Speakers

Almost 20 years ago, Stan Openshaw called for learning how to live with error in our (spatial) data. However, our progress along this direction has been minimal. Today, most analyses using spatial data, ranging from simple mapping to spatial pattern detection, still assume the data are reasonably reliable or the reliability level of data is about the same across the study region. These assumptions legitimize the use of relatively simple and straight-forward methods to analyze spatial data. Unfortunately, these assumptions are the exceptions rather the norms, as most spatial data are statistical estimates derived from sampled observations and therefore each estimate has statistical error, which is often not negligible and not spatially uniform. Considering the error of these estimates even in simple mapping exercises can be quite complicated, not to mention the more advanced spatial statistical methods. In this presentation, I will use two examples to illustrate how error in statistical estimates has been incorporated in the use of spatial data. First example is in choropleth mapping. The class separability classification method, which considers error of estimates in determining class break values for choropleth mapping, will be reviewed and illustrated. Its implications in detecting spatial clusters will also be elaborated. The second example is in evaluating spatial autocorrelation (SA). A new global SA measure based on the Bhattacharyya distance, which compares not just two estimates but also their variances, has been introduced. The presentation will briefly discuss this measure. Future directions will also be identified.

Changes in land cover and land use can occur in response to both human (e.g., changes motivated by economic derivers or public policy) and natural drivers (e.g., weather and climate variability, geologic events). Changes in land cover can also impact local to global scale weather and climate by altering the flow of energy, water, and greenhouse gases between the land and the atmosphere. Land change science is about the understanding the interactions between people and nature that lead to changes in the type, intensity, condition, and location of land use and land cover. Remote sensing data have been widely used to map and monitor land use and land cover change and to distinguish between human and anthropogenic change. The U.S. Geological Survey (USGS) has a long history to produce land cover dataset. Several major land cover and land cover change datasets have been produced by the USGS Earth Resources Observation and Science (EROS) Center including both international and national scale products since early 1970s. Systematic large area land cover monitoring has become prevalent in recent years after Landsat data were made available free of charge by the USGS. These products have contributed to understanding past and present land use, land cover, and land condition change needed for variety research and applications including natural resources management and land change assessments. However, there are increasingly demands for more timely, accurate, and relevant land change products across government and academic organizations. Land change (specifically land use and land cover) dynamics is one of 21st century grand challenges on human and natural systems, including contributions to anthropogenic releases of CO2 to the atmosphere, changes in hydrologic dynamics, modifications of terrestrial habitat, disruptions of species migration, and role in the spread of disease vectors. In this presentation, products of several main USGS land change products including National Land Cover Database and recent Land Change Monitoring, Assessment, and Projection initiative are introduced. The efforts and accomplishments we have made in using remote sensing data to characterize land cover condition and change are explained. Challenges and opportunities for producing future land change product are also discussed.

Unmanned Aerial System (UAS), drone operation, has been recognized as a new tool in conducting field research and business in recent years. Continuous improvements in drone capabilities, including maneuverability and sophisticated sensors, have opened up enormous possibilities for business start-ups and researchers with creative minds. However, when any new technology is rapidly emerging, those possibilities come with many questions and hurdles when users actually try to implement that technology into their projects. In the case of drones, those questions range from drone project planning and budgeting, drone project capabilities, flying techniques and training, data capture methods, available sensors, data processing and analysis including uncertainty and data errors, data visualization, available applications that connect drone data and GIS applications, regulations, and job opportunities in the drone industry. In this session, Mr. Schulmeister, a UB Department of Geography alumni, presents his experience in the real-world drone service industry and provides practical guidance to students and faculty about successful implementation of drone technologies in their projects. This seminar includes a drone demonstrations!

Unmanned Aerial Vehicles or UAVs have risen through the ranks to become uniquely useful tools in various remote sensing and humanitarian applications. While fixed-wing and multi-rotor UAV configurations and their stand-alone operation have become commonplace, their operational envelope remains limited. These are often ineffective in dealing with complex mission requirements (e.g., in search and rescue, cargo transport, and surveillance) that involve some combination of the following: take-off/landing in constrained spaces in a high endurance mission, extreme environments, time sensitive wide area coverage, and map/track spatio-temporally evolving events. This talk will focus on two very distinct areas of research that are addressing these challenges, namely hybrid UAVs and UAV swarms. Both offline reconfigurable and online-transitioning UAV morphologies have been developed in the ADAMS Lab. The latter, named BITU, is capable of transitioning between VTOL, hover, and fixed- wing like efficient forward flight. A computational design framework, which integrates aerodynamic modeling, flight dynamics, uncertainty analysis, and non-linear optimization, has been constructed and tested to design variants of BITU that can meet challenging mission requirements – offering >100 km range with 2 kg payload and VTOL capabilities. On the other end, fully-autonomous decentralized control of multi-UAV flight is being developed to serve time-critical applications such as offshore oil spill mapping and detection of flood victims over a wide area. A novel swarm-intelligence inspired waypoint- planning algorithm, that also incorporates anomaly detection and probabilistic information extraction, has been developed to allow collaborative operation of small UAVs in a manner that is adaptive to their computing and wireless-communication constraints. This talk will also briefly touch upon some of the related work in the areas of modeling, optimizing, and testing wireless communication between moving nodes (e.g., UAVs) and participation (with 3-UAV flight) in a full-scale emergency drill with Buffalo Fire and Police departments.

The beginning of this century marked a milestone in human history. For the first time, more than half of the world’s population lived in urban areas. This trend is expected to continue into the foreseeable future with 6.3 billion people projected to live in cities by 2050. This rapid urbanization will place unprecedented pressures on urban systems and their ability to provide basic of services. To plan for this future, we need to better understand the inherent complexity of urban systems from social, economic and environmental perspectives. In this talk, I will explore how such understanding can be gained through the lens of computational social science (CSS): the interdisciplinary science of complex social systems and their investigation through computational modeling (e.g. agent-based models) and related techniques. Through a series of example applications, I will demonstrate how new forms of geographical data (e.g. crowdsourced, social media etc.) not only provide us with a novel way of analyzing urban systems but how such data can be integrated into geographically explicit agent-based models. In addition, I will highlight that by focusing on individual, or groups of individuals, leads to more aggregate patterns emerging and show how model outcomes can be validated by such datasets. After these demonstrations, I will outline the challenges associated with this program of research, as using such data is not without its difficulties. Together, this work provides a brief overview of the current state of analyzing and modeling urban systems through the lens of CSS.

Studies on a wide range of urban issues by geographers have been conducted largely with a static perspective based on people’s residential location. However, since people move around in their daily lives to undertake various activities, their social encounters and exposures to environmental influences also take place in neighborhoods beyond their residence and at various times. Human mobility is thus an essential element of people’s spatiotemporal experiences, and these complex experiences cannot be fully understood by just looking at where people live. Ignoring people’s daily mobility, the time they spend outside of their residential neighborhoods and their exposures to environmental influences and other social groups there omits a considerable part of their everyday experiences. In this presentation, I draw upon recent conceptual and methodological developments to examine how a perspective that integrates the spatial and temporal dimensions and takes human mobility into account can help identify the relevant spatiotemporal context that influences people’s health behaviors or outcomes. Using examples from my recent projects, I discuss how the collection and analysis of high-resolution space-time data enabled by advanced geospatial and mobile technologies can provide new insights on the relationships between people’s health behaviors and the complex spatiotemporal dynamics of environmental influences.

China’s economy had witnessed double digit growth following the adoption of open and reform policy in the late 1970s. However, China’s natural environment did not improve with the economy. In fact, China’s eco-environmental conditions went in the opposite direction with the economy for decades, leading to devastating natural disasters in the later 1990s. As a result, the Chinese government implemented a series of forest restoration and conservation programs to improve the natural environment. The Conversion of Cropland to Forest Program (CCFP) and the Ecological Welfare Forest Program (EWFP) are two of them. CCFP program is the largest reforestation program to date in the world, involving 32 million households and 120 million people in 25 of the 31 provinces in China. China’s forest cover increased 3% as a result. EWFP is a program that preserves natural forests that provide essential ecosystem services. Both CCFP and EWFP are essentially payment for ecosystem services program. Despite nearly two decades of implementation, the programs’ socioeconomic as well as their ecological effects are not well understood. In this talk, I will present the recent findings from a US-China collaborative project studying the impacts of CCFP on the dynamics of the coupled natural and human systems in Anhui, China. Riding the tide of overall economic growth in China, both CCFP and EWFP have been successful in converting and preserving the land-use, and have exerted profound impacts on rural residents’ livelihoods. I will focus on the program effects on cropland abandonment, fuel wood use and rural out migration in this talk.

Geography has experienced a transition from a long-standing Era of Geographic Data Scarcity (Greek classical period – 20th century) to a new Era of Geographic Data Plenty (21st century and beyond). Geographic data collection used to be focused, time-consuming and expensive, resulting in small but thick databases. Geographic data collection is now open-ended, quick and cheap, resulting in big but thin databases. In this talk, I will ask: How should we do things differently in the Era of Geographic Data Plenty? And, just as important, what should we not do? In seeking answers to these questions, I will share three big thoughts: 1) Opportunistic GIScience: Experiments are not just for labs anymore; 2) Mesogeography: There is a middle path to geographic knowledge; 3) GIScience, fast and slow: Some decisions should not be as fast as our data.

GIS, after continual development in last 60 years, has been widely used in various fields by researchers, governmental officials, businessman, and many professionals and non-professionals. With its root from maps, GIS has more functions including spatial analysis and static spatial modeling. However, many GIS users today are looking for a platform which is of geo-process modeling functions, such as wild fire modeling and air pollution spreading simulation. The framework of GIS with a geo-coded database shows its bottleneck for this kind of dynamic modeling. What should we do for integrating the geo-coded database and the geo- process models? Virtual geographic environments (VGE) could be an answer as a new framework beyond GISystems.