Johannes Feddema
Developing Information Systems for Simulating Anthropogenic Impacts in Earth System Models
It is becoming evident that a number of human activities are impacting the Earth's surface and atmosphere at spatial scales that warrant their inclusion in large scale Earth System Models (ESMs). Simulating these human activities on global scales requires that we develop sub-models to simulate each of the relevant activities, and that we develop the accompanying spatial and temporal information, or parameter datasets, to identify the location and intensity of each of the activities. Besides the impact of human predicted greenhouse gases, models are presently beginning to treat land cover change as an important boundary condition in their simulations. Significant effort is going into developing databases that represent land cover conditions in the present, past and future. To develop such simulations requires development of coordinated technical methodologies that combine remote sensing information with historical socio-economic resources and model projected future land cover changes. One short coming of land cover change models is that they typically do not include land use explicitly, where the distinction between land cover and land use might include human actions that are not reflected explicitly in land cover datasets. For example activities such as fertilizer applications that affect the carbon cycle, thinning forest cover in other locations, and different morphologies and building materials associated with urban land cover all affect climate outcomes without altering land cover classification systems. Ultimately GIScience technologies need to be implemented to develop datasets that represent all of these human actions. A secondary benefit of developing such datasets for model input boundary conditions is that the same datasets also provide valuable resources for climate impact assessment programs because they provide a better source of information to assess climate impacts on human land management practices. [PDF]

Alex de Sherbinin
Socioeconomic data for climate change impacts, vulnerability and adaptation assessment
This presentation begins by presenting a number of vulnerability definitions and formulas, highlighting how spatial data are vital for climate change impacts, vulnerability and adaptation assessment. The presentation then reviews a number of CIESIN applications related to population distribution and hazards, poverty and vulnerability to climate hazards, emerging infectious diseases, and civil conflicts and water resources. In addition to these applications, the presentation identifies major approaches found in the literature. Data needs for adaptive capacity are also briefly reviewed, and the presentation concludes by describing a proposed climate change IVA “data kit” that would meet the needs of low income countries under the Nairobi Work Programme of the UN Framework Convention on Climate Change. [PDF]

Julie Demuth
Exploring Variations in People's Sources, Perceptions, and Uses of Weather Forecast Information
Every day, the U.S. weather enterprise collectively provides multiple weather forecasts to a broad public audience, for both normal and high-impact weather. Although the weather community has a general and often experientially based sense of the public arena of weather forecasting, empirical research in this area can improve our understanding of people's attitudes and behaviors regarding weather forecast information and how these vary among individuals.  Better understanding of the societal aspects of forecasts can then help improve future development and communication of forecast products to improve people's use of the information.  This presentation will discuss recent research efforts to collect quantitative and qualitative information from members of the public about weather forecast information. Among these research efforts is a nationwide, controlled-access Internet survey of the general public conducted in 2006 with 1465 responses.  The survey included questions to assess people's sources, perceptions, uses, and values for everyday weather forecast information. Based on their reported zip codes, respondents were matched with climatological observation data and forecast verification measures to begin exploring how people's attitudes and behaviors for weather forecast information vary based on their socio-demographic characteristics, their experiences with weather based on where they live, and their outdoor lifestyle factors.[PDF]

Marc P. Armstrong
Some Challenges in Using Individual-Level Data in GIS-based Research
Many data sets contain records that have implicit location identifiers, such as street addresses. GIS software enables researchers to attach explicit coordinate and other locational identifiers to such records through a process known as geocoding. Records with added geocodes can then be mapped and analyzed using a variety of GIS-based techniques. In some cases, the attributes assigned to each location can be transferred directly from a corresponding administrative record. For example, if health events are geocoded to points, qualitative or quantitative individual level information can be assigned to that point. Alternatively, attributes may be calculated from one or more point locations and assigned to area aggregates such as census tracts or zip codes (e.g., to permit the calculation of rates). In other cases, researchers may wish to work with individual-level data to create continuous surfaces. Such surfaces can be created through the use of kernel density methods. When points, areal units and surfaces are used, the population attributes associated with the geocoded information can be related to and analyzed with other information using GIS software functions. When researchers start with individual-level information they have the capability to determine the level of aggregation that is most appropriate to their work. This flexibility comes with a price, however, since results will vary with changes in scale. This holds true for aggregation levels as well as for the type of kernel filter that is used to produce a surface. It is also important to note that if individual-level data is, in any sense, confidential, caution must be exercise since it is possible to invert geocoding transformations to reveal individual level information. [PDF]

John del Corral, M. B. Blumenthal, M. Bell
Spatial Averaging Over GIS Features of Time Dependent Climate Data to Evaluate Societal Risk
GIS has become an integral part of the multi-terabyte IRI Data Library of climatological datasets. We bring together geoscience datasets geo-located by latitude and longitude, social science datasets geo-located by geographic feature, and GIS vector representations of features. The Data Library becomes a framework for the integration of climate data and model results with geo- and social science information. Open Geospatial Consortium (OGC) standards are used to store GIS data in an open-source PostGIS (GIS under PostGres) database. The Data Library makes queries and reads from this database. We can overlay the GIS layers on our datasets, and/or derive spatial averages of the data based on the polygons described in the GIS layers. OpenGL and postscript functions are used to view the GIS overlays and to make the spatial averages. We can do projection transformations on our GIS layers with the open-source Proj4 transformation library. We have developed GIS functions and tools to help us analyze datasets from the perspective of geo-located political, hydrological, agricultural, and epidemiological boundary areas. This capability provides a powerful method of correlating geoscience data with social science data and data from other disciplines. For example, gridded rainfall and temperature from models, and remotely sensed Normalized Difference Vegetation Index (NDVI) data have been successfully averaged over administrative boundaries in Africa, South America, and Asia. We generate time series of climatological data and model results that can be directly compared with agricultural and epidemiological data based on administrative boundaries. We also calculate time series of data over river basin areas, and climate zones. We provide valuable information for epidemiologists, agronomists, water managers and disaster relief agencies. A browser-based web interface is used to view, analyze, and extract data from the Data Library. We create topic focused GIS web tools for researchers from various disciplines. [PDF]