Bridge Management System
Josephine Amato, Westchester County Geographic Information Systems (GIS), Department of Information Technology, 148 Martine Avenue, Room 305, White Plains, NY 10601 [Phone: (914)995-3853 Email: Jja9@westchestergov.com]

Over the past six months, Westchester County GIS and the Department of Public Works have been working together on the development of an entry-level Bridge Management System (BMS). Developed with Dialog Designer extension in ArcView GIS 3.2, the BMS provides access to both Westchester County and New York State Department of Transportation (NYS DOT) data, as well as the functionality to include photographs, as-built drawings, and inspection reports for each bridge. While designed primarily for county government, the application will also be made available for local government GIS programs. This proposed conference poster will include a series of ArcView screen shots illustrating a range of data products available in the application.

Assessing Systematic Positional Error in Automated Geocoding of Residential Addresses
Michael R. Cayo and Thomas O. Talbot, Geographic Research and Analysis Section, Bureau of Environmental & Occupational Epidemiology, New York State Department of Health, 547 River St, Room 200, Troy, NY 12180-2216  [Phone: (518)402-7960 Email: mrc02@health.state.ny.us, tot01@health.state.ny.us]

The use of Geographic Information System (GIS) technology is rapidly becoming an integral part of many public health studies and analyses. GIS offers the potential to analyze spatial relationships between disease and environmental exposure to contaminants. Geocoding is one component of this technology that allows researchers to match study participant residential addresses to reference files containing geographic coordinates. Often times the residence location is used as a surrogate measure of exposure and to determine whether that residence is within a zone, which may be potentially impacted from a hazardous waste site or other emission source.

Although many different types of error are possible, this project focuses on developing methods that can provide qualitative and quantitative measures of positional error when residential addresses are geocoded through an automated process. This can better prepare researchers in evaluating whether this error may affect the results of their research. Future research will concentrate more specifically as to how the measured error effects the misclassification of study participants in a health study with respect to exposure.

Westchester County GIS Web Based Interactive Maps
Xiaobo Cui, Geographic Information System, Department of Information Technology, Westchester County, 148 Martine Ave, Room 305, White Plains, NY 10601 [Phone: (914)995-3781 Fax: (914)995-3269 E-Mail: xxc1@westchestergov.com]

Westchester County GIS has recently developed a series of web based interactive maps which are used in supporting both county and local government applications. This conference poster will highlight applications in the areas of the West Nile virus, 2001 County Legislative Redistricting, samples of orthophoto image and planimetric data from the on-going county-wide base mapping project, and an environmental features map.

Westchester County GIS has also implemented the ESRI geodatabase model using standard Oracle relational databases with an ArcSDE application server. The poster will also include an architecture diagram for producing web-based interactive maps which includes ESRI’s ArcIMS.

Municipal GIS Development and Support in Westchester County, New York
Ana Hiraldo, Westchester County Geographic Information Systems (GIS), Department of Information Technology, 148 Martine Avenue, Room 305, White Plains, NY 10601 [Phone: (914)285-4416 Fax: (914)285-3269 E-mail: aeh2@westchestergov.com]

Westchester County GIS staff has been providing technical support, consulting services, training, and other GIS related services to local municipalities for over a decade. This conference poster is a composite of examples of recent work by Westchester County GIS in support of local government GIS projects. Examples include work with the Town of Greenburgh (infrastructure), Town of Lewisboro (open space), Village of Hastings-on-Hudson (trail mapping), City of Mt. Vernon (recycling districts), and Town of Cortlandt (tax mapping and digital orthophotography).

Utilizing GIS in the Support and Development of a Computer Aided Dispatch (CAD) System for Westchester County, New York
Carrie Keneally and Greg Sullivan, Geographic Information Systems, Department of Information Technology, Westchester County, 148 Martine Ave, Room 305, White Plains, NY 10601 [Phone: (914)995-3014 Fax: (914)995-3269 E-Mail: cek1@westchestergov.com]

Westchester County GIS, working in concert with the county’s Department of Emergency Services, is supporting the development of a new state-of-the art computer aided emergency dispatching system. While overall system development requires the integration of several systems (records management, E911, etc.), county GIS staff are utilizing both GIS and GPS tools to maintain and develop spatial data layers such as road networks, mile markers, police and fire stations, district boundaries, and special features such as schools, libraries, hospitals, and similar congregate care facilities. Currently ESRI software tools are being used for spatial data development and maintenance which are then imported into the new Intergraph Public Safety (IPS) dispatching system. This conference poster will include a series of maps which highlight both mapping and dispatching functions associated with the IPS system.

How America Voted: A Spatial Approach for Analyzing Voting Patterns During the 2000 Presidential Election
Arthur J. Lembo, Jr., Department of Crop and Soil Sciences, Cornell University, 305 Rice Hall, Ithaca, NY 14853 [Phone: (607)255-6328 Email: ajl53@cornell.edu] and
Paul Overberg, USAToday. [Email: poverberg@usatoday.com]

Although the 2000 Presidential election was one of the closest in recent history, many commentators noted that the voting patterns exhibited a striking "cultural divide", with urban areas voting for Al Gore, and rural areas voting for George W. Bush. These comments were primarily based on a subjective view of county voting patterns during the election. This project attempts to provide a quantifiable measurement of the voting patterns exhibited during the 2000 election. Specifically, we were interested in determining if a statistically significant clustering pattern existed based on county-wide results, and if each candidate won their assumed cultural association (Gore: Urban; Bush: Rural).

To test these hypotheses, two separate spatial analysis methods were performed on county-wide voting patterns within the United States. The first method utilized a principle of spatial autocorrelation called join count analysis to determine if voting patterns exhibited evidence of spatial clustering. The second method used map overlay and chi-square analysis to determine the correlation between urban areas and votes for Al Gore, and rural areas with votes for George W. Bush.

Combining the USLE and GIS/ArcView for Soil Erosion Estimation in Fall Creek Watershed in Ithaca, New York
Jianguo Ma, Department of Agricultural & Biological Engineering, Cornell University, 58 Riley-Robb Hall, Ithaca, NY 14853

Soil erosion by water has been identified as a research priority today. To model erosion a vast amount of data needs to be included in order to attempt to accurately predict how much soil will be moved from one point to another. The amount of information that must be accounted for and prepared for entry into the model can make modeling a very labor-intensive operation. The automated method allows for storage, manipulation, analysis and display of the model with high accuracy and efficiency. Geographical information systems (GIS) are being evaluated as a means of improving such modeling practices. The study area selected is the Fall Creek watershed in the Ithaca area of New York State.

The major research findings and conclusions are as following: The USLE was used to predict soil erosion in the Fall Creek Watershed. As seen from the derived erosion map, most areas have minor soil erosion which is less than 1 tons/ac/yr. The areas of highest erosion occurred in the places where the slopes are the greatest and also the places located near the water edge. The reason probably is that the sediment travel time before entering the lake is minimum in these places compared to other places. The estimated erosion values mostly are below the usual "tolerable soil loss" of 3 tons/ac/yr. Because of extreme slopes throughout the Fall Creek Watershed, the LS values were possibly overestimated as the USLE was originally developed for mild slopes in agricultural areas.

Development of the Westchester County Base Map
Laura McGinty, Geographic Information System, Department of Information Technology, Westchester County, 148 Martine Ave, Room 305, White Plains, NY 10601 [Phone: (914)995-3888 Fax: (914)995-3269 E-Mail: lam7@westchestergov.com]

In 1998, Westchester County began planning for the development of the first-ever, digital, high-accuracy (1’ = 100’) base map of the entire county. Covering the entire 486-square miles of the county (and a 200’ buffer beyond the county boundary), the project was designed to produce a wide range of digital products which could be used and integrated into the growing number of government applications based on spatial data (emergency dispatching, transportation, infrastructure management, tax mapping, health and human services, etc.), as well as a wide range of basic geographic information systems (GIS) initiatives. The project is being carried out at the direction of the County Executive and the Westchester GIS Task Force, a group which includes representation from county and local government, business, and utilities.

Examples of the various products, which will be delivered upon final completion of the project, will be illustrated in this poster in map form. The deliverables will include; true color digital orthophotography images in both compressed Mr. SID and uncompressed TIFF file formats, over 30 layers of planimetric data with 5’ contour lines, and a county-wide Digital Elevation Model (DEM), which was compiled using Light Detection And Ranging (LIDAR).

Social Accounting Matricies and Landscape Change in Dutchess County, New York
Audra Nowosielski, Department of Economics, Rensselaer Polytechnic Institute 110 8th Street Troy, New York 12180 [Email: nowosa@rpi.edu]

The overall goal of this study of development in the Hudson River valley is to incorporate measures of economic activity, land use change, and environmental quality together in a linked framework capable of evaluating scenarios for policy analysis. The analytical building blocks for the ecological economic model include a Social accounting Matrix (SAM), a Geographical Information System (GIS) of land use and geophysical attributes, and a series of indicators of river ecosystem functions. Most economic models do not include spatial variation of activity, however, location is critical to estimating environmental loading. In this project, GIS data is incorporated to incorporate spatial aspects as well as (1) characterize the household sectors of the SAM model, (2) locate business establishments within the watershed to regionalize the SAM model, (3) provide a land inventory for local development scenarios, and (4) characterize land parcels by location and geophysical attributes.

BASINS Modeling of the Buffalo River Watershed: Affect of Rainfall Variability on Hydrologic Modeling
Mary Perrelli, Erica Somogye and K.N Irvine, Department of Geography andPlanning, Buffalo State College, 1300 Elmwood Avenue, Buffalo, NY 14222 [Phone: (716)878-6204 Email: perrelmf@bscmail.buffalostate.edu]

The EPA's Better Assessment Science Integrating Point and Non-point Sources system (BASINS) (version 2.1) is composed of a suite of mathematical models that can be applied in support of watershed planning and water quality analysis. The BASINS model and related databases are available for download from the EPA web site. The spatial data consist of four types: (1) watershed boundary and associated data such as land use, soil, gauge stations; (2) elevation data; (3) general and detailed stream network data; (4) meteorological data. The system uses ArcView GIS as a platform.

In our study, the Non Point Source Model (NPSM) extension of BASINS was used to compare USGS observed flow data to simulated flow within the Buffalo River watershed. USGS gauge stations on Cazenovia Creek, Buffalo Creek, and Cayuga Creek were used for model calibration and validation. The accuracy of flow estimates at these sites is dependent on the quality of the rainfall data uses as model input. The meteorological data available with BASINS are recorded at the Greater Buffalo Airport. This weather station lies several miles outside of the Buffalo River watershed to the north. Historical model simulations for 1990, 1992 and 1995 revealed that discrepancies in peak flow rates were largely the result of the inability of the Buffalo Airport site to accurately reflect the spatial variability of storms observed at the eastern end of Lake Erie.

This poster presents the use of NPSM to model hydrology of the Buffalo River Watershed from an initial project set up for calibration and end results. It concludes with a discussion of modeling inadequacies, specifically with rainfall data. ArcView Spatial Analyst is used to show the spatial variability of rainfall within the Buffalo River watershed using a network of local rain gauge sites.

A Spatial Approach for Linking Economic Trends to Land Use Change in the Hudson River Valley
John M. Polimeni, Department of Economics, Rensselaer Polytechnic Institute, Troy, NY  [Phone: (518) 377-4849 Email: j.polimeni@verizon.net]

Land use in the Hudson River Valley is determined by economic activities and social policies, and is a key determinant for the health and use of the immediate ecosystem. There is a direct correlation of land use change to the health of the Hudson River, because economic and population changes determine land use, and the land supplies nutrients to the river.

Land use change models sensitive to local characteristics are needed to develop scenarios for evaluation. To achieve these projections, a GIS database is employed to do a socio-economic analysis of households within Dutchess County in New York. The database contains detailed socio-economic household characteristics at the census block level for the 1990 census year and year 2000 and 2005 projections. County-level tax parcel data is used in conjunction with bio-geophysical attributes to calculate and project growth trends for developed and undeveloped land parcels. The GIS was used to calculate distances from each parcel to central business districts in the county, as well as to create a neighborhood index that provides a weighting system for which property class a parcel will likely be developed into.

The land inventory provides the basis to forecast development trends in a dynamic mapping format. Data is assembled with a GIS software program, MapInfo. To complement the GIS maps, a pictorial history, using satellite and other imagery, is used to visually illustrate land use changes under different economic conditions.

Development of a Land Cover Database with LANDSAT Imagery in Westchester County, New York
Francesca Pozzi, Research Associate, Center for International Earth Science Information Network (CIESIN), Columbia University [Phone: (845)365-8977 Email: fpozzi@ciesin.columbia.edu] and
Sam Wear and Carrie Keneally, Westchester County GIS, Department of Information Technology, Westchester County, New York [Website: http://giswww.westchestergov.com]

With the increased availability and accessibility of geospatial data, combined with more powerful and user-friendly mapping software, local governments are building more advanced Geographic Information Systems (GIS) applications. In particular, a wide range of new remotely sensed (RS) data products are being combined with GIS to support government programs in the areas of land use and transportation planning, emergency management, and the characterization and monitoring of environmental and human health conditions. Westchester County GIS has recently teamed with the Center for International Earth Science Information Network (CIESIN) at Columbia University to develop a land cover map of the county utilizing LANDSAT imagery. The land cover map will serve as initial baseline data in comparing earlier Westchester County GIS land use inventories (1988 and 1996), as well as being combined with spatial data products being developed with the current Westchester County base mapping project.

Using Ripley’s K to Determine Data Clustering and Codependence: Applications in Kenya
Ingrid Rhinehart, Graduate Student, Department of Applied Economics and Management, Cornell University, 317A Warren Hall, Ithaca, NY 14850 [Phone: (607)275-9584 Email: iar2@cornell.edu]

Animal herd and human settlement location data obtained from a 2000 aerial census of Amboseli National Park and the surrounding Amboseli Ecosystem are analyzed for evidence of spatial clustering or non-randomness. Data points analyzed include elephants, buffalo, cattle, shoats (sheep and goats), watering holes, and bomas (human dwellings). The locations of animal herds and human settlements are also analyzed for spatial codependence between the species. These analyses are performed using the Ripley’s K function and relevant adaptations for the ‘between species’ calculation. The K functions show evidence that each data set is non-random, therefore indicating autocorrelation. Finally, the point sets are analyzed to show codependence between data sets. The codependence analysis illustrates that bomas and shoats ‘attract’ each other while elephants and shoats display repulsion. This use of Ripley’s K demonstrates some of the possible applications of spatial statistics for analysis within a GIS environment.

Armchair Flow Estimation in the Black River Watershed
Zev Ross, Cornell University Department of Natural Resources, Fernow Hall, Ithaca, NY 14853 [Email: zr24@cornell.edu]

The focus of this analysis is to create a simplified model for predicting stream flow with Arcview GIS using only landscape-based variables and a small number of flow estimates from the EPA (for calibration). Predictions from the model were compared with the actual flow measurements obtained from USGS gauge stations.

A wide variety of flow models already exist through agency and academic sources. However, these models often have a steep learning curve, require an extensive amount of data for program input, and often take days to run. This analysis examined the possibility of using a single function in Arcview GIS, applied to a limited number of landscape-based variables to estimate flow in the Black River Watershed, a 2000 square mile drainage in northern New York State.

The flexibility of ArcView’s flow accumulation function allows a user to derive a wide range of information about a single grid cell’s contributing area (watershed). In particular, the flow accumulation functions were used to derive the area, average elevation, average slope, and average precipitation. In addition, the percent of forest cover, percent of poorly drained soil and percent of urbanization for the watersheds were obtained from other sources. The total watershed area consisted of over 50 million 10 x 10m grid cells.

The ArcView data was imported into S-Plus to analyze the statistical relationships between the variables listed above and average annual flow.

The results show a very strong correlation (p << 0.0001) between mean flow and the area of the watershed. While inclusion of precipitation and soil type improves slightly the accuracy of the flow estimates, the correlation was not statistically significant at the .05 level (p=0.1077,0.1069 respectively). These numbers suggest that while precipitation and soil type improve flow estimates in this instance, the correlation may in fact be solely due to coincidence. Overall, area of the watershed alone accounts for 99.76% of the variation in flow, while the three variables combined explain 99.81%.

Using the area of the watershed, precipitation, and soil type as input variables, the model adequately predicted flow in watersheds greater than 250 square miles. On average, the model predicted the average annual flow to within 2% of the observed values for the 13 watersheds greater 250 square miles. However, the model poorly estimated flow in smaller watersheds with predictions averaging a 20% deviation from observed values. These differences are presumably due to the lower effective sample size a smaller area provides regarding the effect of rainfall on flow.

Development of a GeoSpatial Exposure Model (GeoSEM) for Human Health and Ecological Risk Assessments
William Thayer^{1, 2,} Philip Goodrum^{1, 2}, Gary Diamond¹ and James M. Hassett², ¹Syracuse Research Corporation, 6225 Running Ridge Road, North Syracuse, NY 13212, ²Faculty of Environmental and Forest Engineering, Bray Hall, SUNY-College of Environmental Science and Forestry, Syracuse, NY 13210.

One source of uncertainty in risk assessment is the exposure point concentration (EPC), which represents the chemical concentration to which a human or ecological receptor may be exposed to for a toxicologically relevant time period within a geographic area called an exposure unit (EU). Biased sampling methods are often employed during site characterization. This, coupled with the assumption that contaminants are log-normally distributed, may contribute to overly conservative estimates of the EPC, and associated risk, and may result in excessive cleanup costs. Geostatistical methods allow the spatial information present in sample data to be incorporated in the estimate of the EPC, which should reduce the uncertainty in the EPC and risk estimates. However, the application of geostatistics introduces another source of uncertainty (i.e. model uncertainty) into risk estimates and risk management decisions. While there are some examples of the use of geostatistics in risk assessment and remediation design, the available geostatistics software packages are not designed for exposure assessment specifically and, therefore, require considerable experience in geostatistics to produce estimates that are appropriate for human health and ecological exposure assessment. We have developed the GeoSpatial Exposure Model (GeoSEM), a software tool that combines geostatistical algorithms and mapping capabilities in a format that does not require the user to be an expert in geographic information systems (GIS) or geostatistics.

We believe that GeoSEM will fill a need for a user-friendly tool that risk assessors and remedial project managers can use to investigate the role of spatial information in estimating exposures to potential human and ecological receptors and that will produce more informed estimates of risk. GeoSEM will connect risk assessors with easily implemented, robust geospatial statistical routines, such as kriging (ordinary, indicator, log-normal and normal score kriging) and simulation (Gaussian and indicator), and area weighting approaches such as Thiessen polygons, within a single software platform. GeoSEM is a GIS-based application that is designed to run in the Microsoft Windows^® environment. GeoSEM is capable of using a wide variety of GIS vector and raster file formats including DOD vector product format (VPF), ESRI ArcView^® Shapefiles, ARC/INFO^® coverages and spatial database engine (SDE^®) layers, computer-aided design (CAD) drawings, binary and ascii grids and many types of standard georeferenced image formats such as geoTIFFs, bitmaps, GIFs, JPEGs and ERDAS. Linkage of GeoSem to the Integrated Stochastic Exposure (ISE) model (developed by SRC) provides risk assessors with a complete geospatial/probabilistic exposure model that can be applied to risk assessments at Superfund sites.

Spatial Analysis of 2000 Census Population Figures in Westchester County, New York
Tong Zhou, Software Architect I, Westchester County GIS, Department of Information Technology, 148 Martine Ave, Room 318, White Plains, NY 10601 [Phone: (914)285-3012 Fax: (914)285-3269 Email: taz2@westchestergov.com]

Recently U. S. Census Bureau released the 2000 Census Population data. Preliminary studies show that the total population and its racial composition of Westchester County have changed dramatically from those of 1990. This map analyzed those changes from the angle of racial distribution change of population across the county. It would be of great interest to the public.

The production of this map was done using ESRI Spatial Analyst in order to represent the data in a unique way. Utilizing the Density Function, the software distributes the measured census tract population percentage change of the input census tract centroid point theme throughout the county to produce a continuous surface. The color differences of the surface indicate the various degrees population change over the last ten years. Five variables: Total, White, Black, Asian, and Other are used to produce 5 different population distribution change maps. Other geographic features were added to the map to provide reference and context.