of the database, the data have been incorporated into various environmental research and modeling applications, including mapping global biodiversity, mesoscale climate simulations, carbon cycle modeling, and estimating habitat destruction. Since the early stages of the project, user feedback has provided a means to understand data utility in applications, garner suggestions for data improvements. and gain insights into the technical challenges faced by users. Synthesis of user feedback provided a means to generate a user profilc and derive a list of applications-critical criteria for land-cover data. User suggestions have lead to revisions in the database, including label changes, alternative classification schemes, and additional projections for the data.

Digital photogrammetric change analysis as applied to active coastal dunes in michigan

Abstract: A pilot study was conducted to investigate the applicability of digital photogmmmetric methods to the study and management of dynamic dune systems. Two sets of panchromatic stereographic aerial photographs taken over Ludington State Park, Michigan, one pair each from 1965 (1:20,000-scale) and 1987 (1:15,000-scale), were obtained from historical archives. Stereo models were constructed for the stereo-pairs, using post-processed differential GPS ground control points, and digital elevation models (DEMS) were extracted from each at a resolution of 3 metres. The analysis involved computing differences between the two DEMs at each location, and computing a volume of sediment (sand) flux over the 22-year time period. Maps of elevation change were then constmcted and interpreted to suggest patterns resulting from eolian processes. Processes of dune development, movement, and "blowout" were identifiable and measurable. The project illustrates how recent developments in photogrammet& have enhanced capabilities for monitoring geomorphologically sensitive landscapes I

The IGBP discover Confidence sites and the System for Terrestrial Ecosystem parameterization : Tools for validating global land-cover data

Abstract: The IGBP Validation Confidence Site database provides a set of 379 land-cover maps, each containing an IGBP Core Validation sample. Each map is 448 km 2 in area and is delineated and labeled by photointerpretation of Landsat or SPOT satellite imagery at a scale of 1:125,000. Within each map, land-cover types and polygons are assigned descriptive labels and parameter codes for vegetation attributes, including life form, cover, height, and phenology for canopy and ground layers. These attributes are a subset of parameters defined by the System for Terrestrial Ecosystem Parameterization (STEP), a site model and database that characterizes land surface and vegetation for use in global algorithm training, testing, and validation of land-cover data. Because the maps are linked to the core samples, they provide a large, consistent dataset that is stratified to represent equally all of the orld's major vegetation form classes and land-cover types. The confidence site database has three primary applications: (1) as a set of validation bonchnutrks for allernate regional or global landcover classifications emphasizing vegetation attributes, (2) as a secondary information source for studying core sample accuracy issues, and (3) as a source of training and test sites for regional and global supervised classification of coarse-resolution satellite imagery.

Fractal Characterization of Hyperspectral Imagery

Abstract: Two Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) hyperspectral images selected from the Los Angeles area, one representing urban and the other, rural, were used to examine their spatial complexity across their entire spectrum of the remote sensing data. Using the ICAMS (Image Characterization And Modeling System) software, we computed the fractal dimension values via the isarithm and triangular prism methods for all 224 bands in the two AVIRIS scenes. The resultant fractal dimensions reflect changes in image complexity across the spectral range of the hyperspectral images. Both the isarithm and triangular prism methods detect unusually high D values on the spectral bands that fall within the atmospheric absorption and scattering zones where signature to noise ratios are low. Fractal dimensions for the urban area resulted in higher values than for the rural landscape, and the differences between the resulting D values are more distinct in the visible bands. The triangular prism method is sensitive to a few random speckles in the images, leading to a lower dimensionality. On the contrary, the isarithm method will ignore the speckles and focus on the major variation dominating the surface, thus resulting in a higher dimension. It is seen where the fractal curves plotted for the entire bandwidth range of the hyperspectral images could be used to distinguish landscape types as well as for screening noisy bands.

Mapping the Everglades.

Abstract: The Center for Remote Sensing and Mapping Science at The University of Georgia and the South Florida Natural Resources Center, Everglades National Park, have developed a detailed vegetation database in geographic information system (GIS) format and 1:15,000-scale vegetation maps keyed to 80 U.S. Geological Survey (USGS) 7.5-minute topographic quadrangles covering Everglades National Park, Big Cypress National Preserve, Biscayne National Park, and the Florida Panther National Wildlife Refuge in south Florida, an area of over 10,000 km 2 , National Aerial Photography Program (NAPP) color-infrared (CIR) aerial photographs recorded in 1994/95 were the primary source material, supplemented by extensive GPS-assisted data collections from helicopter reconnaissance missions and field work. The ground control necessary to rectify the vegetation polygons and linear features extracted from the CIR photographs was obtained from geocoded SPOT panchromatic images of 10-m resolution and from USGS 1:24,000-scale topographic line maps. A detailed three-tiered, hierarchical vegetation classification system, the Everglades Vegetation Classification System, was developed specifically for the project. In addition to the vegetation database, a digital database and map products were constructed for off-road vehicle (ORV) trails in Big Cypress National Preserve. The length of trails in the 2,950-km 2 Preserve totaled over 47,900 km. The databases and maps, constructed through a combination of remote sensing, GIS, GPS, and field studies, provide Park and Preserve managers with detailed baseline information on the status of vegetation and ORV trails in 1995. It is anticipated that, with increased concern over environmental preservation, water demand and expansion of urban development, agricultural land utilization, and ORV use, the databases will prove valuable for a range of management and modeling tasks.

Implications of land-cover misclassification for parameter estimates in global land-surface models: An example from the simple biosphere model (SiB2)

Abstract: One of the primary applications of the global 1-km land-cover DISCover product is to derive biophysical and ecological parameters for a range of land-surface models, including biosphere-atmosphere, biogeochemical, and ecological models. The validation effort reported in this special issue enables a realistic assessment of the implications of misclassification errors for parameter estimates within the models. In most land-surface models, cover types are aggregated to coarser groupings than the 17 IGBP classes for estimating parameters, with aggregation schemes varying with individual models and individual parameters within each model. Misclassification errors are consequential only when they occur between cover types that are not aggregated by the model. We use examples of two biophysical parameters-leaf area index and surface roughness-as estimated for use in the Simple Biosphere Model (SiB2) and other modeling applications to quantify the effects of misclassification on parameter estimates. SiB2 relies on satellite data as well as land-cover information for estimating the biophysical parameters. Consequences of misclassification are likely to be greater for those models that do not use satellite data. Mean class accuracy based on those sites for which a majority of interpreters agreed (percentage of validation pixels classified correctly out of total number of validation pixels, averaged over all classes), adjusted by area of each cover type in the IGBP DISCover product, is 78.6 when all misclassification errors are included. By excluding misclassification errors when they are inconsequential for leaf area index and surface roughness length estimates, mean class accuracies are 90.2 and 87.8. respectively. The results illustrate that misclassification errors are most meaningfully viewed in the context of the application of the land-cover information.

Knowledge-based interpretation of remote sensing images using semantic nets

Abstract: The increasing amount of remotely sensed imagery requires efficient analysis techniques. The leading idea of the presented work is to automate the interpretation of aerial images by the use of common a priori knowledge about landscape scenes. In addition, the system uses specific map knowledge of a GIS. The a priori knowledge about landscape scenes, the aerial images, and the image forming sensors is represented explicitly by a semantic net. The definition of a network language allows the exploition of the knowledge base by a set of application-independent rules which provide data and model-driven control strategies. Competing interpretations are stored in a search tree and judged considering their uncertainty and imprecision. An A * -algorithm selects the most promising interpretation for further analysis. Results are shown for the extraction of roads and complex objects, such as purification plants, from multisensor imagery.

Mapping Exotic Vegetation in the Everglades from Large-Scale Aerial Photographs

Abstract: A detailed vegetation study was conducted at a site in the East Everglades to map the distribution of Melaleuca quinquenervia, an aggressive exotic species targeted for eradication. This high-resolution mapping effort involved the use of 1:7,000-scale color-infrared (CIR) aerial photographs and integrated geographic information system (GIS) and Global Positioning System (GPS) technologies to create a digital database of native and exotic vegetation within a 1.5- by 11-km study area. Hardcopy maps were produced at 1:5,000 scale depicting plant species distributions and information on Melaleuca height and density classes interpreted from the ~IR air photos. An accuracy assessment conducted using a helicopter yielded an overall map accuracy of 94 percent. It is anticipated that these products will allow Park managers to assess the effectiveness of exotic vegetation management practices and help to insure the preservation of native plant communities in this section of the Everglades.

Geostatistical Estimation of Resolution- Dependent Variance in Remotely Sensed Images

Abstract: tween spatial resolution and mean local variance of image The variance of a remotely sensed image is determined by data at different scales The spatial resolution at which local the interaction of scene properties with the spatial character- variance reaches a maximum is considered closely to match istics of the sensor Image variance is related to infomation the characteristic scale of scene variation The latter study content, and therefore determines the ability to extract useful and others (Markham and Townshend, 1981; Cushine, 1987) information about scene conditions We describe a technique discuss the relevance of resolution-dependent effects to the to estimate image variance at multiple spatial resolutions accuracy of mdtis~ectral image classification Specifically, The is useful for comparing the capabilities of sen- spatid resolution determines the relative variability between sors with differing spatial responses and within land-cover classes, influencing spectral separabilThe point-spread function (PSF) and the variogmm quan- ity- Marceau et al (1994a; 1994b) use the resolution dependti& the spatial characteristics of the sensor and image, re- ence of classification accuracy to assess optimum spatial spectively A geostatistical model based on these two resolutions for feature extraction elements relates the punctual variogram of a scene with the Fried1 et al (1995) and Fried1 (1997) describe the use of variogram of an image hi^ model forms the ba- a Scene simulation model to investigate the precision with sis for a numerical approach to approximate the punctual which biophysical properties can be inverted from remotely mriopm from observations ~y,~ esti- sensed data This precision is shown to depend on sensor mate of the punctual van'ogram allows analytical determina- characteristics Hu and Islam (1997) present a model which tion of image variance at different spatial resolutions explicitly relates the error of such biophysical models to the Analysis of simulated images confirms the utility of this variance of the input data, which is determined by the spaalgorithm Variance of coarse-resolution images may be esti- tial mated reliably from fine-resolution data Simulations of mul- Many studies of scale dependence take an empirical aptiscale variability show that the method handles more

The DISCover validation image interpretation process

Abstract: Thematic validation of the International Geosphere Biosphere Data and Information System (IGBP-DIS) Global 1-Kilometer Land-Cover Data Set (DISCover) was performed utilizing a state-of-the-practice technique by a team of Expert Image Interpreters (EII) examining subscenes extracted from 379 digital Landsat TM (Thematic Mapper) and SPOT images. The 15 validated IGBP land-cover classes (Snow/Ice and Water were not assessed) were not equally interpretable on the TM and SPOT imagery, Interpreter confidence was highest for Evergreen Broadleaf Forests and Urban/Built-up DISCover classes while Grasslands and Permanent Wetlands were interpreted with relatively less confidence. Analysis of image interpretation in each of the 13 validation regions indicates that confidence in interpretations for North America/Canada (Region 1) and Central Asia/Japan (Region 11) are lower than average. Confidence in interpretations is significantly higher than average for North America/US (Region 2), Northern and Southern South America (Regions 4 and 5), and Southest Asia and China (Region 12). In this study, variations in interpretation confidence are also noted between regions or based upon the geographic location of samples. This exercise demonstrates that Landsat TM and SPOT imagery can be efficiently used to validate high-resolution global land-cover products. The results suggest that the utility of and confidence that may be placed in this technique depends upon the land-cover classification scheme used and the quality of digital and ancillary data available to aid interpreters. Another important factor is the relative confidence of the interpreters to verify the land cover within their respective areas of the globe.

Remotely Sensed Estimates of Vegetation Structural Characteristics in Restored Wetlands, Southern California

Abstract: Traditional field sampling approaches for ecological studies of restored habitat can only cover small areas in detail, con be time consuming, and are often invasive and destructive. Spatially extensive and non-invasive remotely sensed data can make field sampling more focused and efficient. The objective of this work was to investigate the feasibility and accuracy of hand-held and airborne remotely sensed data to estimate vegetation structural parameters for an indicator plant species in a restored wetland. High spatial resolution, digital, multispectral camera images were captured from an aircraft over Sweetwater Marsh (San Diego County, California) during each growing season between 1992-1996. Field data were collected concurrently, which included plant heights, proportional ground cover and canopy architecture type, and spectral radiometer measurements. Spartina foliosa (Pacific cordgrass) is the indicator species for the restoration monitoring. A conceptual model summarizing the controls on the spectral reflectance properties of Pacific cordgrass was established. Empirical models were developed relating the stem length, density, and canopy architecture of cordgrass to normalized-difference-vegetation-index values. The most promising results were obtained from empirical estimates of total ground cover using image data that had been stratified into high, middle, and low marsh zones. As part of on-going restoration monitoring activities, this model is being used to provide maps of estimated vegetation cover.

Unveiling the Lithology of Vegetated Terrains in Remotely Sensed Imagery

Abstract: -Forced invariance- is a processing method that can subdue the expression of vegetation and enhance the expression of the underlying lithology in remotely sensed imagery.

Development of Landscape Metrics for Characterizing Riparian-Stream Networks

Abstract: Sampling methods and functionally related landscape metric~ were developed for characterizing riparian-stream networks using aerial photography and GI~. A sample area was empirically derived by using morphological characteristics of increasing portions of the stream network surrounding points selected on streams. GIs functions were used to band stream networks in 10-m increments to a distance of 300 m, within which land cover was interpreted from aerial photographs and digitized. Incremental banding is an effective approach for characterizing the composition and pattern of land cover as a function of distance from the stream network. Structural attributes that capture the linear nature of riparian-stream networks, such as the composition, width, longitudinal extent, and connectivity of woody vegetation, were characterized. The methods developed provide a flexible framework for deriving landscape metrics of functionally important structural attributes of riparian-stream networks for exploring relationships at varying spatial scales with indicators of stream ecological condition.