Thomas E. Burk

WMS and GML based interoperable web mapping system

Recently the World Wide Web has become a popular vehicle for information distributation and web based geographic information system (GIS) are rapidly evolving and adapting to these new environments. The main hindrance for building true interoperable distributed geographic information systems is the lack of any standard exchange mechanism between the diverse GISes connected over the web. Recent efforts by the OpenGIS Consortium have resulted in several specifications to alleviate these problems. Web Map Server (WMS) and Geographic Markup Language (GML) are such standards for developing interoperable web based Geographic Information Systems (Web-GIS). GML is an XML (eXtensible Markup Language) encoding for the transport and storage of geographic information, including both geometry and properties of geographic features. In this paper we describe a WMS compliant map server and GML based client. This integrated system leads to a true interoparable Web-GIS. GML based client for the first time offers client side query processing capabilities and at the same time provides several challenges. The parsing techniques have performance considerations since the size of GML documents is generally huge and often the queries result in multiple passes over these documents. In this study we also evaluated the two well known parsing approaches - simple API for XML (SAX) and the documents object model (DOM) for single and multiple passes. Our study shows that SAX performs better than DOM for single pass; thus for simple applications like visualization, subsetting, the SAX model is superior. However, for intensive applications involving queries requiring multiple passes over documents or integration of multiple documents in a distributed environment, DOM based parsing offers a better solution.

UMN-MapServer: a high-performance, interoperable, and open source web mapping and geo-spatial analysis system

Recent advances in Internet technologies, coupled with wide adoption of the web services paradigm and interoperability standards, makes the World Wide Web a popular vehicle for geo-spatial information distribution and online geo-processing. Web GIS is rapidly evolving and adapting to advances in Internet technologies. Web GlSes are predominantly designed under a thin-client / fat-server paradigm. This approach has several disadvantages. For example, as the number of users increases, the load on the server increases and system performance decreases. Recently the focus has been shifted towards client-side Web GISes, which are heavy-duty, stand-alone systems. We take an opposing approach and present a load balancing client/server Web-based spatial analysis system, UMN-MapServer, and evaluate its performance in a regional natural resource mapping and analysis (NRAMS) application which utilizes biweekly AVHRR imagery and several other raster and vector geo-spatial datasets. We also evaluate alternative approaches and assess the pros and cons of our design and implementation. UMN-MapServer also implements several open standards, such as, WMS, WCS, GML and WFS. In this paper, we also describe in detail the WMS, WCS,and GML extensions from the interoperability point of view, and discuss issues related to adoption of such standards.

Behavioral and physiological responses of American black bears to landscape features within an agricultural region

Human activities and variation in habitat quality and configuration have been shown to influence space use patterns in many species, but few studies have documented the physiological responses of free-ranging animals to these factors. We combined remote biologger technology, capturing continuous heart rate values, with locational data from GPS collars to investigate the behavioral and physiological reactions of American black bears (Ursus americanus) to a landscape dominated by agriculture (52.5% areal cover). Our study occurred at the edge of the range of this species, with small, scattered patches of forest within a mosaic of crop fields and an extensive road network. However, only ~2–4% of the area contained crops that bears consumed (corn, sunflowers, oats). We used GPS locations to identify the habitat that bears occupied, and to estimate their rates of travel. Heart rates increased with movement rates, rising by over 30% from resting rate to their fastest travel speeds. We used a modeling approach t...

A semi-supervised learning method for remote sensing data mining

New approaches are needed to extract useful patterns from increasingly large multi-spectral remote sensing image databases in order to understand global climatic changes, vegetation dynamics, ocean processes, etc. Supervised learning, which is often used in land cover (thematic) classification of remote sensing imagery, requires large amounts of accurate training data. However, in many situations it is very difficult to collect labels for all training samples. In this paper we explore methods that utilize unlabeled samples in supervised learning for thematic information extraction from remote sensing imagery. Our objectives are to understand the impact of parameter estimation with small learning samples on classification accuracy, and to augment the parameter estimation with unlabeled training samples to improve land cover predictions. We have developed a semi-supervised learning method based on the expectation-maximization (EM) algorithm, and maximum likelihood and maximum a posteriori classifiers. This scheme utilizes a small set of labeled and a large number of unlabeled training samples. We have conducted several experiments on multi-spectral images to understand the impact of unlabeled samples on the classification performance. Our study shows that though in general classification accuracy improves with the addition of unlabeled training samples, it is not guaranteed to get consistently higher accuracies unless sufficient care is exercised when designing a semi-supervised classifier

A Web-based browsing and spatial analysis system for regional natural resource analysis and mapping

Proper use and monitoring of our land and environmental resources for quality of life and sustainable growth considerations require that timely and accurate data on land cover and land use be regularly available. Remote sensing is an appropriate tool and NOAAs 1-km AVHRR time series imagery has proven to be an invaluable input for regional level natural resource mapping and monitoring. Many previous efforts addressing the use of satellite-derived data for land cover and land use classification have resulted in the development of useful products. However, they have seldom been widely applied due to the lack of an efficient and easy-to-use delivery mechanism. The World Wide Web has become popular as a vehicle for information distribution and client/server applications. Web GIS is rapidly evolving and adapting to advances in Internet technologies. Web GISes are predominantly designed under a “thin-client / fat-server” paradigm. This approach has several disadvantages. For example, as the number of users increases, the load on the server increases and system performance decreases. Recently the focus has been shifted towards client-side Web GIS. In this paper we present a balanced client/server Web-based spatial analysis system and evaluate its performance in a regional natural resource mapping and analysis application which utilizes biweekly AVHRR imagery and several other raster and vector geospatial datasets. We also evaluate alternative approaches and assess the pros and cons of our design and implementation.

A test of nonparametric smoothing of diameter distributions

Smoothing a sample‐based diameter distribution can provide a potentially better paradigm of a population distribution. Goodness‐of‐fit comparisons between raw cumulative distribution functions, nonparametric curves and Weibull densities fit to the raw distribution are presented based on two simulated forest stands.

An efficient spatial semi-supervised learning algorithm

We began by developing a semi-supervised learning method based on the expectation-maximization (EM) algorithm, and maximum likelihood and maximum a posteriori classifiers (MLC and MAP). This scheme utilizes a small set of labeled and a large number of unlabeled training samples. We conducted several experiments on multi-spectral images to understand the impact of unlabeled samples on the classification performance. Our study shows that although, in general, classification accuracy improves with the addition of unlabeled training samples, it is not guaranteed to achieve consistently higher accuracies unless sufficient care is exercised when designing a semi-supervised classifier. We also extended this semi-supervised framework to model spatial context through Markov random fields (MRF). Initial experiments showed an improved accuracy of the spatial semi-supervised algorithm (SSSL) over MLC, semi-supervised, and MRF classifiers. An efficient implementation is provided so that the SSSL can be applied in production environments. We also discuss some open research problems.

Long-term growth trends of red spruce and fraser fir at Mt. Rogers, Virginia and Mt. Mitchell, North Carolina

Cross-sectional area growth and height growth of Fraser fir and red spruce trees growing in Virginia and North Carolina were analyzed to identify possible long-term growth trends. Cross-sectional area growth provided no evidence of growth decline. The individual discs were classified according to parameter estimates of the growth trend equation. The predominant pattern of growth was a steady increase followed by fluctuation about a horizontal line. Other cross-sections exhibited a steady increase throughout the series. The only discs that represent declining growth patterns were from trees in subordinate crown position or which had previous top damage. No unexplained growth decline was present in any disc. The results regarding height growth were uncertain. A slight decline in height growth was present although we suggest that this observation was due to problems with the data or the model used to fit height growth. These findings contradict other studies suggesting that a recent growth decline has occurred in red spruce in the southern Appalachians.

Fitting process-based models with stand growth data: problems and experiences

Abstract A process-based growth model (PBM) for even-aged stands is presented which is based on photosynthesis relationships and provides predictions in terms of tree dimensions. The model has been simplified with only the most essential aspects of photosynthesis, physiology, and tree structure included. As a result, the model has fewer parameters than a typical PBM. Some of these parameters are aggregation of physiological and biometrical constants and thus summarize the effect on stand growth of many PBM parameters. Problems arising in estimating the parameters of a PBM are discussed, using the present model as an example. Finally, an example of calibrating the present model for different sites using typical measurements of stand growth is presented.

*Miner: A Suit of Classifiers for Spatial, Temporal, Ancillary, and Remote Sensing Data Mining

Thematic classification of multi-spectral remotely sensed imagery for large geographic regions requires complex algorithms and feature selection techniques. Traditional statistical classifiers rely exclusively on spectral characteristics, but thematic classes are often spectrally overlapping. The spectral response distributions of thematic classes are dependent on many factors including terrain, slope, aspect, soil type, and atmospheric conditions present during the image acquisition. With the availability of geo-spatial databases, it is possible to exploit the knowledge derived from these ancillary geo-spatial databases to improve the classification accuracies. However, it is not easy to incorporate this additional knowledge into traditional statistical classification methods. On the other hand, knowledge-based and neural network classifiers can readily incorporate these spatial databases, but these systems are often complex to train and their accuracy is only slightly better than statistical classifiers. In this paper we present a new suit of classifiers developed through NASA funding, which addresses many of these problems and provide a framework for mining multi-spectral and temporal remote sensing images guided by geo-spatial databases.