Raja Sengupta

Development and Comparison of Approaches for Automated Mapping of Stream Channel Networks

Abstract Accurate mapping of stream channel networks is important for measuring hydrologic parameters, for site planning in construction projects, and for use in hydrologic models. This article compares five existing and two new methods for extracting stream channel networks for use in topographic mapping. In order of increasing accuracy, these methods are: (1) blue lines on USGS 1:24,000 topographic maps (64.6 percent underrepresentation), (2) placing stream heads using a constant flow-accumulation area to mimic USGS blue lines (47.8 percent underrepresentation), (3) constant flow-accumulation area equal to the mean for identified channel heads (30.3 percent combined under- and overrepresentation), (4) variable flow-accumulation area estimated by multiple linear regression (28.9 percent combined under- and overrepresentation), (5) variable flow-accumulation area estimated by a slope-power relationship (23.6 percent combined under- and overrepresentation), (6) identifying stream cells using logistic regression (12.7 percent combined under- and overrepresentation), and (7) extracting stream channel head locations from digital orthophotoquads (DOQs) (nearly 100 percent accurate, but only applicable under ideal conditions). Methods 2–6 require 10 m resolution digital elevation models that can be acquired directly in many areas or can be derived from 1:24,000 hypsography where available; Methods 4 and 6 are new methods developed in this paper. Using DOQs, while extremely accurate, is labor intensive and can be applied only in a small minority of locations where vegetation cover does not obscure channel head locations. We conclude that identifying stream cells using logistic regression has the broadest applicability because it can be implemented in an automated fashion using only DEMs while still achieving accuracies for mapping low-order streams that are far superior to existing USGS maps.

Agent-based modelling environment for spatial decision support

The goal of Spatial Decision Support Systems (SDSS) is to assist decision-makers as they generate alternative solutions to a variety of semi-structured geographical problems, and to evaluate these solutions with the help of applicable data and analytical models. Most existing SDSS, however, support only a limited number of decision-making environments, and are not designed to utilize web-accessible repositories of spatial data and models. These limitations are overcome through the use of ‘software agents’ within an agent-oriented modelling framework, called ‘Distributed Intelligent Geographical Modelling Environment (DIGME)’. The utility of this framework is demonstrated through the development of an SDSS to evaluate the ecological and economic impacts of agricultural policy for the Cache River watershed of southern Illinois.

Spatial Scale and Population Assignment Choices in Environmental Justice Analyses1

Abstract Environmental justice laws protect certain populations against discriminatory actions that may result from a myriad of enterprises, including transportation activities. Previous environmental equity studies examining the effects of transportation-engendered externalities have been criticized on several points, including (1) that the choice of a reference population for comparison to the criterion variable may influence the outcome of research results and (2) that the selection and use of inappropriate methodologies intended to identify and characterize populations may foreordain research outcomes. This article examines the potentially confounding effects of selected spatial scale and population assignment strategies as applied to a study of excessive noise levels at a large Midwestern airport, finding that reported outcomes can vary significantly as a function of methodological choices.

Spatial patterns of illegal resource extraction in Kibale National Park, Uganda

SUMMARY Conservation policy typically excludes people from national parks and manages encroachment by law enforcement. However, local people continue to extract resources from protected areas by boundary encroachment and poaching. This paper quantifies the patterns of illegal resource extraction from Kibale National Park in Uganda, the demand for Park resources by communities bordering the Park, and examines whether designated resource access agreements reduce illegal extraction. Sections of the Park boundary were examined and human entry trails, wood extraction, livestock grazing, and animal poaching signs were quantified. Levels of illegal extraction were compared with the demand for and admitted illegal access to resources inside the Park, collected in a survey of households located near the Park. Extraction was also compared between villages with and without negotiated resources access agreements. The most wanted and extracted resource from the Park was wood for fuel and construction. Implementation of resource access agreements with local community associations was found to be an effective means of reducing illegal extraction, but only if the association members profited from the agreement.

Modeling Enrollment in the Conservation Reserve Program by Using Agents within Spatial Decision Support Systems: An Example from Southern Illinois

Existing models of agricultural decisionmaking based on economic optimization often fall short of capturing the complex dynamics of land-use choices at both individual parcel and watershed-level scales. The complexity arises from an interplay of several factors, as explained by Herbert Simons model of bounded rationality, the theory of diffusion of innovations through spatial contagion, the role of personal environmental values and local culture, and simple historical momentum. This complexity can be captured using ‘artificial life agents’ that model land-use choice for individual parcels by considering characteristics and personal beliefs of the owner or operator, physical traits of the land, and information obtained via social networks. Agents are therefore able to consider holistically a large number of factors affecting land-use choice. The creation of agent-based models of human behavior described herein is based upon empirical data on the acceptance of Conservation Reserve Program for the Cache River watershed of southern Illinois (USA). These models are interfaced with a geographic information system to produce a spatial decision support system capable of anticipating the effects of policies that affect land-use decisionmaking on a real landscape and their economic performance.

Geospatial Agents, Agents Everywhere . . .

The use of the related terms “agent-based”, “multi-agent”, “software agent” and “intelligent agent” have witnessed significant growth in the Geographic Information Science (GIScience) literature in the past decade. These terms usually refer to both artificial life agents that simulate human and animal behavior and software agents that support human-computer interactions. In this article we first comprehensively review both types of agents. Then we argue that both these categories of agents borrow from Artificial Intelligence (AI) research, requiring them to share the characteristics of and be similar to AI agents. We also argue that geospatial agents form a distinct category of AI agents because they are explicit about geography and geographic data models. Our overall goal is to first capture the diversity of, and then define and categorize GIScience agent research into geospatial agents, thereby capturing the diversity of agent-oriented architectures and applications that have been developed in the recent past to present a holistic review of geospatial agents.

Agent‐Based Simulation of Urban Residential Dynamics and Land Rent Change in a Gentrifying Area of Boston

Certain complex processes are most effectively modeled not on the macro-scale, but from the bottom-up, by simulating the decisions of individual entities, or agents. This study uses an agent-based modeling (ABM) approach to simulate residential dynamics in an area of Boston that has increasingly experienced gentrification in the past decades. The model is instantiated using basic empirical data and uses simple decision-making rules, differentiated into four classes, to simulate the process of residential dynamics. The model employs the consumption explanation of the cause of gentrification, which emphasizes the choices of individuals drawn to urban amenities, while testing the production explanation, which suggests that major investments from the public and private sphere attract and explain gentrification. Verification shows that the processes in the model work according to its construction, simulates complexity and emergent phenomena, and may be a valuable explanatory tool for understanding and learning about some processes underlying gentrification.

A spatial decision support system to identify species-specific critical habitats based on size and accessibility using US GAP data

Abstract The Gap Analysis Program (GAP) is a nationwide effort to find areas of suitable habitat in the US, which if protected from habitat degradation, may help to preserve the native animal and plant biodiversity. In recent years, the GAP protocols used to analyze habitat data have become more scale and species dependent. This research describes the creation of a Spatial Decision Support System (SDSS) that applies species-specific parameters of Individual Area Requirement (IAR) ( Vos et al., 2001 ), Minimum viable population (MVP), and Reach ( Allen et al., 2001 ) to determine critical habitats for animal species, thereby eliminating those areas that are effectively unusable because of size or inaccessibility. The utility of the SDSS, and the three algorithms contained within it (i.e., core area, core area growth and aggregate), is demonstrated by creating distribution maps of usable habitat for five species (i.e., alligator, black bear, bobcat, gray fox and wild turkey) commonly found in the state of Arkansas. This knowledge can then be used to guide and prioritize conservation efforts towards protecting usable, and often critical, habitats.

Landscape characteristics of Rhizophora mangle forests and propagule deposition in coastal environments of Florida (USA)

Field dispersal studies are seldom conducted at regional scales even though reliable information on mid-range dispersal distance is essential for models of colonization. The purpose of this study was to examine the potential distance of dispersal of Rhizophora mangle propagules by comparing deposition density with landscape characteristics of mangrove forests. Propagule density was estimated at various distances to mangrove sources (R. mangle) on beaches in southwestern Florida in both high-and low-energy environments, either facing open gulf waters vs. sheltered, respectively. Remote sensing and Geographic Information Systems were used to identify source forests and to determine their landscape characteristics (forest size and distance to deposition area) for the regression analyses. Our results indicated that increasing density of propagules stranded on beaches was related negatively to the distance of the deposition sites from the nearest stands of R. mangle and that deposition was greatly diminished 2 km or more from the source. Measures of fragmentation such as the area of the R. mangle forests were related to propagule deposition but only in low-energy environments. Our results suggest that geographic models involving the colonization of coastal mangrove systems should include dispersal dynamics at mid-range scales, i.e., for our purposes here, beyond the local scale of the forest and up to 5 km distant. Studies of mangrove propagule deposition at various spatial scales are key to understanding regeneration limitations in natural gaps and restoration areas. Therefore, our study of mid-range propagule dispersal has broad application to plant ecology, restoration, and modeling.

Evaluating the Impact of Policy-induced Land Use Management Practices on Non-point Source Pollution Using a Spatial Decision Support System

Abstract State and federal conservation compliance policies in the United States are aimed, in part, at reducing non-point pollution and maintaining long-term agricultural productivity. These policies affect land use and management practices on the nations farms. Farms and farmers, however, are part of a larger agri-ecological system and changes in farming practices affect linked socio-economic, biologic, and hydrologic systems. It is difficult, therefore, to foresee the wide ranging and long-term consequences that are set into motion by changes in agricultural policy. Yet, these consequences must be understood if we are to avoid the deleterious side effects or capture the collateral benefits associated with specific policies. A Spatial Decision Support System (SDSS) has been developed here to help understand some of these consequences within the context of watershed management and the control of non-point source pollution. This SDSS consists of a Geographic Information System (GIS), two spatial models (GEOLP and AGNPS), and a graphical user interface. The purpose of this SDSS is to assist decision-makers as they investigate the impact of agricultural policy on non-point source pollution and the local economy. The SDSS is used to simulate the potential impacts of alternative policy scenarios in the Big Creek watershed. The models and methodologies described here, however, are general and can provide benefits to a variety of decision-makers engaged in watershed management and the reduction of non-point pollution.